ML20083L803

Trip Rept of 820813 Site Visit Re Replacement of Recirculation Sys Piping

ML20083L803
Person / Time
Site:	05000000, Nine Mile Point
Issue date:	09/02/1982
From:	Cheng C Office of Nuclear Reactor Regulation
To:	Hazelton W Office of Nuclear Reactor Regulation
Shared Package
ML20083L677	List: ML20083L685 ML20083L692 ML20083L803 ML20083Q826 ML20084N257 ML20086P013 ML20086P034 ML20086P038 ML20263M566
References
FOIA-84-14 NUDOCS 8209230151
Download: ML20083L803 (16)

Text

{{#Wiki_filter:(( {- { e SEP 2 1981 4 MEMORANDUM FOR: Warren S. Hazelton, Acting Chief g Materials Engineering Branch Division of Engineering FROM: C. Y. Cheng, Section Leader Materials Engineering Branch Division of Engineering

SUBJECT:

TRIP REPORT - MEETING WITH NIAGARA MOHAWK POWER CORPORATION ON THE REPLACEMENT OF NINE MILE POINT UNIT 1 RF:IRCULATION SYSTEM PIPING On August 13, 1982, the NRC staff and its consultant from Argonne National Laboratory met with Niagara Mohawk at the plant site to discuss licensee's proposed plan for replacement of all recircula-tion system piping (nominal pipe size - 28 inches) at Nine Mile Point, Unit 1. The attendance list is attached (Attachment 1). The meeting was focused'on the issues presented in the licensee's August 6, 1982 submittal which provides the basis for Niagara Mohawk's decision to replace all recirculation piping. This decision was made because results of the ultrasonic (UT) examinations indicated cracking in all of the welds inspected. The cracking was confirmed by dye penatrant method and was later determined by metallurgical examinations to be intergranular stress corrosion cracking (IGSCC). Major items of interest discussed during this meeting'are summarized below. 1. Inservice. Inspection (ISI): Although the licensee's ISI con-tractor.was not present at the meeting, it was learned that the normal UT method has been employed in the ISI of the recircula-tion piping welds since 1979. This normal UT method is an improved UT procedure which has been calibrated against the EPRI pipe samples and has been demonstrated to be able to detect IGSCC in the pipe samples at the code required gain or sensitivity level. Using this normal UT method, it was determined that two of the five pump discharge casting to riser elbow welds had code reportable indications. However, when the UT transducer gain

Contact:

.C. Cheng X-28144 .s Been Sent to PDR 20923015 20902 D A-A AW. 000220 CF i _J

(C c-( [ Warren S. Hazelton SEP 2 1982 or sensitivity was increased 10 dB (a factor of about three), the remaining three welds also exhibited indications. In ~ fact, all recirculation piping welds exhibited indications when scanned with increased gain. Two of the five pump discharge casting to riser elbow welds were later examined by dye penetrant methods on the inside diameter. The results of these examinations co,nfirmed the presence of cracks. The generic implications for the apparent deficiency in the normal UT method cannot be established until the actual flaw depths in these welds are measured and compared with the acceptance standards specified in Section XI Code. To aid our evaluation of the generic implication of this cracking incident, a complete ISI history and results of each inspection for each recirculation piping weld were requested. Only partial information pertinent to these requests was obtained so far. In addition, the stress rule index which represents the total stress level for each weld was also requested. The licensee has committed to provide this information in the near future. 2. Metallurgical Examinations: The results of the preliminary metallographic examination of a boat sample taken from one of the five pump discharge to riser elbow welds were presented ~ -by the licensee. The examination war performed for the licensee by the Sylvester Associates (Attachment 2). The preliminary results indicate that: (1) the UT indication is a 360 intermittent circumferential crack, (2) the surface indication is clearly an IGSCC, (3) the boat sample appears to be sensitized, and (4) the crack in the boat sample has a through-wall dimension of approxi-mately 0.02 inches (Attachment 3). The wall thickness around the riser elbow is approximately 1.25 inches and the minimum wall thickness for the recirculation pipe is 1.05 inches. It should be cautioned that t..a boat sample has-the through-wall dimension of only 0.042 inches and therefore cannot truly establish the actual crack depth for the indications observed. Further study of the cracked pipe is needed to clearly determine the crack depth and the cause of'IGSCC. 3. Code Compliance: Although the entire replacement program appeared to follow the apolicable code edition and addenda, two questions regarding the fit up requirements and stress reanalysis were raised by the staff. The licensee has indicated in his August 6, 1982 submittal that the' fit up requirements will be in accordance with ANSI B31.1 - 1977 (with addenda through winter 79) Code for pressure piping. a I e a

