ML20058G860
| ML20058G860 | |
| Person / Time | |
|---|---|
| Site: | Byron  |
| Issue date: | 07/30/1982 |
| From: | Bunch R SINNISSIPPI ALLIANCE FOR THE ENVIRONMENT (SAFE) |
| To: | |
| Shared Package | |
| ML20058G812 | List: |
| References | |
| NUDOCS 8208030435 | |
| Download: ML20058G860 (37) | |
Text
{{#Wiki_filter:. . t
- July 30, 1982 C O -2 gg 47 UNITED STATES OF AMERICA -
NUCLEAR REGULATORY COMMISSION T BEFORE THE ATOMIC SAFETY AND LICENSING BOARD , In the Matter.of )
)
COMMONWEALTH EDISON COMPANY. ) Docket Nos. 50-454
) 50-455 (Byron Station, Units 1 )
)
and 2) AFFIDAVIT OF RICHARD BUNCH - The attached statement constitutes my testimony in the above captioned proceeding. The testimony is true and accurate to the best of my knowledge.
~
+
. Richard Bunch .
Subscribed and sworn to before me this<84ed day of Jul.y, 1982. 7 n q?dh
/ / .8AAou 5NAs Notaryy@ubl'ic My commission expires:
hfti(/9,'lfbV 8208030435 820730 PDR ADOCK 05000454 0 PDR
I, Richard Bunch, affiant, swear that the following facts are true:
- 1. I have read the following material
- a. " Materials Performane in Nuclear Pressurized Water Reactor Steam Generators," by Stanley J. Green and J. Peter N. Paine , in Nuclear Technology, vol.55, Oct. 1981, pgs. 10-29.
- b. " Corrosion af PWR Steam Generators," by R. Garnsey,
~
in Nuclear Energy, vol. 18, April 1979, pgs. 117-132..
- c. "Effect of Some Environmental Conditions on Stress Corrosion Behavior of Ni-Cr-Fe Alloys in Pressurized ,,
Water," by H.R. Copson and G. Economy, in Corrosion, 1968, pgg. 24-55
- d. " Operating Experiences: Steam Generator Tube Performance During 1979," by 0.S. Tatone and R.S.
Pathania, in Nuclear Safety, vol. 22, no. 5, pgg. 636-655
- e. " Effects of Copper and Nickel Compounds on Corrosion of PWR Steam Generator Materials," by Earl L. White and Warren E. Berry," in Nuclear Technology, vol. 55, Oct. 1981, pgg. 135-150.
- f. " System Chemistry Considerations for Nuclear Steam Generators," by Frederich J. Pocock, in Nuclear Technology, vol. 55, Oct. 1981, pgg. 117-123
- g. " Material Options for Steam-Cycle Heat Exchangers,"
by Robert G. Schwleger, in Nuclear Power, June 1979. Pgg. 55-511.
- h. " Materials Requirements for Pressurized Water Reactor Steam Generator Tubing," by P.L. Berge and J.R. Donati, in Nuclear Technology, vol. 55, Oct.
1981, pgg. 88-104,
- i. Memo from Frigyes Reisch, SKI, to Joseph D. LaFleur, NRC, Jan. 20, 1982. .
- 2. I have attached a true and accurate copy of the relevant portions of the aforementioned documents, and ,,
incorporate them as part of my testimony. 3 I have not altered the attached copies in any way.
- 4. I have merely photocopied the documents on a standard photocopy machine.
5 I am not establishing an expert opinion about the content or application of these documents.
- 6. I am selecting facts from these documents.
~.
7 Facts selected from the attached documents are cited as proof to support this response in opposition to NRC's motion for summary disposition of contention 9(c). O S
Il li STATills KARilK".WtSPEKT10!1
- St::kholm Jan. 20, 1982 .
*r h"f 't./'
- swtcisM NUO'.EAM PCM 437EC*CRAT:
/ 1 1
Er ref. va, ,es. R3 - 3.5.4.2-1430/81 Dr Joseph D Lafleur ' Nuclear Regulatory Commission
- Washington D C 20555 a
ggA ,, 1
'n
- i. :
- s. ,
Dear Dr Lafleur,
c
', s. -
3
Subject:
Statum generator tube fretting in Ringhals 3 '.
' (f For your information there is a PM ettached here to present this inspectorate's position on the subject matter. ,,,
i Beside the two reactors in Sweden there are six reactors 8
- in the US, six in Spain and .one in Brazil equipped with -
j the same type of steam generators,* = i !
, - The utility - the Swedish State Power Board - has sent us l
last Dacenser a. plugging application based on Reg. guide 1.121 ; (some material from that is attached here tool . The app-- (. l lication has been rejected. No resumed operation request .. .. for Ringhals 3 or core loading request-for Ringhals 4 has /"* e been received yet, We,will send you later the information you asked for in t your letter, dated December 17, 1981. -- [ [(s
- Should you want sogne further explanation, or should you want additional information, please don't hesitate to s
[ contact us. , Yours sincerely, {, ' Fd ec + T c;5 1 . i Frigyes Reisch , k l
' ces Leif Ericsson Swedish Embassy, Washington i
I I i t 1 . I s T*W8e* . Tebesmeme Lamener . Cows pocas . ,enserewee. Teseseen. Canoe teams
- 0543 5 er Seweseme,mme it its 61 sv.tATOW S see Die swtDATOW. Seeena.eae SIC S STCCweQue TTLFAE i
andt e as
, . . . . . ,. ..==.;==.****= ,
'*****P******"**~*.~'
- ', * ~~ ~ '
/
51AilH1h2H KlfAFTIHSPEKliOH ~ ~ s'n ['s swto s= aucusa ,owen inser:ren.,s 1982-01-19 1(2) i
. t I
e Summary of the Swedish Nuclear Power Inspectorate's po-sition on the steam generator tube frettings in the 7 Rinehals 3 clant. ;
- 1. The fretting in the Ringhals 3 steam generators !
(Westinghouse type "D3") is a new phenomenon, and is of ' I a generic nature. The fretting endves very fast. Leakage
-in Ringhals 3 occured after 3 000 EFPH, i.e. less than the usual interval between inspections. . (,
I 2. The damages in the first rows in the steam generators are very extensive. Most of the damages occour in the inlet part of the preheater section, (between baffle plates 8 5 and 6) but single cases of fretting can be seen also. -
\~ higher up i.e. between baffle plates 11 and 12. f
- 3. The cause of the harmful fretting seems to be . vibrations, but the exact nature of that has not been shewn yet.
- 4. The availa$ problem le safetyisstudies considered indicate to bethata e,afety small problem. All primary leaks increase the risk for large accidents. There is also a risk for simultanious multiple tube leakages in one or more steam generators. The tubes are the only barrier for release to atmosphere in certain operating conditions with leaked fuel.
. 5. The problem can most probably be solved but no plan for j modification has been presented yet. Operation at reduced power with a dant;4d steam generator lika.Ringhals 3 is
(. f- . 1 probably possible. It remains to be shown however that roch operation can bee made without significant further fretting. .
- l. 6. Plugging criteria in accordance with Reg guide 1.121 a
, (J is not sufficient as that guide was developed to take care of other phenomenas. This regulatory guide is suffi-8 cient to assure initial tube integrity, but special analysis
. are required for the effect of increased gap between tube
' and baffle plates. -
7 Full scale model tests are very important, in addition
- to the scale tests performed by Westinghouse. The Rino5als 3
. owner utility, the Swedish State Power Board, is at present j conducting such tests in X1vkarleby, Sweden. There an 1:1
. scala model of the preheater inlet section - between
- baffle plates 5 and 6 - is in operation now. The model
; will be extended further to contain 8 full length tubes.
i Computer models have however failed to predict this 1 phencmena. , t e 7..... .-. ,s t 4 . /
J
- 1. , 2(2)
- 8. In May, 1981, eddy current testing was performed by Zetec on 14 tubes in row 49 after a water ha.c er incident.
