ML20090D879
| ML20090D879 | |
| Person / Time | |
|---|---|
| Site: | Oyster Creek |
| Issue date: | 03/03/1969 |
| From: | GENERAL PUBLIC UTILITIES CORP. |
| To: | |
| References | |
| 90885, DC-31, NUDOCS 8303110161 | |
| Download: ML20090D879 (31) | |
Text
. .-- _ _ _ _ _ _ - ___--- _ _-__-_---_-- _ _ _ - - _ _ _ _ -
- .Ln A& AL 342k
, FOR OFFICAL f) g.g USE ONLY.
,U 3
fo METALLURGICAL EXAMINATION & REVIEW OF CASTING DEFECTS IN Ti!E STAINLESS STEEL VALVE BODIES LOCATED IN TliE CLASS ONE SYSTEMS OF TiiE OYSTER CREEK REACTOR No. DC-31 March 3, 1.909 Q
T.
A s t;,s j
V s\$$
0 9 e > , 1 Hc 'E R'4 9@ j .)
'q .
<~v ; e
'J!d,I 19 RZTURN TO
, DIRECTORATE OF REGULA TORY OPERATION Q FOR OFFIGAl.
f; 4 USE ONLY q oggg 8303110161 690303 ~
-i v ..e
3 gp, 3 g 4 ' 1.gw,: s-a-
- -.. ._- ~ =. .z. . . a . :. . .. .. u : ~=m e - ' > - ' " ' " ' ' ' '-- ~~ -
I Page 1 of 9 FOR OFFICIAL U USE ONLY.
Metallurgical Examination and Review of Casting Defects in the Stainless Steel Valve Bodies Located in the Class One Systems of the Oyster Creek Reactor March 3 ,1969 No. DC-31 I
i t \
'%./
1 4
} Prepared for: AEC Contract No. AT(11-1) - 1658 I Task A Parameter No.68-69A; Subcontract No. 7 I
i Prepared by: R. G. Gilliland Assistant Professor of Materials Engineering University of Wisconsin-Milwaukee 4
' Milwaukee, Wisconsin 53201 i
d, l
Prepared through: Parameter, Inc.
Consulting Engineers
't Elm Crove, Wisconsin S
9
,a v. o.
FOR OFFICIAL if '" USE ONLY i.i
- 4
+
.o
- 2. Peport No. DC-31 O
DISTRIBUTIO:1 COPIES 1
AEC Regulartory Organization Division of Compliance Washington Office (3)
R. II. Engelken, L. Kornblith, G. W. Reinmuth, H. R. Denton Region I, Newark Office (1)
R. Carlson AEC Contracts Division Chicago Operationn Office (1)
- 11. II. Miller, E. IIalman PARAMETER, INC. (1)
R. G. Gilliland (1)
O
._ _._..m_..__..._ _ _ _ . _ _ _ _ _ .._ . . _ . _ _ . _ __ _ ._ _ _ _ __ _ .h
- 3. Report No. DC-31 O
Introduction At the request of the Region I, Newark Office - Division of Compliance, the author visited the Oyster Creek Peactor site near Toms River, New Jersey on February 12, 1969. The purpose of the visit was to review the evaluation and appraise the repair schedule of linear defects observed in cast valve bodies which were a part of the primary and safeguards service system of the reactor.
The author was accompanied by:
Mr. R. Carlson, Senior Reactor Inspector - Region I, Newark Office-Division of Compliance USAEC Mr. W. J. Collins-Division of Compliance-Headquarters USAEC-Washington, D.C.
Mr. R. M. Gustafson-Division of Reactor Standards ,
USAEC-Washington, D.C.
During the visit the following persons were contacted:
L. Loeb - GC - APCD J. W. Eberle - General Electric Company Bob Huggins - Principle Project Engineer - Oyster Creek-GE Nel Strand - Site Construction Mgr. - Oyster Creek Tom McClukey - Plant Superintendent - Jersey Central Power and Light Company (Utility)
Thomas E. McSpadden - MPR Associates, Inc.
The prime contractor is the General ElectricCompany and their Atomic Power Equipment Department (APED) is involved. The utility is the Jersey f3 Central Power and Linht Company, and their consulting engineers are MPR V
i n ~o
4 Report flo. DC-31 p
Q/ Associates, Inc. of Nashington, D. C.; Burns and Roe are the architectural engineers.
The valve bodies in question are cast austenitic strialess steel, Grade CF8M, which corresponds to wrought stainless type 316. They range in size from 4-in. to 16-in. in diameter and are distributed in the follow-ing system:
- 1. Core Spray
- 2. Channel Cooling
- 3. Clean-up Domineralizer 4 Control Rod Return.
There are a total of 43 valves involved.
Forty (40) of the subject valves were purchased by the architectural engineers, Burns and Roe, from the Anchor Equipment Company to a specifica-3 (j tion which required that the valves be 100S inspected by radiography to class 2 of ASTM E-71. Two separate foundries, each with their own separate inspection contractors, were employed by Anchor to furnish the valve castings.
A routine visual inspection of a valve revealed that a casting defect had been uncovered during the pre-installation machining of the cast part.
This prompted the review of all valves above 4-in.in size in the class one systems of the reactor. These inspections revealed that seventeen (17), of a total of forty-three (43) involved required repairs of defectn to one degree or another. It should be remembered that this review was performed on in-stalled valves which had already been hydro-tested. Three (3) of these seventeen valves were found defective to such an extent that replacement of i
the valve was necessary. Three (3) other valves from this group of seventeen required weld repairs. The remaining eleven valves could be satisfactorily
< ~s
(\ 'i repaired by only surface grinding. '/isual and dye penetrant inspection tech-niques were used to entablish the integrit y of all valven beinn_ reviewed. . , , . - , . - . - - -
"N# *W ~ M 'M 9
- e, ,y , _ ,
- S' @e'WD' y hWN epemg.,,,__ _
' 'I I *'Nm'*'*9 *
} -. $ p ,,iy .
9 ,g ,
.y,,,ngeem 1 c- , n a .- -. r f - .~ 2;mm u ,n.2 .} , ,
, ,. m.:.A.,3..: _, m. ;._ ,, ag. a i i Aa ._ . ;- ,,
- 5. Report No. DC-31
/ ;
L/
Radiography was used to establish the as-repaired integrity.
The extent of the casting defects found and the fact that all defective valves were purchased from the same vendor ( Anchor supplied forty out of the forty-three valves in question - APED supplied three) prompted a review of the original, specification radiographs submitted by this vendor. The contractor,the General Electric Company, elected to review the radiographs of only 21 Anchor valves; their selection was based on service requirements and part size. The results of these studies confirmed and supplemented the visual and dye penetrant inspections performed at the reactor site and out-lined above. The valves which were, or will be, replaced were identified as:
No. V-20-41
( ) No. V-14-34
\~
No. V-16-2 Those valves requiring weld repair were:
No. V-20-17 No. V-17-19 No. V-17-54 A summary of the radiographic review by the General Electric personnel is attached as Exhibit A, and is the reviewer's initial statements prior to repair initiation (the replacement and weld repair outlined above was deter-mined at a date later than this memo). A preliminary statement of initial impressions relating to the Oyster Creek site visit on Februar" 12, 1969 was prepared by the author, and it is attached as Exhibit B.
n R ../
- 6. Report No. DC-31 O
Discussion The valve originally detected as being defective, No. V-20-41 located in the Core Spray system, was finally replaced. Sections were taken from the defective valve and submitted for metallurgical examin-ation. The author examined these microsections at the GE-APED labor-atories in San Jose, California on February 20, 1969. Mr. Robert Dodds of the Division of Compliance - Region V - USAEC was present as we met with Mr. Loeb and Mr. John E. Leatherman of APED-GE and Mr. William R.