( ( ( darren S. Hazelton SEP 2 WS2 The staff recommended that, recogn 4NSI B11.1 was the original construction code and if tha replacement piping was ordered and is being m ..o the 1980 Section III Code through winter 1980 addenda, a simpler at least from QA viewpoint to follow Section III 1 mirements instead of those requirements of B31.1 code. Tht. l' ices. ;e contended that aside from the NDE requirements which the licensee has already committed to follow the requiremerts of Section III and XI codes, the fit up requirements of ANSI B31.1 are practically the same as that of Section III. A comparison of both codes should be pursued to ensure that there is no incompatibility. The issue of whether there is a need to redo the stress analysis was also questioned by the staff regardless of the licensee's argument that the configuration of the pipi,ng syst, will be the same as the original design. Since the original cc

ruction, many modifications had been made on these piping systems, especially in the area of supports, hangers, restraints, etc.

The piping stress is strongly affected by the exact locations of these piping hangers or supports. Even with a detailed benchmark, piping replacement program of such a scale would probably result in some misfit or misalignment of pipe ends which would certainly require some reanalysis of piping stresses. The licensee indicated that the decision to redo the stress analysis or not has not been finalized yet and they would certainly take all these concerns into consideration. 4. Replacement Materials and Cutting / Welding Procedures: The licensee indicated that all replacement piping materials will be 316 Nuclear Grade or equivalent with a carbon content of less than 0.02 percent. These materials do meet the piping material selection requirements of NUREG-0313, Revision 1 and therefore no augmented inservice inspection is required. However, the staff emphasized that the equivalent materials should also meet all the mechanical property requirements of 304 stainless steel so that no stress reanalysis will be needed. With regard to the welding procedures, the staff was reassured that procedures similar to that used for safe end replacement will be used for recirculation piping replacement with some minor modifications. The safe end weld procedures were previously approved by the staff.. Documents for these modifications were obtained for further staff review. 5: Pipe Samples: To further evaluate the generic implication and to uncierstand the true cause of this pipe cracking incident, pipe samples containing cracks or indications from the recirculation piping were requested by the staff so that those pipe samples can be shipped to ANL for independent study. The licensee has committeed to provide the samples needed for this investigation. munuu m-u-um mm uumnu

( SEP 2 1982 Warren S. Hazelt'n o 6. ALARA Consid'eration: The issue of dose estimates and control was briefly discussed. The licensee has committed to the ALARA objectives and will exercise whatever measures available to achieve that goal. At the end of the meeting, the licensee requested staff's approval for removing the recirculation piping so that they can proceed to cut'the piping on Monday, Adgust 16, 1982. A telecon was therefore made with the Project Manager to. discuss this request. After the t91 econ, several issues were conveyed to the licensee and requested him to reply ir, writing before the staff can approve . removing of the recirculation piping. The staff requested that: 1. A tentative schedule for the replacement program.should be provided. 2. If there is a change in the essential parameters in the cutting / welding procedures, the licensee should notify the resident inspector and obtain approval before its use, 3. An ALARA program similar to that for the safe end replacement program which was previously approved by the staff should be implemented to mitigate the radiation doses to the workers, 4. Torch cutting should not be used for the end-cut at the valves or pumps and the end-cut for the loop piping, and 5. A detailed man-rem exposure estimate for the replacement of recirculation piping should be provided for staff review. -In summary, the results of the preliminary metallurgical examinations of the. boat sample removed from #15 recirculation pump discharge to l riser elbow weld did not clearly establish the crack depth, the i cause of crack initiation and whether the piping was completely I sensitized. Since the cause of crack initiation and crack depth cannot be determined at the present time, its generic implication is not clear. To further evaluate the cause of IGSCC so that the staff can determine the generic implication, the following action plans with the lead responsibility idenified are recommended. Recommended Plans for Action: A. Plant Specific: 1. SER on approval for removing the recirculation piping (NRR/DL & DE). S e