No dents were found on the preheater tubes, but "it was suggested that several tubes exhibited signals which should 1 be monitored at a future inspection" SKI was however not ' informed of this recom:nendation nor was obviously NRC. The. tube which was showen in May, to suffer substantial wal3. thinning was that one which then leaked in October. L c . ul . e .
# I.
i
.(
4 l
- h %*
I I
< 1 t
i I I I I. -.... . , . - . . . - . . . . . . -
. /
e
- l l .
MATERIALS REQUIREMENTS FOR .mmm PRESSURIZED WATER REACTOR STEAM GENERATOR TUBING Ph. BERGE and J. R. DONATI Electricitide France Direction des Etudes et Recherches 2 Rue 1.ouis-Murat, 75384 Paris, France Received November 10,1980 Accepted for Publication March 12,1981 M "r"m W S.,"_Mi mmamm.mm W.<. m m M7M a m wW L, _R T ~~M aHCT the different vendors show marked differences in Considerable research and development work on in their resistance to corrosion in normal or faulted the behavior of steam generator tubing materials has operstmg conditions. Our p'trpose, in this paper, been performed during the last ten years. These ss to compare these properties and show how they studies relate mainly to such austenitic alloys as meet the requirements of thns application. 1810 stainless steels. Alloys 800. 600. or 690, same- ,, times with different heat treatments or surfacefinish. fl. MATERIALS FOR TUBES-DEFINITION From these studies. tentative conclusions on the best ggg pggpgggggg selection to be made for this application are drawn, the different candidate materials being compared with respect to iI.A. Material Choice-Meat Treattnents
- 1. mechanical and thermal characteristics that
' *n the r mecha ical nd thermal properties can act on design and manufacturing processes and on the.st corrosson resistance, the best materials of steam generators f r steam generator tubes appear to be only austenit:c stainless steels and a!!oys. The ,mittal choice was
- 2. susceptibility to corrosion product release in 18-10 stainless steels (AlSI Types 304,316. and 347s.
the primary coolant, which is responsible for Thus, the steam generators of the French CHOOZ a good part of the irradiation dose received power station are equipped with tube bundles med: by the operation and maintenance inrsonnel of Type 316 stainless steel, which have opetuted
- 3. corrosion resistance and. in particular, stress satisfactorily during 10 000 h of service. Nonetheless.
_ corrosion susceptibility in different solutions these stainlcss steels are known to be very susceptible that are or can be met in operating conditions. to stress corrosion cracking in~a number of environ-ments and notably in the presence of chlorides and
'^
oxygen. Different operational problems have stimu-
- o. m
+a r <
resistant to this type of corrosion. Consequently. Alloy 600 has been mtaduced. I. lNTR0 00CTl0N.. and since 1967 has equipped a large number of steam generators, notably those produced by Westinghouse For the last ten years, considerable research and Electric Corporation (WEC), Combustion-Engineering. development work has been performed on pressurized and Babcock & Wilcox. An alloy with intermediate water reactor (PWR) steam generator tubing mate- nickel content, A!!oy 800, has equally been con-rials. They relate mainly to such austenitic alloys as sidered for this application and is used in the Federal 18-10 stainless steels, Alloys 800, 600, or 690, Republic of Germany for the power units built by sometimes with different heat treatments or surface Kraftwerk Union. finish. It is well recognized that all the problems that Laboratory tests and the results of service ex-have affected the PWR steam generators cannot be perience show nonetheless that Alloys 600 and 800
. solved only by good material seicetion. Nevertheless, e:m be, under certain extreme conditions of mechan-the alloys that have been, or could be, selected by ical stress or contamination of the secondary water.
88 NUCLEAR TLCitNOLOGY VO L. 5 5 OCT.1981 0029 5450'xilooto-00n502 00/0 at981ANS
1 Y. Berge and Donati MATERIALS REQUIREMENTS FOR PWR STEAM GENERATOR TUBING susceptible to stress corrosion cracking.To overcome higher than Alloy 600, to perfonn more this problem, studies have developed in two directions satisfactonly with respect to several corrosion simultaneously. phenomena 3 2,45% ferrite duplex stainless steel, considered 1/. A.J. Improrement of Performance of because of the better resistance to stress
' Existing Alloys corrosion of this type of microstructure.'
Improvement of the performance of existing It should be emphasized that, while Alloy 690 alloys notably seeks to reduce as far as possMe the does not present any major difference with respect residual fabrication stresses. which form a major to Alloy 600 for industrial tube fabrication, extrusion part of the total stress in service. The different f and heat treatment of the duplex stainless steel are solutions envisaged have been described and com- both very delicate operations. pared in detail elsewhere.2 As far as Alloy 600 is i conce'rned, cooperative studies carried out by Elec- II.B. Properties of tne Different Materials tricite de France (Ed F), Framatome, and WEC have led to the adoption of a 10- to 15-h heat treat- ff,g,'l. Chemical Composition ment around 700*C (715 2 15'C) on straight tubes in the factory, completed with a shorter duration The specified chemical composition for each treatment at the same temperature on the bends of different alloy is laid out in Table 1, and for Alloy
<250-mm radius. 600, the particuhr specification used for Fr.ench This treatment has several effects, which are power stations is shown. One should note the reduc-dealt with in more detail below. tion in the maximum carbon content and the desire to maintain the chromium content in the upper
,1. It greatly reduces the residual stresses due to . part of the allowable range (to improve the resistance straightening and belt polishing of the tubes, as to intergranular corrosion) and the cobalt content
- well as those particularly high stresses found m small as low as possible (to limit the presence of "Co in radius bends (see Sec. III.C). the primary circuit).
l 2. It removes the cold work introduced at the surface of the tubes by the same straightening and ll.B.2. Mechanical Properties _ polishing operations and by sandblasting the internal Table 11 compares the tensile properties at am-skin, which results in a reduction m the stress level bient temperature and at 343*C for the different m large radius bends (see Sec. Ill.C) and probably materials under discussion. Note that Alloys 600 and a reduction in the rate of corrosion product release 690 have the best characteristics among the austenitic m the pnmary coolant (Sec. Ill.A.1). materials.
- 3. It makes the alloy less sensitive to intergranular Generally, the results for finished tubes satisfy corrosion by selecting a sufficient time at tempera + the minima specified vithout difficulty. In the case ture to rehomogenize the chromium content in the of Alloy 600, for which we have a large numbcr of vicinity of the precipitated carbides. In addition. test results, it has been shown nonethekss that it is the quantity of chromium carbides likely to precipi. desir:21e to avoid the carbon contents that are tate during subsequent aging at the service tempera- sufficiently low to allow a total dissolution of chro-ture is considerably reduced (see Sec. Ill.B.5). mium carbides during the final annealirg treatment (cr.rbon < 0.0107c for 980*C, for exampr:). In such a
- 4. It improves the m. trinsic resistance to stress case, one would risk, due to grain growth and to the corrosion in caustic solutions, and probably also in absence of intragranular carbides, obtaining weak pure water, m:h the hypothesis that these tw values for the yield stress. Heat treatment of ~10 h phenomena arc essentially. of the same type Oce. at 700*C can equilly produce a slight decrease in the for example. Sees. Ill.A.2 and lit.B.3). yield stress. However, the maximum value of this drop (~50 MPa) applies to the tubes with the highest
//.A.2 Comideration of Other Alloys degree of cold work due to straightening and which have, therefore. a yield stress well above the specified Because of the importance of tests needed t minimum.