Schmidt of MPR Associates.
A close examination of the metc.11ographic specimens taken from the top surface of the rim was performed by the author. These ser.tions were taken from the area of the visually observed defects and were located in
() both longitudinal and transverse orientations. The valve had received, prior to machining, a homogenizing treatment of approximately one-half hour at 20000F, followed by a water quench, An example of the defects observed in these microsections is shown in Exhibit C. This micrograph shows that these indications are a com-bination of microshrinkage voids and hot tears. More specifically, it appears that the microvoids (shrinkage) are essentially linked by hot tears. These defects were metallographically observed, in some cases, to be greater than one-inch in length. No evidence of excessive mold-metal reaction was metallographically observed.
A preliminary statement relating to these metallographic examina- I tions was prepared by the author, and it is attached as Exhibit D.
O t
e*
- t.
9 7 g 7 a q g -gy g g aparg jey m m .,..,...-w...,, - - , .
n _ _ _ _ _ . _ , _ _ _ _ _ . _ _ _ _ _ _ _ _
- 7. Report No. DC-31 1
O A properly designed casting, a properly prepared mold, and correctly melted metal should result in a casting having no defects. However, the combination of perfectly executed events such as these cannot consistently be expected, even with the utmost in decign. Thus it is not unusual for casting to have defects and particularily those which form during the
[ solidification process. It was these types of imperfections which were l
observed in this case.
A hot tear is an intergranular failure (along the grain boundaries) that occurs while the casting is at a high temperature within a mold. A hot tear may form in a larger section of a casting that is at a temperature just above the solidus temperature and is subjected to tensile strains in-duced by the solid contraction of an adjacent, already-solidified thinner j
() section. While the mass is just above the solidus temperature, it consists of grains of metal that are surrounded by a thin film of liquid that cannot support a tensile load. Thus, a parting of the metal grains at the grain boundaries results, leaving closely spaced, or even linked, microvoids.
After the area completely solidifies,but is still near the solidus temperature, the small spaces between the microvoids often fail or crack as a result of solid contraction. Other microvoids can form from the natural contraction during solidification and areas of liquid starvation.
Therefore, the point here is not that the castings had defects in them, but that these defects were not detected and repaired. A personal review, with Mr. Collins, of the original vendor radiographn of selected valves re-vealed that in all cases at least the major defects (those which ultimately resulted in valve replacement or weld repair) were clearly visible.
- I'fMIh [ M'N D4] f "' &$ ++
$?$ - **ff=." ' ' - " = = * =
~
y , e.g i . e w , ,- -q -,
m ,r~ wvm m .u.. ._ -
,a: u, . , .,w,, , :. . e,.a u . - .. 2. x .>_p h w e n 1:.:.2.c
- 8. Report !!o. DC-31 Weld repair procedures and techniques being performed on the three valves mentioned above were found to be most satisfactory. The good practice of very low energy input, guaranteed by a low interpass tem-perature and a stringer-head welding technique, was being closely main-tained. The welding was being performed using the tungsten-inert gas process with a bare, type 316 stainless steel filler rod.
Conclusions Based on the above facts and observations made during the inspections performed on 12 February and 20 February,1969, the following conclusions can be made.
- 1. The casting defects found in the valve bodies installed in the
() class one systems of the Oyster Creek reactor were metallographi-cally determined to be the result of solidificat ion shrinkage and hot tearing and were manifested as linked microvoids.
- 2. The metallographic examination revealed no excessive mold-metal surface reactions.
- 3. Pre- and post- heat treatment radiography indicated that these defects were truly casting imperfections, and that the heat treatment sequence did not propogate the defects.
4 The defects observed, both radiographically and metallographically, are common to the casting process. Such defects are frequently expected in designs such as these, and the fabricator usually tries to reduce their occurance by gating and risering techniques and, when required, weld repair.
- 5. The real point here lies in the breakdown in the control of the review and analysis of non-destructive test results. Had this situation not existed, defective parts would not have gone undected.
- 9. Report No. DC-31 O This fault must be borne, in a major way, by the primary valve contractor, the Anchor Equiprent Company. However, it is felt that the prine contractor, the General Electric Company, and its architectural engineering firm, Burns and Roe, must jointly be held responsible for their negligence in not maintaining, by con-tinuous review, the quality assurance of their vendor. Had the code-required procedures and specifications been properly adhered to the current problem could not have existed.
- 6. The procedures and techniques of weld repair of the casting defect, and subsequent nondestructive testing, are considered to be proper and correct.
- 7. Finally, it is considered important , if not critical, that the re-mainder of the Anchor valves in the primary system (totaling 19) be reviewed similar to those already examined. This statement is supported by the results of the re-examination of the other 21 Anchor valves, considering that this review caused major repairs to be required on 25% of the parts in question, which had pre-viously been termed as meeting Code requirements.
It is the author's opinion that the General Electric people are making a reascnable effort to correct what is a bad situation, and that they have a good understanding of the problem, have performed an adequate analysis, and are aware of its cause.
% g h=
O .
g '
.,4., .
r' g 9 " . *', * *e ,9,,.. ,
s%,[p~# 7 EX11IBIT A
. u p ,
- - re : , e p /J , ,. r r r. . * '7
- N U C l. E A R E t1E R G Y
- ', i , . ', 7 ;9 . ..' ) ! I a C . .'. . .. C . .' / . J ('- j hl
- DIVISIOfl ,
b*d 16** O . .1r
. .* .\ y 1 i
( ; g
(/ 173 CURTN;R AVI., SAN JO5!, CALIF. 9512 5 .. ARE A CODE 408, TEL. 297 3000 l .
TWX No. 910 3M 0116 ATOMIC POWER EQUIPMENT sEv'ARTMENT Februa ry 12, 1969 RF, PORT OF RT-REVIEW AND SITE-ll'5?ECTJON OF ANCHOR STAINLESS VALVE BODY-CASTINGS FOR, PRIMAR'l AND SAFCGUAhDS SERVICE IN OYSTER-CREEK NUCLEAR POWER PLANT
___ L_
l %' ('. -f .
y
~ k (. j r~
V. s
~
n,
/t.4e fW/..
l D. E. Tackett E. H. Franks Field Applications Engineering, N.S.P.6P., GE-APED Summary
- At the request of S. Naymark, Manager, DTP Project Engineering, we have co'n-pleted a review of radiographs for all stainless steel body-castings in primary and critical safeguards service at the Oyster-Creek nuclear plant. The fellowing valves required replacement, najor repair or re-radiography for the reasons noted:
()
V - Valve.V-20-41 is being replaced due to defects found during. site inspectica; these defects were such that a field repair was not practicabic. We have reviewed and accepted the radiographs for the replacenent body-casting.