Warren'S. Hazelton SEP 2 1982 . 2. Meeting with the licensee to discuss whether there is a need for stress reanalysis's (NRR). 3. SER for approval of installing the vent piping (NRR). Bc Plant Specific and Generic: 1. Request for additional information on the fabrication history, inspection history and results including UT procedures and stress rule index for each recirculation weld to determine the generic implication (NRR/RES/IE). 2. Obtaining the pipe samples for further study to determine the generic implication by ANL (NRR and RES). C. Generic: 1. ISI and UT inadequacies: Issue a Regulatory Guide or propose a Code Case for UT of stainless steel piping to detect IGSCC (RES). 2. IE Bulletin on Information Letter to BWR owners (IE). 3. Longer term studies at ANL (NRR/RES). /A/ C. Y. Cheng, Section Leader Materials Engineering Branch ' Division of Engineering ' Attachments: As stated cc: H. Denton T. Ippolito . DISTRIBUTION: E. Case S. Pawlicki DMB - Central Files R. Vollmer-C. Serpan - MTEB' Reading Files D. Eisenhut W. J. Collins W. Johnston P. Polk E. Sullivan R. Bosnak G. Lainas J. Halapatz E. Jordan H. )n d DE: EB C. Y. h W. o on, Act. 8/ /8 '/82 4 \\

r. b (\\ --(. \\,.. s ATTACHMENT 1 Meeting with NMPC on Recirculation Pipe Replacement (8/13/82) ATTENDINCE LIST NAME ORGANIZATION S. W. Wilczek, Jr. NMPC-ENGR. R. J. Pasternak W. J. Collins NRC - IE C. Y. Cheng NRC - NRR W. E. Ruther ANL R. H. Todd NMPC - QA H. Gray NRC - REGION I G. Gresock NMPC-ENGR. T. J. Perkins NMPC A. Hulse NRC - REGION I S. Hudson NRC.- Resident Insp. 'T. Gillman Newport News Ind. J. B. Keating NYS PSC J..G. Roberts 'H. Gruttman NMPC-Rad. Prot. E. W. Leach O t ) '\\ a m'-

i (( (b ?< l ATTACHMENT 2 l Metallographic Examination of sample removed from #15 Recirculation pump. Your P.O. #7771. J.G. SYLV' ESTER ASSOCIATES INC. JUNE 1982 a 4 9 N-t t x e

(f ( k. J. G. Sylvester Associate.s. Incorporated Metallurgical and Welding Consultants Materials Testing = __m _.x ,_2 900 Hingham Street. Post Office Box H. Rockland. Massschusetts 02370 0607. Telephone-617 878 9000 June 23, 1982 Mr. G. J. Gresock ~ Niagara Mohawk Nine Mile Point Nuclear Station - 3 t P. O. Box 32 i Lycoming, NY 13093 l

Subject:

Metallographic examination of sample removed from ' Nine Mile Point #1 Number 15 recirculation pump. Your P. O. #7771.

Dear Mr. Greseck:

a A one-inch long "bo.at sample" was ground from the number 15 ) recirculation pump and submitted to.our hot. laboratory for metallographic examination. The sample is shown in Figures 1 and 2. A crack-like indication is seen along the mid rib of the sample in. Figure 2. This indication was studied by our metallurgist, Dr. Thomas Eagar, under the supervision of our health physics personnel 's in NRC licensed f acilities. The sample was mounted in clear plastic and cut in two equal sections. Although. the surface crack-like indication ran through this section, no crack could be found on the transverse section at 200X magnification. There was an indication at the edge of the

)

. sample of less than 0.001 inch depth, but it could not be il classified as a crack. Following this, the end of one piece was u polished and etched per ASTM A262, Practice A to reveal the crack d shown in Figures 3 and 4 at 100X and 200X, respectively. The ] surface indication is clearly an intergranular stress corrosion I crack, (IGSCC). ,j r It shou.ld be noted that the surface indication is approximately { three-quarters of an inch in length, while the depth ranges from 5 less than 0.001 inch to 0.010 inch or greater as shown in Figure 3. ? The crack'is typical of IGSCC and the etched grain boundaries j indicated that the steel in this section is sensitized. (W These results have been conveyed to Prof. W. F. Savage for his E review and comment as requested in your letter of June 16th. If we may be of further assistance, please contact us. [ S Sincerely, 4 0m $\\ Dennis.C. Yea to n General Manager { D C'Y / l _i o h Enc. Quality Testing For Industry i}

. c. ('- mw db A ( W a, .b W.,;.. s..e..."x.?..u-ett '%-.; ,H. ::','8'-T.,*.%*'J'.~*~~.'.~~-P.T*"i. - i.m%. ~tE.,C- ~ 'tT:

• ~f e-p e

+ ,.. w. < ;, m .-e-.. e,,... n n- .,s m;4 3.sf;2(;. .d..;s:~.,: n..,w ;m...w..