qualify a new material for service use, solutions of Alloy 690 is often delivered with a final annealing this type have taken longer to develop than the heat
, temperature on the order of 1040T. This high value treatment referred to above. The studies have been does not appear to be essential from the point of view carried out principally on the two following mate- of stress corrosion resistance, and it would appear preferable t'o use the same temperature as for Alloy
- 1. Alloy MO, with an austenitic structure, which 600 M80 C) to benefa from the improvement that is likely, as a result of its chromium content can thereby be achieved in mechanical properties.
r
, Berge and Donati MATERIALS REQUIREMENTS FOR PWR STEAM GENERATOR TUBING ///.,l.2. Stress Corrosion Cracking in the of boiling ' has led to many different types of corro-Pnmary Aledium sion. Material resistance to each type of corrosion After various operational periods, the stea must be considered separately.,as their mechanisms generator tubes in Alloy 600 of several PWRs have , are different.
shown evidence of cracking starting from the primary
///.9.1. Corrosion by Phosphate Solutions side. These defects are found at the ends of the q roll expanded zonesia (in such cases, they propagate? The phenomenon of acidic ' sodium phosphate p:rpendicular to the highest residual stresses resulting. concentration leading to a general attack of the tubes from the mechanical expansion), or in the vicinity of. by the formation of iron, chromium, and nickel the tube support plates when denting has occurred,'3! phosphates, either nonprotective or soluble, has or also in the upper part of the tubes bent to the) been studied at length.:o.:: The different materials smallest radius (first row). 2 These defects are attrib-j compared appear to behave very similarly in these uted to intergranular stress corrosion cracking; concentrated phosphate solutions. In many cases, despite a chemical composition of the coolant gen j treatment of secondary water with phosphates has erally considered unlikely to provoke this type ofi been abandoned.
corrosion in austenitic steels and alloys. But since, 1959, Coriou et al. had drawn attention to the l ///.B.2. Pitting Corrosion risk of stress corrosion cracking of Alloy 600 under t high stress in pure, high-temperature, oxygen free ' Model boiler tests have shown that when feed-water. These results have given rise to long controver-i water is contaminated with seawater, pits can ap' pear sies between research workers for nearly 30 years. on all the austenitic materials that have been tested. It is now accepted that this type of corrosion can in contrast, pitting ~ has rarely been observed on steam affect Alloy 600 (Refs.15 and 16). The time to generator tubes of nudear power stations and, as rupture varies greatly from one heat to another
~
far as we know, has never given rise to leaks from - and appears to be very sensitive to the test tempera- primary to secondary medium.Therefore,it does not ture between 290 and 355'C (Ref.17). The cracking appear that resistance to pitting corrorion in service mechanism in such a medium and the scattering of conditions is an important criterion of selection for the results have never been clearly explained. We have these austenitic alloys. In addition, comparative introduced the hypothesis that the same phenomenon tests between Alloys 600, 600 thermally treated, could be studied in dilute sodium hydroxide solutions 690, 800, and sustenitic stainless steels carried out (0.4% by weight), bearing in mind the numerous at EdF have shown that resistance to pitting is similar similarities between the phenomena produced in in high temperature chloride solutions. Low-tempera-both media.'8 Corrosion in dilute sodium hydroxide ture corrosion tests have shown the favorable effect solutions is presented in Sec. Ill.B.3. of molybdenunt for 18-10 stainless steel and the The phenomenon has never been observed on unfavorable effect of thermal treatments giving struc-Alloys 800 or 690 nor on stainless steels. Coriou tures heavily sensitized to intergranular corrosion et al.8* indicated that it only affects austenitic alloys (I h at 700*C for Alloy 600, for example). By extend-with high nickel content. Heat treatment at 700*C ing the duration of the treatment at 700*C to 10 to gives Alloy 600 a more resistant structure to this 16 h, this unfavorable effect is suppressed. It was also type of corrosion *"*" and greatly reduces the / noted in these tests that the surface preparation of level of residual fabrication stresses in the tubes. I the tube markedly affects the susceptibility to pitting Thus, it reduces the risk of this type of corrosion ! corrosion: A high roughness has a very unfavorable appearing in plants using such thermally treated effect' as was noted earlier for stainless steels in tubes. ferric chloride.23 The risk of " pure water" eracking of Alloy 600 after a certain time, which can indeed be long if - ///.B.3 Stress Corrosmn Cracking the stresses are reduced. is certainly a weak point for this alloy com pared to the other alternate #/.BJ u.in //re Pmenn of C/dorides. The favorable materials. - r le f nickel with respect to stress corrosion of austenitic alloys in chloride environment is wel 111.8. Corrosion from the Secondary Side n wn. The main reason for adopting Alloy 600 for steam generator tubing in seawater-cooled plants The greatest number of observed incidents of was its resistance to this type of corrosion. The corrosion starts from the secondary side.":" The 18-10 type stainless steels are known to be very presence ol' impurities in the feedwater frequently sensitive to chloride stress corrosion cracking. Duplex due to the accidental in-leakage of cooling water stainless steels are much more resistant.' Nonethelesr. through the condenser and the concentration of these model boiler tests * (with a heat flux of 10 to 17 impurities or other additions on the tubes as a result W/cm2) have shown that a duplex stainless steel tube
{
. Berge and Donati MATERIAL.S REQUIREMENTS FOR PWR STEAM GENERATOR TUBING TABLE V Comparison of Candidate Materials for Steam Generator Tubes to Alloy 600 Thermally Treated at 700*C Considered as the Reference Solution Benefits Disadvantages Versus Alloy 600 Heat Treated at 700*C Better resistance to stress corrosion cracking in pure Susceptibility to chloride stress corrosion cracking water or low concentration caustic solutions y g ; ; ,
Alloy 800 Lower corr sion product release in the primary g g coolant Thermal expansion coefficient higher than for carbon steels Better resistance to stress conosion cracking in pure , Limited industrial experience water or low concentration caustic solutions Thermal conductivity slightly lower Much lower corrosion product release in the primary Thermal expansion coefficient somewhat high.er Alloy 690 coolant gg Shorter 700*C stress relieving treatment possible . without sensitizing the alloy Good resistance to chemical cleaning agents Pressurized Water Reactor Environments." presented at the 17. T. S. BULISCHEK and D. VAN ROOYEN, " Stress INCO Power Conf.. Harbor Island,1968. Corrosion Cracking of Alloy 600 Using Constant Strain Rate Test." paper 183 presented at Corrosion /80-National Associa.
- 9. G. C. BODINE and J. W. CARTER, presented at Corro. tion of Corrosion Engineers, Chicago, March.19,1980.
sion/74. National Association of Conosion Engineers. Chicago, Illinois. March 4 8,1974.' 18. PH. BERGE, J. R. DONATI, B. PRIEUX, and D. VILLA'lD. " Caustic Stress Corrosion of Fe.Cr Ni Austenitic
- 10. G. F. TAYLOR. paper No. I19, presented at Corrosioni Alloys," Corrusion. National Association of Corrosion Engi.
77 National Association of Corrosion Engineers. March 1977. neers 33,12,425 (1977). I1. "EPRI Workshop on Cracking in Alloy 600 U. Bend 19. H. CORIOU, L. GRALL, C. M AHIEU, and M. PELRAS, Tubes. Denver, Colorado, August 20 21,1980.
" Sensitivity to Stress Cormsion Cracking and Intergranular Attack of High Nickel Austenitic Alloys." Corrosion. National
- 12. H. J. SCHENK. "Expirience d' Exploitation des Ginfra.