- Valve V-14-3'. is being replaced due to defects found during exploratoig grind-ing of a surface indication found during site inspection; field repi.;r w uld have cauced distortion of the seat ring. Radiogcaphs for the replacenent body-casting will be revleucd and accepted prior to shipment of ti.e valve body from Anchor.
- Valve V-20-17 requires repair of unacceptab2e defects found during RT-review f of origit.al-filn in Sea Jose. Exploratory drilla tig in ;pect ion a t site con-fitned that field repair has very good chance of success. A replvcement casting has been ordered. ' his .'alve body vill be either repaired t o an i
acceptahic condition or will be replaced.
- Valve V 37 lached cenplet e RT-coverar,e due to two fil.as that uere missing..
The portion l acking cove ra;',. consists of two arcar each about 3-inches by 10-inches. There arcan wil.! be re-radiographed by the site in accordance with a shooting ske tch provided by D. E. Tackett. The retake filns vill be reviewed by Tackett/ Franks and will be filed wis.n the original filra. No repair action is anticipated due to satisfactory quality shown in adjacer.t radiographs.
v e
./
e,N P y-*w : ~ ~,s s.
4o . +s s - ...y. .. . . , , .
.o
v
. - v.,w ~ ~ . m u.u - :3-~- J.T- . a n k .% -. *
.a eu &hi .a.X15.' h '. .~ x
/ -
lf*
i i v ,
- Valve V-16-2 had an area of unacceptable inclusions that was rarked for repair on the original filn.s.
RT- review of original / repair filns in San Jose failed to confirn that the repair had renoved enough of the inclusion-area to result in an acceptahic condition. Also, in another area of the same valve body, original filns showed hot-tear indicationr. for uhich retake is required to evaluate as-Installed condition of the casting. The affected areas of the body-casting will be re-radiographed by the site in accordance with a shooting sketch provided by D. E. Tachett. The retake filns will be reviewed by Tackett/ Franks and will be filed with the original film. If repair is r equired (either or both areas), field repair should have a good chance of nuccess. Based on review of partial-coverage retake-films from the site, it appears that repair of the inclusion-area will probably be required; final recommendation must avait review of full coverage retakes of the questionable areas.
Valve V-17-19 had a borderline-acceptance hot-tear indication in the original film that could not be correlated with any of the reported surface-defect repair-areas at the site.
The size and location of this indication is such that an as-installed retake radiograph is required to evaluate the adverse ~
effect of the defect on service integrity of the valve. A retake of the questionable area is being made by the site in accord with a shooting sketch provided by D. E. Tachett in telecon to R. Hayne:, (2/10/69). Final recom-p menda t ion tais t avait revier by Tackett/ Franks of site retahen.
Y]'-
Based on the Tackett/ Franks RT-revleu in San Jose, all renaining body-castir.gs were found, to be acceptable by the standards of AST:' E-71, Class 2. All rem ining body-castings were also found to be acceptable by liquid penetrant inspection of the outside surface at the site; sor.;c valves required grinding or ninor repair uelding to remove shallou surface defects. Upon coTpletion of the above-not ed repair /re-inspection / replacement, based on the RT-review in San Jose and on the satisfactory results of liquid penetrant inspection performed at the site, we (Tackett/ Pranks) consider that the subject valves hcve trore than adequate structural soundness to fulfill all design requirements of the prirary and engineered safe-guards service-applications where these valves are installed.
fBCkrmul d.
All of the cubject valves were purchased by Burns and Koe f rom Anchor en MR purchase order BR-2299-92 to neet the requirenents of E&R specification S-2299-61.
Paragra;>h 5.h of the lat t er required that body-castings for the r.ubject valves be 100?. inspected by radiography to Class 2 of ASTM E-71. Th e "'. a l ve No . " l i s t ed in the accompanying table is the systen identification as shown on the PLlD drawing. The "l ten ::o." J isted in the cable corresponds to the item nuc.ber in the purchase order (nu-2299-92) and specification (S-2299-61) for the subject va2ves. A1] of the subject valve bodies are cast sustenttic stainicss steel, Grade CFSM, which corresponds to wrought stainless type 116. Castinga for these valve bodies were obtained by AucLor from one of two foundrie. Roe.ne r d.- Vo l can .
Cr. Lings poured by Room. were radiographed :oed liquid-penet rant inspected by
()
Conam.
Castings poured by Vulcan were radiographed and liquid-penet rant i n s pe c t.ed by X-Ray Engineering. Weld-repair of castings (when required as a result of original
\.
a, ,-- ~ _ _ - - ~ _ . - _ _ _ . - . . . . _ _ - ~ . - - . . - . - -
_3..
p radiography) was perfor;ed by the foundry that poured the casting. The foundry V responsih]e for each valve hady-canting is listed under "Feundry" in the tahic.
'he foundry sta yed cach body-canting uith the " Heat No." and "Dody No." as listed in the Labic, which combination constitutes a cerialization of each easting. Radiographs are idcatified to the casting by the heat / body cor.bina-tion. To ensure that. the radiographic fila reviewed was of the body-casting for the listed valve, the heat / body identification was obtained directly from the stamping on the installed valve; it is this identification which is listed under "licat No." and " Body No." in the tabic. 'ihe heat /bedy identificat ion
/ so obtained exactly agreed with the radiographic film identification except for valve V4 6-2 (see Note 12 of the tab]c). In the case of valve V-16-2, the body-casting was stauped heat "4-8" hody "1" by Roemer, whereas the radiographs, inspection reports, and material test reports are identified to heat "4-48" body "1"; Anchor will transmit to us a letter certifying that the body-casting stamped heat "4-8" body "1" is, in reality, heat "4-48" body "1". In addi. tion, comparison of defects in the site-retake films with defects in the original films positively identifies the installed casting to be the same casting originally radiographed as heat "4-48" body "1".
Site Inspection A suspicious surface-defect uns noted during visual inspection of valve V-20-41 after installation at the site. Liquid-penetrant inspection of the affected area (near bounct rin) revealed radial crack-like dcrects across the t.op surf ace of the rin; some of these indications continued doun both inside and outside unchined p surfaces. Careful exauf nation of original film in this area o,f the casting V reveals a hot-tear into the top of the rim. Caref ul visual exanination (at SX) of these defects by D. E. Tacket t (durlog si t r visi t on 1/31/69) shoved there defects to be hot tears; such defects occur on initial cool-down of the casting juct belou the freezing temperature. These defects should have been detected by the original liquid-penetrant inspection and rejected at that time (see belo'.'
under Conclusions). The hot tear indication shown by radiography in the original film is considered to be a Class 2 borderline-acceptable condition. Due to the nur.ber tnd location of defects in V-20-41, field repair uas not practical and a e repl:.ce.acnt body-casting was ordered. Liquid-penetraat inspectim of the outside surf aces of all stainless-steel body-castings of Anchor valves in primary and safeguards npplications was perforned at the site. Indications which required grinding to obtain acceptnble liquid-penetrant test results are reported for each valve under the colunn headed "No. Surface Indications Explored" in the tah]c.