5. y..-n

. ~ - .... e.. w :. c._., a.. .v L?is.~.".' 'W;:' E".;E *c/A :- s 4*: 5;.: = *:~.Wi;U '. - +'. ~. ' ~ - - '.".?. 'r. -;. :..,.....->:::. ' 4 : n s 'i. es :v -y"i .a. x .1.~ ~ ....w u. -m...ti,

..y

.-. : e.<,>.:,a - ,._f. p; a wit.;r.c.- =

'b-;;&...'. 2 5,,:--

.o. ?.- r.,... ,.,,, m ; ,r,:... ..s.

• • w

1. ?,. y :.:'-.

r s. w g.. Ue .< *w c... -.e - ~ g

3..].. ;.. -

. ::;.l. f. 2.,y., -

~+ M -- _

,,, 3,:: - pw ,..r. h.u.,. y: v v, w~, -

7.,.. : cM.;.,.W.Gf,y/.@;;.-

' d.y.... u g ..: w.. \\ _ : c:.:v:i2MR&,::- @s,a &w,;r. 4;,, @; ~.-- ::.s. :.:. y : Y .-.,..v t ll-1 ~~i l l-.7l

g/. 'i ' - ;.?j..

@s. .... m., . g. 9 x.,. g.. -s y;g;,.~j*.,.;.J.. 83[d-T T79.i:.s : .g ... c. y' y-

..\\s

-n. - y L *.M Wrw s. .er=m w ,v.. a 8

• kb T.-

.f? _ _. ' sk J-h ,a . ' ~. '~- .*&A.W. ' _^&;:Q'~ '"?shkQ. -~.'?, .. = %'.$.f' Qv.'. '*. P Figure 1 (2 Y1 pm 4.::w Ma

f. :-os w.: -!

-.c 7 Kdp.- w% p s waw_ n @d@f*;;-- sut %b M. ~ u aU N. N~ OF8 -r-. % w:.#.~. _ w. sq. u w%.= w... m m. m &m..y.e....:. m .gy 4 l7,,3 wa% mend %- m&:.9=' b.w-a-,.w&&%m..n.R.-2M:.i

~-

we w n

r.o.

d .z. :e. v e m .. 4.,.k'z:w.:RW474%4. %pr n wv,=w, a*fJyWgp"6*d4t%mu..--s.'m.Xi.' ~. . C%. 7. y'o. 2e- ^- er.ennue98 sr. r. 2 # JJ':*i,*1*# 7 ' ? M'+a N A., w.. - e i:, b.cm._. ,w s v r. s 4".-*':.f f.ssN vM t*-*** 2*C 4.g.pdR%-~gy. g%'":: 4.- a $g W,. .w F ug.e.,... _ . r.z. Q. ,n W s M. s.g - W' =q L f* *4 I -*i i e 4"4

~. w y-e

.'.Y k g..<--

Figure 2 ((y) igd, 4...- -W.

. a w-m3 m _ m rL .:p ggyjg.y;3 y \\, e-A kb hk. ~ P L 4 s.er vav&idh a d ex.4 ~'.&'h. r..MW.' < s^ WR:xc.: ' Wye ', E:.h.&. k? Q 'T Q &M5d~ -~ FW < &~:% 74

=.& 144

's'.b

C 4, s

h..t$. ' p; . :v.y th5,p;s c.,.nm2=w a d: gh ha %ewftw(t >..,.;l.:;&./r4

s w ::-s v.eQw 9'c. e g'n..4\\.'.;**

3.- ( .qs.. b ptNN;W'.q* N M' qa'IA

=.- y.- pw..a y

I N

Y.s e. k h... ./*,% h.: '*.., - \\~.m]?'.,' - Y...g i. gs 'g

1..-

N,i ha'. 'N*f d1 Y.'$. ~ *

• u.% '%T,.1,.:..,2w
• 5 v',
• d

y b ::

w. n*'.:f.. e ;. x.n.=..,; .. ~ : t..

s. >. '

/ 4 Y'x.,,.m,..f... a !,,..~ ;n ?. t ....2 .r =k 8;4 ,.t*.7 t.,.. -{'.....':(-,.:,e,...' .s. -. y.,. :.. w ;t,:,a .u .. i. :, *. k. 4 W;;.x:-;' 2 _ -....., ;. C.. : .. ~ .a h. ( ?. ..'--@G%'.% ' 'Shh w?* fg:.*.; u,.R '...

- % )g..