Anociation of Cormsion Lngmeers,22,280 (1966). teurs de_ Vapeur de la Centrale Nucliaire d'Obrigheim," presented'at Journies AIM de Liige. October 1974,
- 20. R. GARNSEY, " Corrosion of PWR Steam Generators,"
- 13. D. G. EISENHUT. B. D. LI AW, and J. STROSNIDER. N ict. linergy. 18.2.117 (1979).
" Summary of Operating Experience with Recirculating Steam Generators," NUREG 0523. U.S. Nuclear Regulatory Commis. 21. A. BAUM. "The Mcchanics of Concentration Processes sion (1979). in Recirculating Nucicar Steam Generators," presented at the ADERP.EdF Mtg. Water Chemistry and Corrosion- in
- 14. H. CORIOU. L. GR ALL. Y.'LE'G ALL. and S. VETTIER. Nuclear Power Stations.Scillac. France. March 1980.
Trans. 3rd Suclac Atcrallurgy Cullouturum. France. 1959, p.161. North llolland Publishmg Company (1960). , N. PliSSALL. A. B. DUNLAP. and D. W. FELDMAN,
" "" ***"'" Y ' ".' ^ Y "
- 15. D. VAN ROOYEN. " Review of Siress Corrosion Cracking of Inconci 600." Carrosion. National .issariation uf Curnniani **U*'"'.* "'U" * ' " ' " * ' ' (".
" ""'"'" """#'^""'
Engineers. 31. V. 327 (19751
- 23. Pil. BERGE. l' roc. 3rd Int. Gurgress Atcrollic Corrusion.
16.11. A. DOMI AN. R.11.1 M ANULISON. L. W. SARVliR. G. J. TilEUS. and L. KATZ. "Eifect of Microstructme un Mauw. USS R. 2.406(1966). Stress Conosion cracking of Alloy 600 m liigh Puniy Water." Corrosion National .lssmiarion ut Girrmion /;nrinecrs. 33. 24. Electnene de Franec/CEA Technicatome, unpublished
- 1. 26 ( 1977 L result s.
wucu uuwevumma wtutJut_mstm _ __ __ _ _M
- Mi i
ez1979 Vol 123 No. 6 fOWer_ * .
- "p sM%?q.$
- ver.~. . a s.t s 1~J
'I hh.k:
' . . . .ws. y W[h'8
' A - w:: q r - .
k . [. . . b l 1
~$ p,. . .
'l$g y$ c .t.c :"rnw:-<,g; :
.% c .
.e . . ,. : ..
' l':- . .v . * '%C.Q 3 ., , . , i dr. g M4 '. #5MM m,' , m. m
- pf'S
*- d.' ['kIk -[s -
i specialty stainless stects for con-demers
~
.d sy,, - ; i The key eyele components are 1
,, , (N ,Q- %v pD !
4 steam generators, feedwater heat-llp.'
- ~ ia 3 y i
' .- 9. : . .
3 , , 3 c.. ..
.;., m - ,
;; . *j . ; J p' q g j g g Q' ) ers, and condensers. 'I he discus-uon of steam generators is divided
/
4 .,
.' qts #' a into ections covering fossil-fired
* ',j j l
,'yp-h.
3h sasimy eruce somhoom ,e
.g j ,
; . 1
., ,pt b ,
ll, ,
~l
',y and nmicar svucnn, became of ihe dntmct ddlerentes between are tifitenme demands on '
- 7 .'
! the two, the discussmns of fe'ed-s 7 e + , ,
*".4 the raator e.v pments med q ,
j water heaters and condensers ap-fC ' ll' 3, j ,- >
- r. . c
. . sicam.p ner.iimg w.te n n at [ , ;
) ply to both nuclear and Imsil i{
j c i , l ,.ntrat stanon powe r plants 9 ,
'i ,-
q-y 4 j > 1 ,- 7
.4 '. 3 apphcatium, escept as noted.
YYi picsc are conipinnded by the 4'
- i ; '
1 lhn report draws upm infor-viluted sonditions of todayi cool- mation dcrived from many indus-
![ 4
,.g..aier uipphes, and by the
'.I
, J ' '
? J irial and laborainry sources, to 3 ,
1., ;. uianons impned by environ- : f
' which we are indel-ted. Represen-on treat. tatives of the follommg manufac-t
{g unt.d sonuderatiom used in plant 'l j t uring oritarniations enntributed
.cnt chemical.s
,,.- j
,csond.iry spiems otenovely to the sections indi-4 These sondmons produce a vari-
' I .
, cated liabensk & Wilcox Co and s- l
' ' ' *h .
p of harmlui stiests on operating , p ( 'ombmuun I:ngineenng Inc (fos-in ludmp tempera- n ut and nm. lear acam pencratorsl;
' % " ,,unponent
,~' ia re and prewure stresses, strew [ Westinghouse I;lectric Co. Joseph and erosion. Oat Corp. and Whitlock Mfg Co
. 4 ,y sluced coe rouon.
.. g (,1carly. this calls ior thorough lll dl ll nuclear sicam generators); Foster i
- 5. j (Wheeler linergy Corp and Fosterl f 1. Wheeler E($y 3,alisn ot the cap.ibihty, availa-$j Mts. production control, and per. 1 td (feedwater heaters, g
- y. brrnanse of the !cading candidate g J, condensersh in particular, we ac-
' ,* . . , yterials, to ensure optimum knowledge the awiitance of *.icss- f
" . .,':[ pnt rehabdit . Without intelli- l l'- ers J A t hitty. L I;. 'steele. N1 J ]
dt gent selecimn, m.nntenance re- lleil, it Van I cer. C K Paulson. ! M remenh san kyrosket. Resuh W (; Poubon. K D Kolberg, W H l( a
' 7.~. '? I bviended plant oma ges, addmg i aunan. Ni A Cordovi J R Pres-
. f. [. * (,ndroh ..i t hem.inds nt .hdla'rs ~ lev. A I; e innck W l flow. 1, C i g
i
',, lN uh tu piani opcraung sous ggggg '"4 f a lineberger. \ \1 Impagli.ir7o. and
.: M.'f thi repot w as umlertaken m ! fl P Wdhamt {
,;-(. se k now led ge that no ungis fg l
. _. J
; l
.aurse addre es the Inll spectrun' Repnnts of this and other Spe-
, '"* W f fistors that impast on the selec- [ g l 4
!, d.j cial Reports are available at
$ .vn of matenah tor ma,mr ueam ~lIEEiI E nommal cost. For a complete
( onditions es e
- dr.:*g* ,,cie componenttrn.ned are deugn. operanon, anti @ gpnce hst, wnte to:
gt' Power Repnnt Department
\taterials opuons s x 35 1% t Ave of the Amencas
.. ,"'* W , reg.
e fromrelathe tel.nue.d ed atntenitis ' y' I E New York. NY 10020 e.. ; , pt le ntic stech for boilers to ths ggj Qg g BBQ m , __ , , , _ _ . _ m a in
,-. y uaa il 5
. sc.cr. more esoue titanium ane
~' Robert G Schwieger, Assoaafe Editor . Sheldon D Strauss, Senior Editor
.. m a
; gryp~gl < 24ykY??Q& A 5 f,%}~& *Q*Q M, y l
was avoidance of furnace-sensitized radiation embrittlement. The first h h eration stainless steel and care in fabrication to under close study, with the hope of rnci- up, plai i remove trates of potentially damagmg mi/ing such etTetts I he second pnubih- to mim j j M ((( l pickling and cleaning agents I ow alloy ty was recogni/cd early by supphers of stects were selected - A50M Class il for nuclear steam i> stems and regulator)
. mulatu gener.e forgings. A53.1-B for plate, with %-in. agencies, and a surveillance program .as Reci cladding of type .108 stainless. Improve- prescribed to detect this. and thc onset of steam I ments in materials and fabrication have other effects (changes in toughness increased yield strength from M.tMMIi wi tensile stiength, etcl. It involves place-to 75.000, and even 90.0tx), and radi.i. ment of metal coupms near sensame tube u first st suppm O
u0I88f eeaeeeeee tion >ensitivity has been reduced by pomts of the pressure vessel wall as stan-dropping the NDT (nil-ductility transi- dard procedure before initial startup. tion) temperature well below operating Reacting to radiation levels several tirncs I ing. II.