Except f or valve V-14-34, either no repair welding was required or only minor a repair welding was needed. In the case of V-14-34, grinding and liquid-penetrant inspection revealed defects near the scat-ring ueld to an extent that field repair velding would have resulted in excessive seat distortion and would have required access from inside surface; therefore, it was decided to replace the body-ccsting instead of attempting field repair. Three valves, V-20-12, V-20-18 and V-17-54 were selected at random from the subject valves for re-radiography at the site.
These site-retake films have been revicwed by , Tackett/ Franks and shou the s .nc identical defects that were revealed by the eriginal films; all defects in both original films and site-retake films for V-20-12, V-20-18 and V-17-54 are a.ceptahic to Class 2 of ASTM E-71 (except see Note 17 regarding rejectable hot-tear in G V-17-54 which was repaired prior to site retakes). -
V 6
2 1
- v .
l m J
l' L
l l t
f>
Radiography Revleu Radiography films made by Conam and X-Ray Engineering of the as cast body-l castings for the subject valves were obt ained f rom Anchor for review by Tackett /
i Franks in San . lone. The total film-record for the subject valves consists of l several thousand films (including body castings, bonnet castings, cast stems, j and discs); out of several thousand film only two film (Dand F on valve V-14-37) j could not be Jocated, requiring partial-retake of that valve at the site (see j[ Sunnary above and see Note 10 of the table). Of the film provided, al_1_ films of the body-castings were reviewed except for radiographs of the inside throat (the inside throat of these valves does not, constitute a pressure boundary and i does not contribute to body strength; its function is merely to support the seat j
ring). Radiographs of bonnet-covers, discs, and stems were not reviewed.. All indicatiens in each filia (including acceptable defects, artif acts, and surface-contour shadows) were recorded and evaluated as to relevancy, acceptability to Class 2 of ASTM E-73, and effect on the structural soundness of the casting.
, Each set of film for each body-casting was checked for appropriate identification j, (heat / body nunbers) to the systen valve and for overlap of adjacent exposures j to ensure that full RT-coverage was obtained. The following definitions apply l to the significant defects noted in the subject valve body-castings:
i?
5
- Shrink. This defect consists of an irregular-shaped multiple-branched void in the metal caused by local liquid starvation during freezing of the car. ting.
Shrink is connon in and near rims, ribs and other section-changes where there i hq is a possibility of thinner sections freezing solid while th' icker sections
- are still partly molten; the result is blockage of flou of liquid metal from i the risers into the thich section to compensate for the volumetric contraction that occurs at freezing. Since the metal freezes inmediately at. all surfaces
! of contact uith the nold (due to the chill cffeet of the cold nold) forning j! a continuous skin from which the solid-netal grains grow toward the center of the thicknens, shrink tends to be unar the center of the section thickness l and is rarely open to an au-cast rurface.
I Cas. Yhis defect consists of empty voids in the netal due to trapping of gas when the netal freezes. Gas-voids are globular (blow-holes) or elongated-l round (worr.-holes) in shape and are not branched. The liquid metal hau a higher solubility for gns than the solid metal; therefore, considerable gas l is released during freezing. If there is not an open path for this gas to escape through liquid metal out of or into a riser, the gas vill remain trapped and form a gas-void in the solid metal. The same factors which promote shrink also prouote gas entrapment; therefore, these two types of defects are fre-quently encountered together. For the sane reasons as for shrink, gas voids are rarely open to an as-cast surface.
- Incl usions . This defect consists of filled voids in the metal due to trapping of nold-sand. The hot liquid n.etal is very erosive to the surface vf the sand mold. If the netal is poured at just the right temperature and rate of fica through the uold-channels, and if the nold-sand has been properly packed, a o thin glass vill form on the n.old surface and prevent entry of significant d sand into the molten metal. Pouring of the metal either too cold or too hot, P
- i e
f
...g,7g_ .
O
\*/
or pour #r; at too faat a f3ow rate through the t.cid (too s3ou a pour vill resul t in internal chills), results in washing away of the surface of the sand nald; the cand thuu vashed away is trapped in the metal (normally in the form of a glassy inclusion or partly funed sand). Trapped sand is frequently associated uith small shrinks (tails on the inclusions). Sand trapped at the metal surface results in a cand-hole.
Hot tears. This defect consists of an intergranular separation when the metal is stressed as the casting cools doun fren the freezing temperature.
Hot tears occur near the freezing /nelting temperature. Due to composition unbalance, and also due to' local composition-variation due to freezing process,
. al] the t'alten metal does not freeze at the same temperature; the last r.etal to f reeze is in the grain boundaries and freezes at the lowest tenperaturd.
If the tetal is stressed during this critical temperature range (about a hundred degrecs be309 freezing temperature), the grain boundaries vill separate uhcre still liquid, resulting in het tears. Hot tears may or may not breck through one of the skins of the castEng, depending.on severity and location.
Hot tears are f requently found as radial cracks into rins due to the circum-ferent al stress that results fron the restraint of the mold-core; hot Jears i
are also found in thin sections between thick sections parallel to the ' change in section.
porderline-Acceptable. This is a classification of defect-severity according to ASTil E-71 radiography standards for castingn. The term in no way implies marginal st rength or questio.udile soundnent of the canting. 'norderline-acceptoble is were]y t he mamier of defining the Jimits of acceptability of various typea of defect s. Any defect or combination of defects that conctituted a coadition approaching the l'imits of accept abili ty of ASTM E-73 uns reported as " borderline-acceptable" during this review and van carefully evaluated regarding effect on the structural sounducss of the casting. Any, hot tear, regardless of size or location, was considered "Dorderline-acceptable" and carefully nt.udi.ed ar noted above; an exp]anation is givea in the table not.cs f or all hot-t ear indicat ions accept ed during the RT-revicw. Site retahr.
were requested for all hot tears that could bave an adverse effect on service integrity of the valves. In the acccupanying tabic, the number of borderline-acceptab]e conditions in each casting in reported under the columns headed
" Borderline-Accept able RY I ndicat ionn". Inclusiona and can holet are con-hined in reporting the noube r of boi derl i ne-accept abl e d4 foett nince such defects have a uiqor effect on the b m ic structural strength of the casting.
, To the maxinum extcut possible, the RT-revicu was correlated with the results of the surface inspections performed at the site. Many of t.he significant defects found during the RT-revleu were corrected and clininated during the site inspec-tion; where positive correlation of RT/PT indications has been nade, thin is reported in the appropriate note of the table. The not result is a considerable reduction in the overall nouber of " borderline-acceptable" conditions remaining in the as-installed valve b~odies. In addition, some re-radiography was performed at the site and the site retake-fi]us were carefully compared to the original im .