.t.%. :. - g; 2 % K. a 4E:lj. '...n\\1ar .,t. 7.~. c. pip +: _,i.1 Q $CM M'**--._.'M V M-^-'6?.F 54 W.[*h, 3ll @fg g h ~. ? 3 o l l Figure 3 ( / cox) N. kkb~~ .n. ".. t..% s 4 ~ & q4 r 7 .g, ?. Q- ~-

• p.t.,m.. e s.~~ :. gw..c.

.__g w. ~ 9v m~ =. 9.. ;.2- =. _. waw% i. ua..., u.- r L3 .t--

--. ~-

. v*4,.Q.-

". 6.- .s. M '-L.=;?'d%51"R"7'Ef#5.q1 Q. Q"Q G,...,..-. % q.. y -2 6.G m.u. m, '-%. w.h-4-a ?.0:l .c.. u W c,w n,- d.M&it-hy*m:%.t.9'a'..- w e T. pu' d' c'2 -tBl'-5 q; y'. ....,:m.-~. ~; @:=>., =mm. w ~ =- W"===..'~. G, kfu s.= A.~..% mwt-iN4.,-E h.te w=%n.% gTQ73 f f*epd.,--. [ lim .~ w .g 4w WK,. %,& J-u } ic*w u 3.. L;; W yt g#c. y..m;... a .a _v c--. Mr.c . m :$.y$, '. na?, " 4.lk?$$ yn A ..f. n~ jk" u V f k 4... ;%. %e.". /',t.:. %;

1. .y b=~~~.&m--~.@. h

• ?. !,.h,.,.

b. &

W r.- 4. t .5./. .. :n. t ..= t M.g:< N& $f {hh;c.?:~h' $fN$U-Sb.N h = s a. ^~ w - v w-2 e h.Y h. ~ Y dd Y.Y '- a's is. &:e - "L i i ~ .e m l ~,. s m

g-~

y- - - - - g.. s 7a -EN .1= _ _ _ g_,. :,_ l l Gia 9ti Figurc 4 ~ (aco X) M l g .~

@TERNAL CORRESPONDENCE {I.-013 ( (~ M YNIAGARA r' EM U MOEAWK w m n one. IROM R. H. 'Ibdd DISTRICT System f EL Messrs.:. R. O. Norrix DATE July 30, 1982 FILE CODE QA82925 G. J. Gresock "D" SUBJECT Visit to J. G. Sylvest. 3-N2.1-S12 Associates Laboratory General On July 8, 1982, I travelled to the J. G. Sylvester laboratory in Rockland, Massachusetts, to personnally exmine work done on a sanple. removed _fra Nine Mile Point Unit il Recirculation Line #15. We sample had been removed frm the pipe ide of'the weld joining the pipe to a casting on the discharge side of the pmp. fL. arrival at Sylvester's laboratory I was shown the report they had already done m for Dr. W.1 F. Savage and G. J. Cresock. %at rapw.L, dated June 23, 1982, and addessed to Mr. G. J. Gresock frm Dennis C. Yeatcn, will be referred to here as the Sylvester royw.L. We Sylvester report concluded that the surface inlication on the sanple was not'a crack at the mid-point of the sample. However, when a seu md plane was polished at 1/8 from-an original tip of the sample, a crack was found which extended ccmpletely through the sample with no indication that the crack would soon stop.. @ e cracking mechanism was described as IGSCC. W e extent of-the crack at that . point 'was reported to be about 0.010". Measuring frm the #ducierograph and apply-ing the reported magnification factor of 100x, I found the crack to be 0.020" at the point where the crack runs out of the sanple. W ese measurements should be s used as rough estimates only. . -on Magnification: - % e Sylvester report presents the phot m iu up aphs in that re- _ port as being at 100x and 200x magnification. he nminal magnifications given by the Sylvester laboratory for tha ydu.udcrographs they supplied for this report wer'e 50x and 125x..By_ wrative measurement and especially by measurment of the width of the sample shown in Figure 1 of Sylvester's report, wherela portion of a ruler is shown in the same view, I have arrived at the magnifications of 100x and 250x for the l pet.w i u vp ops used in this x=pw.L. I believe these magnifications to be consistent with those used in the Sylvester. report. It should be obvious.that measurements taken from photmiuvposs may introduce rather wide errors unless the lens assemblies have been calibrated. ' Laboratory Findings When I had the veyortunity to talk with the Sylvester laboratory personnel, I l 'found the sequence of activities had been'as follows: h e boat. sample was encased in its entirety in a transparent lucite mount. %e sample was then cut at its mid-l point and one_of the cut faces was polished. mis would be plane #1 ~in figure' #2 of this report. Plane #1, as ragw.ud in Sylvrcer's report, showed only a small dinple but no crack at the location of,the original surface indication. Attention was-then given to the tip opposite plane 91. %is tip was ground back about 1/8" I to plane 42. 'It is plane 42 that was shown in' etched form in Sylvester's report. I felt the amount or depth of crack shoun by plane #2 was not sufficient for i sQ, full characterization and I asked the lab personnel to polish the planes shown in. L figure'#2, this report. Plane,83 when polished, did not show any crack or indentation. f 9 e V - 'I