,mg pr.
with 8m temperaturcs, and climinating copper that impacting the base metal. ard L from wclds and base-plate.'d examined at periodic mtervals, the As will be discussed below, water coupms will indicate changing rnetal l sttribi After Alloy 808f t0fS chemistry shares the spotlight wit h properties well in advance of pmsib4 or C metals in the nuclear steam system rupture. Goal of today% technology and l choic. (NSS) with respect to operating reliabd- the guidelines of Sections lit and XI of heat ity over the plant's 40-yr design life. the ASME Code is to eliminate .ny fine 9 111l11llllllllllll Since nothing can be added to the RWR threat to the integrity of the pressure steam circuit, the goal is to maintain vessel over 'ts design lifetime. prove to v neutral water chemistry at all times.
- ate I ll Moreover, every etTort is made to limit Pressurized-water systems -
oxygen entry in the external portion of The PWR (Fig 13) features an indi-cont depo fired the loop (feedwater heaters, condenser, rect steam eycle, with a steam generator materials for nuclear application connecting pipe), to mimmi/e corrosion separating the primary loop from the men S tarting point for the evolution of was already-existing steam gener- and consequent deposition of corrosion secondary Consequently, maintaining a i um ation technology. Elevated temperatures products on fuel-assembly and other compatible environment for the preuure. impi and pressures were, of course, the component surfaces. vessel metal n a simpler matter here age common denominator, with nuclear ra- While full-flow demineralizers (supple. than for the RWR. Oxygen dissolved in l' diation providing the added variable. mented by reactor-water cleanup sys- the coolant is held below 100 ppb cy of The boiling water reactor (BWR) is tems) are used to deal with corrosion using deacrated makeup water, and by esse
, the nuclear counterpart of the fossil-fired products formed uiwtream, the chenus- maimaining sutlicient_ hydrogen m solu- , frei boiler (Fig II). The essential dilference try target for feedwater to the core is tion to suppress the ouygen that is II"'
is the use of a large, water filled pressure 50-100 ppb oxygen. Note that water produced radiolytically from the water. , wat vessel,in which steam is Fenerated under undergoes considerable decompwition in I.ithium hydroxide is added to adjust pl{ bec the effect of the nuclear chain reaction the high-radiation hehls ut the core, to the desired level. cat i occurring within. Materials for early hence designers have had to deal with The steam generator and balance.of- tra i BWR-as well as PWR (pressurized- 100-300 ppb Or in the reactor water. plant, however, present a far more tai j water reactor)-types evolved on the Two areas of potential conecrn still comples situation. Careful materials - mr 4 basis of experience in marine-propulsion, exist: the emibility of crack formatma Selection must be combined with rigor- , I r ntral-station and petrochemical-pro- and growth in the vessel, and neutron- ous water-chemntry control -during op-
- fa' , ce3t technology. For the small demon. l strai;on units, the materials of choice for y l
the moderate temperatures and pressures 11. Solting-water reactor tBWW system produces steam en massive, thick-wull steet 12 involved were carbon steels A202 and pressure vessel. Neutral water Chemistry and Careful Control of oxygen entry into Us m. l A212-B; an austenitic steel cladding external circuit protect reactor internals against deposition of corrosion produe:a (18% nickel, 8% chromium) provided f;t protection against corrosion, as it does to vansurnevarator ter e ier 1 this day.' s < s As unit sires grew to the 200-500-M W src.gn M-- > ' range, the need for strength and tough- ,, h f
, ness in thicknesses of 3-5 in. dictated the neacior - r-- --
r7 ve"'d> , e use of A302-B, a low-alloy steel. Options go' ""'""" ff r"U [-7 (f . .
, l".
existed only in the method of fabrication: , T ~ Separ dors f Assembly from flat plate sections a f- I ""d ' " r * * ,_, [ {] _ h: l I, forded high-quality constructmn by ex-perienced domestic Suppliers, but placed s f_.
, y,, yy, /l .
conven.e, h a
' longitudinal welds in regions of maxi. Corc k7 ,
)_
\ [~1, , _ _ f~ "
L mum stress and radiation levels; ring- = 01,.u.ron sican, [ section forgines avoided the latter prob- conden .no e __ b j , lem, but ran the risk of quality-reduemp / W """er
- cu ,
it plant scaleup to todayi iUOO- N , j-(, MW plus sizes came the need for larger nycy c u,,,,on L. ,] q m, - h} vessels and greater thicknesses-8 in, for pumps A q - l jg the belt line cylindrical course ennstitut-e{ h(1 {- - x
*k 4 h plants c niinued to use it, and to expen- exchange demineralintQa.v,elame into
,d cration as well as pre operati nal start-first .
ence tube wastage. Most others changed "<- _dern-bed =ad necce fhe typet, of mi
- J p, plant outages, and extended layups .
1 to mmimise system cornnion and accu- lo a treatment approath that had been The former is preferrcil for its Inrger pg a u!.ition of corrouon products in the concurrently under study as an .shcrna- capacity. and can bc regenerated without
. pliers tive to phonphates but one that did not requiring shutdown if supplied redun-Ifenerator. itecire predominales. Most PWR otter the desirabic bulfering action. It dantly. .A note of caulinn- Da i== .rulat 7,,,**
4 'M mvolved a chemniry based enlirely on " thrown" from runh types have entered d geam generators are resirsulaimg 11-og g some sicam gencrutors and cuumed corro_- q tuoc umts (I g 14). Those used in the the use of volande amines All-volable uRhness, first stanons had carbon steel shcIls and treatment ( AVT) comprises two compo- s_ ion of bot h t'ibinnt and tube sunnarts, 3 g System converting to AVT registered
' % unpport plates, with stainicss-stect tub- nents:(1) use of ammonia or a substitute se, derivative that volatiliics readily, adjust- carly success. While metallic sludges I ,gas a t "" -j eng. ing Before
,arobicm long, however,inaunits materialized tube-crack-cooled ing pil as needed to limit prebuilcr seemed to form rather quickly, there was E' .ith brackish water, a phenomenon corrosion product deposits, and (2) a littic evidence of tube wastage. However, rel ti new phenomenon came to replace it tsi, n 3rtnbuted to chloride stress-corrosion. volatile oxygen scavenger, such as hydra- a bore,c iong. oiscove,cd in ie74. when
,ine. as a fu,ther check on co,,osion..
a i,, ,, 3 sfic, intensive ,ca,ch re, site,nai v ,. ,Becan., c nr a, enacern about loss of steam-generator tubes were found to All y 600-sometimes known as Inconel. g met ] er Croioy -emerged as the metal of the acid chloride.neutrali/ing capability _ have necked down, thus preventing the pay;g m
^ i With appropriate fmal formeriv miroduced by the phosphate _ passage of internal-inspection probes, it logy and nd XI gr '.) heat choice for tub treatment ng. or higher), a additive, some PWR operators began to, is described as tube " denting"(see box, (to ll50F f a fine-grained structure is achieved to look at the need for mventing_ so(ids _ p S 9).'
iatt any This is the primary heat-exchange pr Q )8, provide the needed measure of resistance transport to the steam generator,. Note that this necessity had long been recog- problem that has preoccupied nuclear M t.o stress-corrosion cracking. Finally, ater chemistry has to be carefully nized for boiling water systems, because designers over the past four years.