?. l
.. g . , . , _ . - . - . _. .. _ _
o e
{
r l p
!b
[ films to dctcruine whether there war, any significant change in the defects.
j Since the original fi]us warc nade prior to solution heat-treatm.zat of the l
castings, this conparison of site-retahes to original films constituted a conpari-son of defects before and after heat treatment. Radiographs of the follewing l' valves were compared to determine changes in defects that night have occurred during heat-treatuent:
Valve V-?O-17: Revicu of origi.ncl fi]ns revealed a severe shrink near the bonnet rin (this shrink was rejected by the reviewers, see Suggry, above);
site retake fi.1w of the sinne area also showd severe shrink. The total length of this defect was about six-inches, including a netuork of multiple branches. Careful conparison of the original and site-retake filns showed that the defect had not changed during heat treatment of the casting. .A cuestionable shrink near the guide-rib in the center of the body (this Indica-tion was also rejected by the revicueru, see Note 3 of the table) shoied on original filus; comparison uith site-retakes showed no change during heat-treatuent. There were ilo defect indications in the retake fiJns that did not show as identical defects in the original films. -
- Valve V-_]6-?_: Review of original films revealed an area about six-incbes in d i m..e t e r , nea r the center of the body on one side, having a high density of sand inclusions and small shrinks (this area was rejected by the revievers, see Sur maryy above) . The same indications would have been borderline-acceptable pV in a thicker casting; however, as located in this casting, the accounlation of defects van conside md unacceptable for the thin section between the sont-rings. Couparison of sit e-retaken with original filmn showed no change in the linear defects in this area. and shoucd no development of additional defects. Site retahes of borderline-acceptable hot tear indications in original files are being made but are not available as of this writing (see _Incomp_1gee L'ork below) .
- Valve V-20-18: Revieu of original fi]na revealed several gas holes of moderate size, with brenching suall-shrinks originating from cach gas hole (the conbination of defects was considored by the reviewers to be borderline-acceptahic to Class 2 of ASTb1 F.-71 and are reported in the table as borderline-acceptable under
" Gas l'ol es", evcn though the gas or shrink,.taken separately, would not have constituted a significant defect). These defects were located near the bonnet-riu of the body-casting. Careful couparisoa of site--retaken with original fiJus shoued no change and no deveJoptent of additional defects.
Valve V-20-32: The singic hot tear shewn on original filns was at top of rim and did not shou on site retakes; this tear was probably removed in excess nnterial taachined f rom top rim during finishing of the casting. The sin 3]c 3 art,c inclusion shoun on original films also shoued identically on site retakes.
Careful comparison of site-retakes with original films showed no change and no developacnt of additional defects.
m f ) .
v b
o
,-^ M**'y W;*rvt? ;r ', ,*,
n.7 g _5 E m b ~* .. e- -
m>
(g"
/
Valve V-14-34: A'ninor surface indication Icd to exct.vation into the thick sectica behind seat ring; this excavation uncovered shrinks / tears extending to insida surface (this valve-body is to be replaced; see Sun =rv above and see Note 7 of table). Site-retake radiographs of the (xcavated area show d exactly the sabe indications as were shown at thin location en the original filu; therefore, it is certain that the defects were due to the casting process and not due to heat treatment.
- Valve V-17-54. A full set of site retake films was compared to correrponding localions on the original filma. Also, sketches of defects removed by grinding or repair-uelding were conpared to natching indications in original filns. A shallow hot tear indication about 6-inches long (this should have becn rejected on the original film) exactly matched the surface defect 5:hich was excavated / repair-we]ded at site; retake-film now shou this area clear.
Similarly, all other ground or repaired surface-defects found during site inspection were matched to corresponding indications on the original film.
Also, minor buried defects in r c retake films were carefully compari e' to same defects in original filns; no change was noted nor did any additional defects develop. All of the foregoing provides strong evidence that there was no growth or addition of defects during the heat treatnent process.
- Valve V-37-19: A significant hot-t ear indication (borderline-acceptab]c) 11.
the original film corresponds identically to a surface defect ubich was excavated / repair-ucided at the site; the affected area is clear on site (v) repair-rctahes. This is evidence that the curf ace -defect or.iginated during casting and not during heat treatment. Site retakes of another border]ine-acceptable hot-tear in LMs valse-body are being made but are not availabic as of this writing (see Incomplete Uork below).
In all of the above, a considerable nuuher of linear defects vere coupared between original-f.ilu indications (as-cant) and as-installed indications (either l ie,n i d- p ene *_ ra n t or site-retake-filu indications). Yhere uns absolutely no evidence of development or growth of def ects as a result of solution-annealing hen t--t rea t me n t which uas donc after original radiogrnphy. Therefore, we (Tachett./ Prat.hs) coasider that the original ac-cast radiographs are fu))y representative of th'c as-installed condition of the subject valve body castings with respect to evaluating structural soundness and effects of defects on service integrity.
_I_nco glete Work .
The following work rcmains to be completed on the subject valves so that the body-cm. tings will have unquestionable attuctural-soundness for the intended service:
Valve.V-14-34: The radiographs of the replacement body-casting have not yet been reviewed by Tackett/ Franks.
m f .
\, /
7 y
i n
1 I f i V 3
Valve-20-17: The repair / reinspection / refurbishing of the rejected / ques'.ionabic areas of the body-casting are in progress; the final post-repair radiographs I
have not yet been reviewed by Tackett/ Franks.
I ,
- Valves V-34-37 and V-17-19: 'lhe required site-retake radiographs have not yet been revicued by Tackett/ Franks.
t
- Valve V-16-2: The required site-retake radiographs providing full-coverage of the questionab]c areas have not yet been reviewed by Tackett/ Pranks; repair will probably be required.
1 1
2
. Conclusions f
/ Some discuc,sion of the original intpection of the subject valves appears to be
] needed. If, as we maintain, t.he surface defects found at the site existed in j the valve bodies fron the tine ti sy vere cast, why were these defects not j detected and corrected at the original Jiquid-penetrant inspection? Also, how t uns it possib]e for radiographa to be accepted when the illus showed obvious
- indications of cicar3y unacceptable defects? Uc feel that the following
- factors in the original inspection satisfactorily cnswer the chove questions
j - Liquid penetrant inspection of castings requires carefully controlled tech-
! niques to casure of f ective detection of defects. The inherent rough surface p of the casting presents special probicas in removal of excess penetrant. prior i
V to applicatioa of developer. The unst frequent cause of lo'ss of effectiveness
, for J iquid- penet ri.nt inspection of castings is over-rer: oval of excess penetrant.
! Where water-washable penet rant-dye in used, encess in frequently reucved by l high-velocity water-jet; ubere s ol v en t--d ye is uned, exce: , is frequently l reooved by spr.ylun or fioshing the surface with liquid solvent. liot h of
.j the above ne tho.h are very effective in renoving excess penetrant from the '
j rough casting surface; however, they are also very ef fective in rcuoving f the penetrant from surface defects. Such abuses are con.uon practice ubere I
liquid penetrant inspection is perforncd on n sub-contract basis eith no
, vitaescing er process control by the purchaser. Also, the finn) liquid-
] penet rant. inspection uas performed on the raw castino ; nachined areas of the
, casting were not reinspected. Since 3inear defects such as shrinh, hurled tears, etc., are likely to be uncovered when the skin of a casting is removed by nachining, liquid-penetrant inspection should be donc af ter final cachining.