i (, A. \\_ ( . d482625 July 30,-1982 { ~Page 2 Plane 64 - showed a small indent but with netal pretty nuch enclosing the indent. She character of this indent was smooth-sided, not like a corrosion crack or. ' fatigue' crack. Planes 1, 4, 5, 6, and 7 were separated by approximately 0.020", i 0.010', 0.010", and 0.020", respectively. Planes 5 and 6, when polished, showed no crack, but they did show a deepening indent with' a metal roof alnest closing off the indent hole frcan the original surface. Plane #7 is shown in fugures #3 ~ through 7, this report. She character of tle hole we found fits the following theory of occurrence. This theory was described to Dr. Thcznas Eager, of Sylvester Associates,-and he did not find disagreenent with it. 1heory of Flaw Merence It is postulated that the original defect began as a welding undercut in stainless steel piping which had not been properly solution heat treated. She undercut was paztially renoved by grinding with a coarse abrasive. The coarseness of the abrasive is attested to by the grooves seen on the inside surface of the pipe and joint. Evid6ntly, the crown of the weld was high and was ground down to natch the adjoining base metal. In doing so, the operator of the grinder actually pushed see weld metal over the undercut groove. The coarseness of the abrasive made it = par as though the undercut had been renoved. In fact, the undercut groove had beccane a closed chamber in which crevice corrosion could do its worst; and it was surrounded by severely cold e rked metal. The sensitized structure of the pipe shoas up in Sylvester's report in figures 3 and 4 asi in this report in figures 6 and 7. 7he sensitized condition can be described as' partial ditching after exposure to ASIM A262 Practice A. The photo' sequence in this report on plane 67 also shows that there is a plane in the heat affected zone and parallel to the fusion zone which is fully sensitized and displays a full ditch structure. The weld metal is ditched also, indicating the deposited weld metal was relatively high in carbon. In this envirarment, corrosion residues collected in the partially closed crevice which had. originally been a weld undercut.. Wh=~Iuent corrosive action

attacked the sides and corners of this chamber and established a true crevice corrosion crack accelerated by cold work stresses.. Sporadically, at intervals around~the circumference, the corrosion penetrated to the fully sensitized plane which parallels the fusion plane of the weld metal. -On entry into this sensitized plane, the rate of corrosive action increased to a considerable degree.

Discussion Crevice corrosion and stress corrosion can obtain a start in otherwise fully corrosion resistant metals by means of local areas which retain very high residual stresses. When both a deep groove capable of retaining stagnant material and highly stressed metal occur together, stress corrosion cracking has an excellent chance of occurring even in the nest resistant alloys. Once corrosion has gained a start, anode-cathode reactions extend and accelerate the local corrosion.in patterns known as pit and crevice corrosion. These highly stressed local startire points can be produced by abusive grinding. NMPC has.had at least one example where a newly installed field weld leaked upon hydro test..The nechanic assigned to renove the ~ defect by grinding actually closed the crack with-the grinder so ughtly that it S'-