.e 4 controlled to preclude corrosion and the direct cycle made the reactor core Extensive research has been done on the an indi. j deposition. As in lower pressure fossil- caternal highly part susceptibic of the circuit. toThis deposition is not vendors fromand thebydenting the Steam phenomenon, Generator b tenerator d 'I red units. the key feature in this treat-rnent approach was the addition of sodi- true of PWRs, however; the steam gener. Owners Group-a group ~ formed in 1977 -rom the am phosphate as a butter, to precipitate ator provides protection for the reactor. under the guidance of the Electric Power
- aining a
'n, purities entering through system lenk- But now the need for condensate treat. Rescarch Institute. The following cle-pren ure- ..
ment began to get mme a,ceptance in ments have been identihed to date: ter hue 2 4 # P"en a. e Denune is the result of acid-chlo-
,olved in j in thh program, maintaimng contnit PWR quaners. t was l'reat ment focuses on heavy metal nde cornnmn of ihe carbon-stc'el supgr,t._
ppb b 1 of the mihum-to-pinwphate rano plates, at the points of, intersection with_ ( cs,ential to preclude the occurrence of enmsmunds pron. corner) and diwolved and b salb (valium. uhe a l Im nal melhnds tub nr m wl $ free ca min; m the set on.ta r y water a l he cornr.nm pn=luct formed is leatmed pohslung hHcts Ine ecmoval of that is J. .llowever, with inleak. ige ot mme m ,o mamin.mm.centn.ionhn ,anefresh- so,reedea ma.d e,. des 1 hese w en- magm nie ite.o.i. .md hi,ih copper and
,, w a , ,,'
deust pH .j becomes dillicult, and Inconel cracking gracrally spiral-wound tubes, precoated wit h a cellulmie fdser aid. Though cornwinn.chlorides are present within th can occur. In athlitmn. thermal concen- a The cornamn pnaduct capands into 1,s nce-of, tranon of phosphate in heated crevices adequate for normal conditions, they did I can produce a more acidic solution, lead- not protest a g.u mt diunived whds the annuhr space (drilled holes) around ir ,nore enhnny via system leaks in condens- the tubes, occupying about twice the naterials I ing to tube. wall ihmning. volume of the material that it replaces. th rigor- Phmphate ' chemistry generally lost ers, primarily. favor at t his point - although some For this purpose, two types of ion- e In expanding, it exerts enough iring op. l 1
- 13. Pressurized-water reactor (PWR) has indirect steam cycle. Steam generator is the steel 12. Pressure vessede (PWR sh n here) critical component, sets rigorous demands on materials and water-chemistry controlin I
to l use low-alloy steets, require in-service
' fnorutonng to ensure life-long integrity entire secondary loop. to minimize corrosion that leads to extensive tube fauure
.ts y.
.; q.. . _W~ait.t_nic_krcss.
. . , in
%ro,sture-secaratort retseater
~ 8
--,.,,,- Steam j l i
J00 tons i .w) l [** l g' ' Pressuraer I I Generator To l . .
. . . .J "eneraro, d, g
- --~q q )l Turtune LJ'"1.,
g s 400 :orte
}W .NCan1 ' L onQenser r r, h ;
j s i ....c,""' - i 0 *" e u.a.ent %*ater
~
- w "'
n, l
-f o. ..u, , .
, ' H'#~t' iQ ,.n...,, ,
i i 1 lI f, , s.,,, .,' y ( 'i A m..o ,,
; ,r /
o,a.n,
.,o..,..
=-=:
i tli x
./ i q%.
1
- 5. nextor 4 l
_. ,o w,,, q iai v , -- j a
.uaen,a,e mo tone
% ctore x o nt
- . - .a Qua n g s ,
! na
.. s - , - . . . _ . - - ~
..%,.-- 2.~ .e _-
. M% mal 5MN5[-_^Q - MI# hM hMNE O h_ -
L i b * - Nic.un ouuur. 15). whic h were meriuluccil for fmnil. h nitrogen
' slad v One n (rackmg m the open rd r
/ hred systems m the l' nth t he nuclear i length at the upper tube tsundle tep any case.
, gy , , , , , version Icatures a straiyht.shclband-quhe apparently due to high-cpit fatigus l' l w wo u, d88 8 8.9 9..
rg. nan.u design. The advantage heir is the he.u. combination with s orrosuin resula4 1 , C0"Ld 8""' D tramfer benelit of operaimg the scosmi- trom p.mu chemntry contro! Anatiisi ., and s.db N@ to a f/ paser sici arv (shcIll side temperalme ami prewure miergranular corrouon m sescral tubo Ifdle H5 8 q j ch$se to the averaFe conditu.ns on the at the upper tubesheet, aw.uated will IlL bf
- d"I-l M __L -
l primary (tuhe) ude. limes t hus suppl > the entrv ol' sulf ur Probable tatoe u *dlCf P' gj p _. r b sicam with 15-50 degrees superheat at madequate control of leedwater purdy designs full load. capabihi g".. g! f u . rianut.o" b d A hi of . - Secondaruhuip materials are of pngs the comi h me concern in \> stems with once-thro @
"l ,
,7' primary structural consideration was team generators. llor etample. stain;, , e ndem.
ma terials compatibihty with regard to .tcel is the requirement for tubint 8 U"d
}l i - , -
thermal expansion-not a consideration low-pressure feedwater heaters (see Group - in U tube geometry. Since economy p S 12) This minimites introduction or studies k O l , 7,o, :,,ygg,,3 dictates a carbon-steel shell. carbon-steel iron. In no ease is coaper alhn alloweJis Problem j l ^, tubing would be preferred if it were not the feedwater tram. On the other hani - ff ' - Tuoe cunoie for the boric acid environment of the stainless is recommended for moisture- [ primary loop. Surprisingly, the coetti- separator /rcheater tubing. but both car- [ ,,,, a ~ cient of expansion of nickel bearing allo) bon steel and copper alloy 140/10 with provides a better match than stainless nickell are opimns with delferent advatb f [ " ".."."l{y }
. l l
t v.11st . . e ' and so Inconel n ihe choke. lions the st.ut, the hidiauhs s aint i.i res relative lo pil kno:ainins and t t
.4 kA m.