'l - Padiography of critical cantinpn or uridu nts requiren independent rc v iew l j , of the fitus by at l e a:. t two qualified in t erp re t e r: . No raat t er hop conscientioua j and careful en interpreter is, i t. is inevitable that a feu serious defects l will be nissed or misjudged in the course of J ooking at several thousand j filn. It is very easy for an interpreter to lose his place in a st.ack of j -
film, due to an interruptica, and to niss one or more films entirely. Also, crrors of judgement can occur since the standards are, at best, a vague guida
! to acceptance. Uhere films are revieacd by more than one person, the probability i of missing a significant defect due to h aan crrors is reduced to essentially jp zero. Ic is not at all surprising that the original acceptance of the. subject 1 V valves' radiographs, colely by the inspection sub-contractor's interpreter, resulted in overlooking or misjudging a few serious defect indications.
4 s s.
1 3,
'*'*"P*
- T,"f "" ' QTR h*
- 4 18Fpy g %.p. - .-.1 ;* Q ' * **
)) -
"O e
C <
J r
s -
l . .
oor er..no s ecyteale ,.
l 4 RT 7n.ii.. c .3 sc' g. . c c l,alve Itcs ) IIeat Body I i
_a. -
3o. ,
Teundry
, I n US " a lch rin'< ; E C .--a
! .nu cations -l C - nt
,0. s. ..o . -
- c 'm Tm! - Nw !
b r'a .
19 5-12 4 Roe =er lV-20-15 1 1
- 0 0 .
c.:cepted jV-20-21 19 ,5-41 7 Roccer 0 0 0 o ,
.s:cepted
'l lV-2C-40 19 4-9 6 Roemer 9
- 9 0
0 See Note 1 1ccepted j V-20-41 '19 4-12 8 Rcener 4 3 1 .s..uncrous
, j Valve Removed Scrapped V-20-41 19 1-1 -
noener 0 0 0 -
.:,eplacement Valve iccapted V-20-12 14 5-11 5 Roemer 0 1 1 1
,See Note 2 3.ccepted
,V-20-17 21 4-7 1
, Roemer 2 1 0 I i 0 . See Note 3 '
-f , Rc; air !
g V-20-18 14 4-10 2 Roccer 0 1 , 0 2 l
l See cote 4 /.ccepted
'V-20-23 ! 21 4-8 3 Roemer 2 0 1 0 See Note 3 3
' Accc7ted '
V-14-30 l
'73 i4-14 5 Roemer
- ] i 7 ,See Note 6
' "cce?ted V-14-31 8 74 4-23 8 Roemer ij .
,See Note 6 A;;e,,ted ll ; V-14-32 : 73 4-15 4 Roener !
- t. , l 3 See Note 6 l l .
a _, --<c .' d ' ,
, ; ,' V-14-3 3 l 74 ,4-4 3 Roccer l
- j l 3 lSee Note 6 3,c c;.p ted I
p S
.4 f
.3 3
4
~
>1 -
N '
J a
G O O
- f. .. -
~ m
3orderline .iccaptabic i RT Trticaticas No. Surfaca I Valve ! Itea !? cat 2ody Inc;usions j ) I c.d t e a ;-ior. 3 l Current l No. ! No. o. !No. Foundry Shrink 5: C-~ : ales iTea rsi '<r7ored Dera ri:s Status I I
l V-14-34 71 l4-18 6 Roemer 0 0 2 1 Sea Note 7 Scrapped l i
V-14-35 71 4-22 7 Roe::.o r 1 0 1 0 See Note 8 . Accepted a
a
- 61 4-21 '
9 Rocter 0 2 2 0 See Note 9 '
Accepted l V-14 l l
i v-14-37 61 7-20 , 11 Roemer 2 0 1 0 See Note 10 Retake i V-15-29 l113 3232D 1 Vulcan 0 0- 0 0 . Accepted I I' I '
' 160 5-9 10 Roemer 0 1 0 1 -
See Note 11 Accepted O
s i
lV-16-1 i V-16-2 161 4-S 1 Rcemer 0 4 1 0 l
! See Note 12 ; Retcke j i
V-16-14 160 4-48 9 Roemer 0 0 0 1 See Note 13 l Accepted ,
l .-
l V-16-61 161 5-8 5 Roeter
- 0 2 0 0 ; Accepted V-16-62 135 321SD 3 Vulcar. 0 0 1 0 See Note 14 Accepted j i i i lV-16-63 134 5-32 l 7 Rocter 1 0 1 0 See Ncte 15 Accepted'
, V-17-19 l 4S i3206D 2 Vulcan 0 5 3 4 , Sce Note 16 Retake <
i .
l jV-17-54 l 43- , j3206D 1 ; vulcan 3 1 1l 5 ; see Note 17 ,
Accepted P
0 1
1 O "
O O -
o m -rr-- - - -e. ,.. ....r. ..-e--- 3---- ,3---------~a._ .- s -~ ~ - w w ey -re rr--e r rere a r* * = " - = ; - - .---- *m eem
. s- w ornye we t ~. p-+--c. -o, 3 y -, +_ r g q y , y p gyy
4
~
!, e
('
l Note 1 - Reviev of V-20-40 body carting RT indicated that one filn station
) in Bar.d B had borderline coverage in the thick section, nissing approximately 1/4" at the edge with the next band being singic film.
, liovav e r , the general quality adjacent to the missing area together with the section thickness are sufficient to assure adequate structural soundness.
I '
- Note 2 - Reported surface defect has been repaired. Review of a full set of j site retakes rhowed no change in original defects and no development
! of additional defects. The hot tear reported on original films was I a short tear into top of bonnet rin; it is presumed that all or nost
{ . of this tear was renoved during final machining of bonnet rim.
9 I
Note 3 - Review of V-20-37 body casting RT instcated that in addition to two l (2) borderline acceptable shrinks and one (3) borderline acceptable inclusion noted in results, a ]arge unacceptable shrink uns located in the bonnet seat ring area. This indication uas confirmed by RT
) of the disassecLled valve in its installed location at the site. The l unacceptable defect will be repaired to an acceptable condition or l valve body will be rep] aced. A questionable indication (possibly surface-pattern) requires retake in another part of same body-casting.
Note 4 - Reported surface defects have been repaired. The borderline-acceptable l condition consisted of a combination of gas-holes and small shrinks,
}d such that the corbined defects constitute a borderlihe-acceptabic condition. A full sct of site retnkes were reviewed and shoued the j same identical defects as the original filr..s; there was no change in I the defects and no development of additional defects.
I e i Note 5 - The hot tear reported on original films vas a short tear into the top
) '
of the bonnet rin; it is presuaed that all or nest of this tear was recoved ducing final unehinion of t he bonnet riu.