.QA82925 {~'r {-(k July 30, 1982 , Pbge 3 I could not be sh wn by dye penetrant examination. Heavy pressure and excessively coarse grit smear and actually cold weld a thin layer of metal at the surface. We bond produced in.this manner has sufficient strength to prevent penetration by. dye but will open up inmediately under application of load or in the case of re-pair welding. We detrimental effects of abusive grinding can be alnest entirely eliminated by finishing off the grinding task with a fine grit abrasive nounted or held in a mediu:n which will flex or fracture if too much pressure is applied. W e size of abrasive needed to aliminte stress corrosion at the ground surface seems to vary under differing xnditions. A grit of 150 mesh size will not, under nest circum-ctances, pumeLe stress corrosion. If a surface has been harshly treated under coarse abrasive, application of very fine abrasive sufficient to just remove the coarse grooves will not renove the stress fields under those grooves and SCC may still occur. It is necessary, then, to follow coarse grinding with a graduated sequence of. abrasives which will end with a size finer than 150 mesh. Incidentally, the act of following a sequence of sizes will achmily save tine in the rencmd of ' deep grsoves and scratches. If the corrosion action descr N a above is to Ee stopp x3, all of the metal containing the existing cracks must be renoved. Bewelding on this same pipe has a high probability of producing another fully sensitized plane next to the new fusion zone. Any happenctance which can place process water in contact with that sensi-tized plane can then initiate full penetration of the pipe wall by IGSOC. It is not practical to attempt solution heat treaunent and water quenching of the original plpe material. Recamendations he piping for this system shcadd be replaced by either pipe which has been fully solution hest treated or pipe which has a loa carbon content. (It is assumed that the original stainless steel pipe will be replaced with stainless steel.) ne intant here is to remove the possibility that a sensitized plane can be formed in the heat affected zone by assembly or field welding. ' We replacerent material should be corrosion tested in accordance with ASM A262 Practice E. W e corrosion test coupon should include the heat affected zone of the base metal after applica-tion of a full-size weld bead using the same procedure and technique as will be used on field welds. Surface finishing or grinding of stainless steel, such as blending of weld crowns, removal of defects, etc., should end up with an abrasive of 180 to 200 grit size. @e abrasive grit should be neunted or bonded with either rubber or resin. Crinding tools shall not be used on other steels or netals prior to being used on austinitic stainless steels. We final abrasive direction should be at .c distinct angle to the previous grinding direction so that complete removal of all prior scratches can be verified. R. H. 'Ibdd i Attachments: Figures 1-7 Assoc. Sr. QA Engineer RET:jmn, s xc:' D.P. Dise R.A. Dowd C.V. Mangan T.J. Perkins .G. Yang T.E. Tenpges

July 30, 1982

(./ (- r - QA82925 s u Perje 4 l \\ / g 0.0 2 L Eide view End view y O.106 Top view 1.0 inch Figure 1 Sketch of boat sample at four tim s actual size. Dimensions were taken frcxn the report by J. G. Sylvester Associates, June 23, 1982, by scalimy frczn Figure 1 of that report. 62 crack in9acation Plane il f5 16 ~ Curf f gg p 12 \\,/ y /./ n,_a \\ Y. / 'g s 'j ' Dashed line s indicates \\p - f ! I Plane 67 y,. Figure 2 Schematic of boat sample to show relative positions of the planes studied versus featmres found. No scale . t, \\

5 r /L' (. \\. .gg{a_-- 3 ..g - A ~ ~ l..

• y,. g

. - ~ ~. ~,., 2..M .i .e n w Q w c .p a.. m-y - o t*, , *.i Q H .e s .s-e p , = - ,. - *.f';. C 0 = ,a -.. r. e .8, a .M =, a .,4 N c H m 'g ~,-- g O N ~. W -ri O ..b C W th .C d-d , %2 m [n Q

• . ? '.a.;~

g sees m O CO e .Q .G f 'et ~, 3 5 a m .d d ., a.W.- 3....; l 'A M 6

• -.'.9'*

'4J pg

./.. *%. a. 5" g -

$gl' e9 *s .g

$ -l W & a..'-

e ~~ ae o wm o a./. b o 8 5 ,-p 5 96 oo .g .g m .g 'g ,, 1 su x x_ a na m, ,s w"y. O y .h> n e a .l- .,d, 1*U ** g% e ' &n:: -~~ C .$L < g; s- . 6.. pe. e. lQ =

• r r.

v4 ,5 7 -c4 el U [, gg(' .._ vw3 N

• g C

9 .8 'u Us 8 4 5.pyi,*.*** *r *.* */*M "'.a F.1.*J *.". Q g j;,( N I7.k-, I T.~y,'.^. Q M...,.~, W, %. i. % ',;,..... M N$' h.j,N.b... j..s:.,.. M e 1. r...,,.,p. n.r., '_. t ... g z. g ES. l2 i[w. -. .V.

• ** S: n?,!-

.. r; f. . t * * ,L o 's.. ) *.,.. ;', ., t..ep..er. ;,

• s.*..

To y. ;.m..c.. Wy. y,y.';4 . y,-.):J O .7 's;

e. -\\...,.,y,,.s 9 e.m e.

.,3,.