's
.un -r water thenmary were asugned cotwal
\'
+ '
lus.c .hver roles in assurmg reliable OTSG perform- ' N[ ance.* Considering U tube denting proh-neati..,coou ,,-uct CORT tgY Li i a.o,ao, .",e
,,,n ,, ne ,,, pg ,,, s my n,,,. co,,, n n , . the i ed greatly m ihn respect. . lust .n m the L t.u..o. no . .uurt u g. a nes of %nsumal '.'P Steam
- 14. U-tube. tecirc-type steam generator is predom6nant. Tube thinning, enrrosion plates with drilled holes provide suppirt
.ilong the length of the tubes. s gggg W
dC h' '8 "
sta locus of industry research program Recognifiny the need to mimunie the g[.)' ,[ ] % to ret. annular contact area where meoming The conlammants can collect and initiate eg ,,offie f' f
- f#
**"' 0 compressive force to deform the tube and corrosive action, each hole n broached at E' . l stress
[ the support plate ligaments between the three point s, w hich are spaced 120 MJ L I > nuctua tube hole and water-circulation holes degrees apart. They provide relatively rI U 7 others c One side-etTect is a narrowing of the large openings for flow of* steam and WM ; ; later. center portion of Ilow slots hAated along water around and along the tubes. avoid. the center line of support plates which ing sohd5 concentrating mechanisms that 3 , ,, s . , m., o , suppo ouyy, A N i
' M O N' creates an "hourglassing" effect.
i n Another is the nblongation af mme narrow crevices (Fig 16). tubes at the U bend regionaccompartied ean result from steam blanketmg m , noe.* l c" %, [/e i suppo and ai The importance of feedwater chemis- ! "l 4 wir (C). w by cracking and leakage. Tube denting has occurred within as re' circulation in the steam generator to try stems from the fact that there is no ((
/* ,
Decer of all little as four months after conversion accumulate solids. As a result. most i A k N5 from phosphate to volatile treatment. solids entering the OTSG and dissolved 5 j [ us, ,,. One.
"O*'
although prior use of phosphate has not in the superheated steam are passed g ; ,.,cu.e always been a prcrcquisite. To enable along wit h generalud steam to the m ; m ens desig continued operation. vendors are work- t u r bine. Jmt as m the ItWR thn '*"""" A " "' l ing with uscrs of their steam systenn to precludes athlision of anvihing but vola-
'd l ', , ar mu-identify and elimmate pomis of mica- ulinne chemicah to the feedwater to O On Lage of contaminants. particularly chlo-l
{" l , ( prevent secondary system corrosion. 7 b ,, l'Y - rides and onygen. To minimife further An all volaule t rea t ment has been dentmg n canimg plants. t hey recom. l m.,arne.t { / L *' prewnbed for O ISG oper.itmn from the ' et d" f as- i "'*" mend ught adhcrence to AVI procc- st.ut. Il features ammoma and hydraime '""'"" l * { dure.s together with ngoroin control of for pil and osygen control, respecinely. condensate and leedwater chemistry plus full.ilow condema te poinhmg to I h -; -
.,y, y ,
'*S" during both operanon and shutdown - to maintain feedwater purity. Ty pical feed. U mmimite corrmion in the secondary water linnes for normal operation are 20 " ' 'd'" "C ' 'h ,- - p l system and transport of corrouan prod- ppb ulica 10 ppb iron. 2 pph copper. I ' - 7 '
# \i Di ,psiwr '""'
( ucts to the team generator. l%omic td pph lead. 50 ppb intal soh ls.' gggq[ OD* Urmjikt lowdown b _wateri.100-150 ppb With 15 tubes out of 270.000 show m g N f ano chlorides, pH of 15-9 0, and 1. ppm ~ iodi leakai;c over a span of 30 reactor years '** suspended solids. of operation, perf ormance is considered "".u rui / '"" I ~~ ! Once-through generators good. Tube unrfaces appear umformiv '"d""' - Sh* l """" 4 clean, except for iron oude accumull- ' ' ) About a fourth of all the operatmg tion on the lower tubesheet of one umt. 15. Once-through, rtr'g%-fre dMon u nh s .m n i .. L
I nitrogen blanketing during startups. In groupingv (1) w a t et .c hemist ry pro- e t'hangine tube supports to stainicss
- open tube die region, any cec, the high water solubility and grams to keep out corrosive chemicals or a0% or Jf W to redu_ce_ corrosion.
fati;:ue in be satur,sted4 team solubility of some mimmire their impatt (M nientification a Full depth rolling plus hydraulic resulting o,nta mina nt s present a problem: CO, of alternative comtructum materiah few cipanuun of tubes into tubcsheets to Another is .nd salts such as Nacl tend to concen- suweptible io corrou.in; i 1) improved einmnate crevices at pints. crai iubes' .ite in the draint With drains pumped thermal' hydraulics and t 1) methods for a t 'se of \llov 690 for tubing: this _ iated with istmard, as the system calls for, feed early diarniwn of problems Out of the e adds enrrosum resistany 01., chromiurn sause n .ater purity is thus degraded. llence, mwntigations and others by individual 00'* 1 in aqnMn..thmi.ng_highqcmpera,- r purity. Jeugns for all plants mcurporate the N5S and metah supphers have come ture water, to rc.sistance of nickel (6M) a of prime , apabihty to allow drams discharge to several promising developments. Some of tif chlornic stress-es,rrouon cracking., e-t hroug h the conder.>cr hotwell for cleanup by the the more interesting'" are: a Chemical cleamng of dentec steam
, uamless condensate p. hsher. e llydrogen monitoring as an inthea- generators in a threc stcp process:
tubing in 1:nder EPRI manatement. the Ow ners tor of corrmion attak m progress. removal of ludp from corroded region, ters (see 6roup is sponsormg a wide variety of a Rc of boric acid or small concen. sohent-cicining of crevice; removal of intion of a udies related to the steam generator t ra t ium of calcium or phosphate to copper deposits and passivation of allowed in problema These include four broad mhibit corrosion. surface. her hand, e rnont u re. twth car. .mc.un vuoter l e/IO with . l m advan. Se m iu,e # ) C \ ums and
>< parator . . gM, .i .
- . j 0 Corrosion and retubing: Profiling m,pe, sin.,o -
c.W .
- t. ~t -
C the experience of one utility : { 3.);g
* "" $leam-generator prot)lems at nuclear plants rettect the complex .. < M interaction of mechanical design, materials, and water chamistry o !
the esperiencn at onn plant typittee ttm liantory, nnd nrm nppeonch (- l - ic restore the plant to full capacity '% ' f A
$ (
The Surry nucleur station used sodlum-phosphntH water tennt- p ; ] [ j inent Int its first two years of operation When tube thinning urid g , ,, , ., ,y .n '#. L I ? I stress coitosion cracking made their appont arico at suver al i j i ryctear plants with U-tube steam generators. Surry and most , .i (o g ners converted to an all-volatite program. Four to 14 months f utse l uniste v f r cifwater later. tube dentmg occurred at Surry's units (A). mostly at '"'C' wgport plate regions having fewer coolant passages Adr1stional gg' 0 Go og g3
. ys problems followed: Compressive stresses narrowed some .
ppport p' ate fiow slots, producing an "hourglassmg" effect (DI. I "C '*# N T'
.) n@
, ynd an amial Cracts developed at the apen of an inner tube U-bend
!,Q With more than 20% of the tubes out of service as of tast
.h[-l N dtcmisant o
o* Jrer
* *f j December, management decided to replace the tube assemblies CN' s"*'N' '"###
l l 8 all sin steam generators with new units. l 8 peplacement units encompass several design modifications. 3 f,I-,y I '
. ye involves ther addition of a flow-distribution baffle to direct y .f 7 1 ,',,
en ? .i *inu t.u ,
,, , gee laterally across the tubesheet. Two others are "quatrefoit" p 1 i .u,,,9 ' 2esign for sucport plate tube holes, and addiiional lancing ports. , 5 l
' amrer j u three aim at mmimmng accumulation of sludge at tubesheet Q.r :: .,
t enuti and f acilitating its removal. g,?
"""-d pg .