)
l Note 6 - The RT review of these body castings was conducted in October 3968 and reported to W. M. Scott on October 17, 1968. The casting quality
, was acceptable to Class 2 standards of ASTM E71. The surface indica-i tions have been bround and/or repair velded as required to obtain an acceptable condition.
t Note 7 - Excavatioti of a surface defect uncovered an area of shrink / tears that
, vould have required repair uciding more extensive that would be practicable in field; therefore, this valve body will be replaced with a new casting.
, Radiographs of the replaceraent casting are not yet available for review.
j The defects showed on the original fi]m of the affected area, but the l KT-indication was acceptable to C1hss 2 standards.
1 J
sv .
? .s 7
i
! . a - , . -
_________~._.._-.w.. m m.m a L
Y .. . .
l I' l' s Note S - The singic hot tear reported uns in an area rejected for hot tears by original radiography and repair welded by the foundry. A small residue of one of the original hot tears was not removed. This hot tear rennant is considered to be buried in the octal thickness under the repair veld deposit.
l Note 9 - One of the reported hot tears is a short tear int o the top of the
- bonnet rim; it is presumed that all or nost of this tear was reuevsd j during final machining of the bonnet rim. The other reported hot tear was in an area rejected for het tears by original radiogrrphy l and repair ucided by the foundry. A small residue of one of the i original hot tears was not removed. This hot tear remnant is considered l' to be buried in the nota] thickness under the repair veld deposit, i Note 30 - The single het tear rcported is a short tear into the top of the bonnet j . rim; it is presuacd that all or most of this tear was removed during k final machining of the bonnet rim. Two exposure locations of original j filn could not be found; retake at site is"rcquired to provide full l RT-coverage for this valve body. '
1 -
f Note 11 - The single surface indication explored on valve V-36-1 was faired by j grinding and found satisfactory by PT uithout repair.
! Note 12 - The valve casting heat number reported by the site for the installed l valve V-16-2 was "4-8 Body 1". The film reviewed by.Tackett and Franks j was marked as lleat 4-431:ody 3. Verification that body narked 4-8 Bady 1
! corresponds to film marked 4-481;edy 1 has been made through Anchor l Valve Company and ui]1 he confirmed in writing by Anchor. In addition j to the four (4) borderline acceptable inclusions noted in the results, j a large cluster of unacceptable inclusions were located in the body wall bcLueen the seats and up along the guide. This area may have been repaired, but repair film coverage of idjacent. arens is not adequate to verify removal of the unacceptable coadition. The questionah]e area will be re-radiographed, evalunted and repaired to an acceptable condition, if required. The reported area of hot tears consists of two short, paralle] hot tears in the thin section betueen the seats and adjacent to the thich sect ion into which seat ring is installed.
These tears may be open to the inside surface; therefore, site retake and inside-surface exaufnation is required to cvaluate the need for repair. If these tears are found to be buried in the wall, they uould be borderline-acceptable to Class 2.
Note 13 - The single surface indication found at site has been repaired to a satisfactory condition.
Note 3 4 - The single hot tear reported uas in an area rejected for hot t ea r.4 by original radiegraphy and repa:r velded by the foundry. A small residue of one of the original hot tears was not removed. This hot tear remnant is considered to be buried in the metal thickacss under the repair weld deposit.
\
4
1 4
- e l _)3_
i
}
( (v n) Note 15 - The sing 1c hot tear reported was in an area rejected for hot tears
] by original radiogtaphy and repair ucided by the forndry. A small 8
residue of one of the original hot tears was not renoved. This hot tear remiant is considered to be buried in the retal thickness under the repair weld deposil.
Note 16 - One of the three reported hot tears was detected as a surface indica-tion during site inspection, excavated, repair welded, and the area
, re-inspected by radienraphy. Another of the three reported hot tears 1 was in an area rejected for hot tears by original radiography and repair veldcd by the foundry. A small residue of one of the original
- ! hot tears uas not renoved. This' hot tear remnant is considered to 1
- be buricd in the neLal thickness under the repair veld deposit,. The j third hot tear reported is adjacent to an area repair-ucided by the i foundry and apparently opened as a result of that repatr; site retake j is required to confirm that this defect uas removed or to confirn
] (with inside/outside surface inspections) that the deftet is buried j' in the surface (in the latter case, the condition would be considered borderline-acceptable to Class 2), ,
- Note 17 - A ful) ,e t of n!te retaku; (n
- ade af ter site repairs were complet ed) g were reviewed and con: pared to the original f 11rm . The three areas
- of borderline-acceptable shrink PT-indications were found to hava been due to a surface partern which uns removed during eriginal. finishing f,J of the body casting by Anchor. The single horderlin'e-acceptable hot-g tear indication reported in the original film was a short tear into the t.op of the bonnet rim; it is presuned that all or r
- ost of this hot tear uns reraoved during iinal nachining of the bonnet rin. In j addition to the borderline-acceptable conditions reported, the original j fi] ras shond a six-inch long hot-tear indication, narked " surface".
. This RT-indicat ion van clearly rejcetable. This hot tear was detected
- during surface inspection at the <ite, was excavated, repair-weided,
- and the area re-inspected by radiography. Thorout,h review of site-retakes showed no change in original defecta (except for site repairs)
{
nnd no development of additional defects.
J Cene ral No_t_e :
b
- The borderline-acceptable hot-tears which were accepted by the 1; viewers
) (Tacket t/ Franks) as noted abeve were of two types:
? - Short hot tears into the t.op rin of the as-cast bonnet-area of the body-casting.
Much of this area 20 removed in finn 1 nachining. Sna11 tears that reraain in this area are not considered injurious to the structural soundness and service integrity of the valve due to their location outside the bonnet-scal area.
t .
1 /7 i V 1 .
9 a g h
. , - , +
w.. -., g p eg ,,,.n. , ,.,.e,. ,,y, , , , . g , .., ,
i .
j -l k
^
p a '
r -
Small rennants of hot tears that were not comp 1etily removed during exca-vation and repair velding by the foundry. 'Ihc tears are buried in the wall of the casting, under the repair wcld deposit, and therefore do not '
have an injurious ef fect on structural soundncas and service integrity any
-! t' ore than acceptabic small shrinks or acceptable IJncar inclusions.
Ilot-tears which do not f all into one of the above categories will bc (or have been) repaired, regard 1 css of acceptability to Class 2 of ASTM E-71.
4 I
i f
I-l J '
i
!O '
4 4
4 i
I
,I
.)
9 e
.Ir
\.
.r e ~,. . . .7 m., p . %,, , , . . . . . , . .. . , , . _ . , , _ ,. , , .. ,
- e
[ i t
i Exhibit B i
!O I TO: R. Carlson - Region I, Newark Office
( Division of Compliance - AEC f
L F.o N!:
( R. G. Gilliland - Assistant Professor of Materials Engineering l
College of Applied Science and Engineering University of Wisconsin-Milwaukee
! THROUGH: Parameter, Inc.
i, Consulting Engineers ,
Elm Grove, Wisconsin
Subject:
Metallurgical Examination of Casting Defects in the St'inless Steel Valve Bodies of the Primary Circulating System of the Oyster Creek Reactor - Preliminary Statement Dated: February 13, 1969 O
Introduction At the request of Region I, Newark Office - Division of
- Compliance, the author visited the Oyster Creek Reactor cite -
near Toms River, New Jersey on Februarv '12,1969. The follow-e ing is a preliminary statement briefly outlining the initial i conclusions of the metallurgical analysis of Casting Defecty
{
found in stainless steel valve bodies in the primary circulat-ing system. All the valves in question are CF-8M (similar to type 316) stainless steel cast material and range in size to 16-in. J[
in diameter.