3. :,:

.e.,s,.c...c.,.... se r w*<- en a c+.,,.. ,...,... S

• '.*... : W.~..

.f.e, s s...~ -,r 6,.Q.;, a:_ 2.,,., p( *

e. a

...m. V %, F,y, f.J.. s.; .,. : w ' ;M " ' ~..

• ?.

p.. ; y.. g _e. _-., - - o 4 .gp-x,v . I;.a,e. m..s::c. : s.,,,*. s } i.W,.y &;.. 2 : s ;,.p... 4> %.. .. p.-s.:?,%.y -;**).2& - :g. 3,~ *, e.f 'f. y )S .;( :.s;q. ; *-:..p.%e&>~ a, Qy. g,2-l:,i '8

• t 4 >,~

....vs..

....r w. f y " r G *.f.1' C s :'.0..h ': 'A..';' :;.=*? ';?.:&. ; Q

k..' *.'./. y

'.:.,R' ' L..'~ r..'.u-f..t.p..f.v. -t.s?,H.

O d r.s.. e v p.. ; s.

u.. s -

..: - ~ ,.s. m ... ; ;; c - -,.,:e'%.~.1. -h"v;.*.u. :O, s. 2. :e y .z.i....t '. v :.? Q. A *

.r-;,,- Q pa; ^.L

.: a., &.. w Q.. - 4..? f :W.n,.t

n..,v

. ~ ~. :- n e :. &..s$gh 4, ....;;,\\ m.: 4;.,2;.W.;; M '9:r.. r\\.,;.s:,,..m ,...:. ;-. q.-r -.q g;g M.,, :,. c'... g. :,.= .s:,. J4,p.

f. W-lt.

.:w n -

t..-:.:,, n%;# y :. :... m :, o

. i.u ; e v.:v-' x.:. ?.~.: m.. k.g w..? rf:r;. @ W C. - Sc 2 ' M ,c c;.;: e Q .f..,q, y,,:. qRet.g::p g g. . W ::f7,. %:: Qu,,r. .:.c;:..n+.s-c.:.c.c. -w r e c..r. 4;.r y. sd.'.if.~. n e- . a.1. : .. yw:. p ::y. Q..?: u,...,.f ) Q. ~. z.? -::q w. o . - s y: g ;.w :.s ... ~, w n em - a ,..., y a. - =t m .m.m- - ;a .f.m.. -p.w p~. h Y' M. %r, ge.w>m%g y.a-a:.:y o b "-y~9.~.,:.i;*Q.t.y.gdQs&y a- &u y ggs ; o, m

k., ~..,.?.;$: kk;:.f.rhk..:c. n

-.?Y YQt..%,,,. kN8 n w - m 5 .s v- ,u a a% %. : - ?.t ;:c: ;..m 9.c.t.e A o a sA .c D 3 -;Od.N<.m.U$.5M.h.,m e!M m, h, M,, a. ','- ?$p'. "

• ,..3 *a.>

p-lJ ,n.. : s. M, e.u m r ..n... Q,Q......?.'.4.?*.. i.. W.,....%. w M G.: '. %. ~ o o,7g,.. . f. v. .,ya '.a'.'. A...5*a. M a. ~..- w m..

m. ~,I b x.-:g g

. s. : w.. n wa ' OT-W itiO M E .. n m. s -...:.w.x.ww.,a

7N July., 1982' QA82925 Page 6 h 9 v: y,.mt..'~ ?O.$. l n j k ?'I fy. { g,f,g hh Y ~ \\. f'S*b + ,\\ '.g.,. \\ = n s *g 'n ' . y.. 4? lh '

.y.:.

y tG:) .. y 8:q.-) ,s) - x,) ~ / I s :,, .% f,~. k *y :o j#i,. j,.{l($ik fM*.,.: ( .I~' l { ' 's f .m

54 W

J /.' ,..,.l, $j l h2 2[. < Ji -2 . Y g i.d ;,[ l Figure 6 at 250x. Plane 17 etched with 10% Figure 7 at 100:c. Plane #7 with same etch as or.alic, 6v for 90 seconds. (dpivalent to figure 6 at left. At this magnification an interconnected ditch can be seen paralluing M/cm ).

• 1his is the etch described by the weld metal-base metal interface. 'Ihis ASDI A-262 practice A.

Besulting structure O. indicates the metal in this plane is fully shows intercentritic ditches in the weld sensitized. 'Ihe' top end of this ditch can be l and a dual structure in the base metal. seen in figura 6 as a groove in the lower center portion of the picture. 5}}