,~ -O Discussion ,
of the forty-three (43) valves involved , sixteen ( 16) valves s
y E
'N l.
Wysivspsar , nr .w..e -
-,,,,m.m_ . . _1
?-
o (y
required repairs of defects which were discovered after installa-tion in the reactor system. Two (2) of these sixteen valves were found defective to such an extent that replacement of the valve was necessary. Three (3) other valves from this group of sixteen required weld repairs. The remaining eleven valves could be sat-isfactorily repaired by only surface grinding. Visual and dye penetrant inspection techniques were used to establish the in-tegrity of all valves being reviewed. Radiography was used to establish the as-repaired integrity.
The extent of'the casting defects found and the fact that all defective talves were purchased from the same vendor ( Anchor Equipment Company) prompted a review of the original, specifica-tior. radiographs submitted by this vendor (Anchor supplied 40 of the 43 valves in question). The; contractor, General Electric Company, elected to review only 21 Anchor valves; their selection was based on service requirements and part size. In all cases, at l ust the maior defects (those which resulted in valve replace-ment or weld repair) were clearly visible in the original radio-l graphs, t
Conclusions.
Based on the above facts and observations made during the visit on roter tary 12, 1969, the following preliminary conclusions can be stated.
l
- 1. The casting defects found after installation, which O) v .
\
l <
i- , ,
- ..- . ~ - . . .
s' O' resulted in necessary repairs (grinding, welding, or replacement), should have been observed by the vendor during their nondestructive testing program . Thus, the quality control of the valve supplier was greatly lacking. In addition, it is felt that the contractor, General Electric, must be held responsible for their negligence in not maintaining by continuous review the quality assurance of their vendor, the Anchor Equipment
' Company. Had the code-required procedures and specifi-cations been properly adhered to the current prcblem i
j could not have existed.
- 2. The procedures and techniques of weld repair of the
-I casting defects, and subsequent nondestructive testing, are considered to be proper and correct. The good ;
practice of low interpass temperature (200of max.) and stringer-head technique was being employed.
- 3. The cause of the casting defects is suspected to be a result of solidification shrinkage, hot tearing, and/or mold-metal reactions. These very preliminary and gen-eral conclusions can only be specified as tentative until an analysis of the proposed metallurgical investigation can be made.
- 4. It is considered important, if not critical, that the l
remainder of the Anchor valves (totaling 19) be reviewed, l
- () .
/
o
.-i. .
I I -.
'l .
O similar to those already examined. This statement is supported by the results of the re-examination of the i- other 21 Anchor valves, considering that this review j; caused major repairs to be required on 25%~of the parts in question which had previously been termed as meeting
- Code requirements. -
o This " preliminary statement" report is made subject to revision by a more detailed document to be prepared by the author in the verynear future.
e , i ..
/ t
'h ~(
) L l (. s. < -
- O -.S
}
t o
i <
r l-i o
I t
O .
- r.~- . ,. m .
N- -
G_.-____.
Exhibit C O
)
s a v 3 I < ,
. e .
Ce7 ~ .
N . , J ,.;.at
,, v" n~
.o:
,f,n ,. , .'s, %p; a
A \f -. s %. ,r t, t .. .
L'~, t '- y, s
s . ,,
J.s3 r, .
' s
, Q ( ' ,-
>?
s
- +
4.
j, .%~ y pg
. <g*, / . -;
(, I ', '*(..,
.V ,
_)
. ~ - -
O
, a . .
.- a 4
1 i
i Micrograph of a typical casting defect found in valve number V-20-41 during the metallographic examination performed at APED-GE-1 San Joso, California. Lightly etchad - 250X.
l l
l
}
d
. _ ~ - - , _ , _ ,_ , _ . . , _ _ - . , , - _ - _ _ _ _ _ , - . _ - - . - _ , - - . - .- - - -.- - . , - - . ,
-Exhibit D The University of Wisconsin - Milwaukee MILWAUKEE, WISCONSIN h3211 COLLEGE or ApptiEo SCIENCE AND ENGINEERING February 25, 1969 Mr. Robert Carlson, Sr. Reactor Inspector USAEC - Division of Compliance Region I - Newark Office 970 Broad Street Newark, New Jersey
- Subj ect
- Preliminary Conclusions Pertaining to the Metallurgical Analysis Performed by APED-GE on a Cast Stainless Steel Valve Body from the Oyster Creek Safety System.
Dear Bob:
As requested I visited the Atomic Power Equipment Division (APED),
General Electric Company in San Jose, California on February 20, 1969.
() The f ollowing is my preliminary conclusions of the metallurnical anal-ysis performed by APED-GE on a cast stainless steel valve (V-20-41) from the Oyster Creek Safety System. This valve was the first of a total of three valves to be rejected in the current reevaluation of the safety circuit valve system.
! 1. Three sections, taken from the top, vertical portion of valve No. V-20-41, indicated considerable micro- and macro-shrinkage voids. Some of these shrinkage voids were ob-served metallographically to link-up, forming what in com-monly called " shrinkage cracks". These shrinkace cracks were observed, in some cases, to be greater than one-inch in length.
- 2. These shrinkage cracks were observed to possess blunt tips, further verifying the fact that these were shrinkage cracks j rather than hot tears. In addition, this observation sup-ported the earlier conclusion that the cracks did not pro-pronrat e during boat t reat mont .uul wa t er quench i ng.
4 O
t
~' ',, ' , l' i, _ J . '_~~~f[~~_.___ , .C ,_ _ _ . . - . _ _ , _
2.
(~b
(_) 3. The defects metallographically observed are believed to be casting
, defects, common to the casting process. (The fact that these in-dications were observed in the as-cast radiographs of this valve, and other similar parts, supports the statement that these were casting defects.) Such defects are often expected in designs such as these, and the fabricator usually tries to reduce their occurrance by gating and risering techniques and, when required, weld repair.
- 4. The breakdown in the control of the review and analysis of non-destructive test results is the principle reason for defective parts such as these being undected. This fault must be borne, in a major way, by the primary valve contractor, Anchor Equip-ment Company; however, the absence of backup review by the General Electric Company must be considered, and at least a portion of the responsibility be placed with reactor centractor.
- 5. Finally, it is my judgement that the G.E. people have a good un6er-standing of the problem, have performed an adequate analysis, and are quite aware of its cause.
Best personal regards.
Very truly yours,
',.,m v.
a.A
- d. G. Gilliland Assistant Profer.sor Materials Department RGG:jlm cc: Mr. Richard A. Lofy - Parameter, Inc.
Mr. G. W. Reinmuth - AEC - Washington, D. C.
O
_ , _-.