- Other changes-indicative of those being made m the indus- ..l$ Ag,+)-- ' 'm Stief.- q 1,y_,nvolve materials and fabrication techniques: ; M S. :
1 e_Chanmnc succon rime" and parties from carbon to territic, , j if iF; W 'i l I i pniess stees (Type 405), for greater enrrnen reontance. / 'i. l e Full-depth espansion of tubes in tubeSheets. g .,
- ' b- . i M
e Heat trnatment, to improve grain mictr)structurtt an i remove . l i .' l
'" i ,
i ", , ' , 4 r
', esiduni strasses The retubing procese (foHowmg preparatory work) starts with '
,.i. -
i aittioq the steam-generator sheit at the transition cone ( t). and I
, .c ' l aimovenQ i'm upper assembly Neal. altos rutting ni manwny ! - . r I enmgs m the lower shell, the tube buintio is cut honzontally lust I e '";
( 2euve the tubesheet (2). st tne same, lirna. another machma hj[.
' .emres tube wesos from the other side of tne sheet Fmotty. the jag p"' ,
. v.'
. riire tut e bundle is raised into a cash positioned afnp tho cut Q"nl TQ1 $*'4 .
! eit (3). and tube stubs are removed hydraulically Preparations ,o . .i g ,
' o# r mstalling the new bundle involve reconditionmq the sheet anr1 *[$'mM9 W
l ,,, j :norougn decontamination ,i l l 5.,, j total time reauired? wnii. .i demo ionii 62 days. actuai condi- t ; cons-inciuding radiation eiposure-w it pronaney add to inis. ._. : (
- u m
.o; .
.. ..r:'ysG2
- w . . .
~..: ;;.
.s ...~ >xm,,,gg's.,. '.j v :wg s > , - nj T.
I'l .rf'.
, _, j fQi '; _
_< djif
=
'2 -
J +f-s.e . , ' . ; . .
> . T - - . . . .
NUREG-09093![ ~
.]
9 NRC Report on the January 25,1982 Steam Generator Tube Rupture at. R."E._Ginna Nuclear Power Plant U.S. Nuclear Regulatory Commission h & I i .? g - i : l s \...../ - pe l p l C ( . y .
<- y 1
2 t- _ _ _ _ _. - ._ _ _. 7.3.9 Tube Rupture Mechanism The licensee has not reached a conclusion as to the exact failure mechanism for the ruptured tube or the sequence of events that led to the tube rupture. The licensee and Westinghouse Electric Corporation are continuing to investigate flow-induced mechanical tube loading, mechanical wear, impact damage and combinations of these phenomena. The licensee removed the ruptured tube from the unit for metallurgical evaluation and reported the results of a cursory visual examination to the Task Force. The ruptured tube (R42C55) showed what appeared to be a fresh, axial - band of fretting wear with transverse striations. There were also indications of previously oxidized wear markings. The total length of'the rupture was 4.1 in., with the wide mouth of the rupture measuring 0.7 in. The tube locally ballopned to a diameter of 1.15 in. at.the rupture location. The wall thickness at the rupture point was found to be reduced to approximately 5% of the original wall thickness with the rupture stopping at a point where the wall thickness was estimated to be approximately 50% of the original wall thickness. The rupture exhibited Considerable ductility. The appearance of the failure was similar to that shown in Figure 40~of huREG/CR-0718, Steam Generator Tube Integr'ity Program - Phase I Report, which shows burst f ailure mode of elliptical wastage (85 to 90% deep). The licensee removed 9-in.-long segments of the ruptured tube and five other tubes in the No. 4 we'dge area for metallurgical examination at the Westinghouse Research and Development Center hot cell laboratory. The tubes removed for examination by Westinhouse were tubes R42C56, R44C55, R43C55, R42C55 (ruptured tube), R44C54 and R43C54, and a segment of an unidentified loose tube. Visual examination by a member of tt;e Task Force indicated the following: All tubes exhibited axial outer surface markings similar in appearance to classical fretting wear with transverse scoring caused by surface asperities.
permission for Ginna's return to power. The staff's review will include i evaluations of the cause of the steam generator tube failure, the extent of , damage to the B steam generator, repairs and modifications performed, and _ I
' future action and inspections that the licensee has proposed. Additionally, as '
reported in SECY-82-72A, a proposed management structure has been developed for an NRC/ Industry coordinated program to address steam generator degradation. This program was established to develop any necessary short term fixes and long term research activities that could promote resolution of the steam generator degradation problem. 7.4 Other Equipment Problems During the Event - 7.4.1 Pressurizer Power-0perated Relief Valve The two pressurizer power-operated relief valves (PORV) are Copes-Vulcan, 3-in., Model 1513, valves with Model D-100-160 reverse-acting pneumatic operators. Figure 7.18 shows a cutaway view of a PORV. Both PORVsi ad been replaced in 1980 during a modification outage to provide positive valve indication in accordance with TMI Action Plan requirements. Prior to this time, the PORVs had been part of the originally installed plant equipment. The licensee stated that valves similar to both PORVs were tested under the program implemented for the licensee by the Electric Power Research Institute (EPRI). This program consisted of full-scale operating tests on prototypical valves and was' completed in December 1981. During the course of this test program, a Velan, Model B10-3054B-13MS, block valve was also tested. This Velan valve was of the same model as those installed at Ginna; however, the valve motor operators were slightly different. All test results were judged to be statisfactory by EPRI and the licensee. In 1979, the reactor coolant system low-temperature overpressure protection system was installed. This system uses the PORVs to prevent overpressurization when the plant is shutdcwn and at a low temperature with the pressurizer completely filled with water.
SYSTEM CHEMISTRY CONSIDERATIONS .%-ire FOR NUCLEAR STEAM GENERATORS FREDERICK 1. POCOCK Babcock & Wilcox Company Alliance Research Center, Research and Development Division P.O. Box 833, Alliance, Ohio 44601 Received October 20,1980 Accepted for Publication February 27,1981 4 sihC, d2~ZZi3EEZjE[CC3] industry in the U.S Prior to the introduction of once-through fossile supercritical steam generators in the The evolution of water trestment control has early 1950s, there were two principal philosophies'of been in response to the effect of water contaminants internal boiler water treatment. One theory for high-on observed damages in steam generators. pressure fossil boiler internal treatment included the The accumulation of deposits from condenser so-called " free caustic theory." This was essentially leakage constituents in combination with alkaline operation at a pH of ~10 to l1 (reference temper- - boiler water additives has caused corrosion in recir- ature, 2S*C) with a few ppm of residual phosphate culating fossil boilers. This is mitigated by the use of (usually trisodium phosphate) to precipitate any phosphate treatment only at controlled pH condi- water hardness constituents inadvertently introduced tions. The same fossil water technology is applied to with the feedwater. Two principal corrosion problems nuclear boilers with adverse results, especially in units arose from this treatment: " caustic gouging" and tubed with Alloy 600. "caus' tic embrittlement." The advent of once-through fossil steam gener- To combat these corrosion problems, a second alors led to the use of very pure water, since anything concept, called the coordinated pH-phosphate control not soluble in the steam was available to concentrate method, was developed in the early 1940s (Refs. 3,4, and deposit in the boiler to enhance corrosion and and 5). Essentially, this method assured that there heat transfer problems. This fact necessitated the was sufncient phosphate present to prevent the introduction of condensate and feedwater polishing formation of " free caustic" (sodium hydroxide) on by filtration and ion exchange. When corrosion concentration of the boiler water chemicals to t' ear problems were encountered in nuclear steam gener- dryness in crevices or "under deposits," thus reducing ators due to phosphate chemicals in combination the probability of under-deposit " gouging" corrosion with can' denser leakage constituents pure water attack and/or caustic embrittlement in crevice loca-treatment philosophy was adopted, and with it tions.*J condensate polishing came more widely into use in The purpose of the relative high pH used in both these units. Pure water with condensate polishing was cases was, of course to reduce the probability of c/ ways applied to once-through nuclear steam gener- boiler steel corrosion by improving the conditions ators. for the formation and constant repair of a tight Since no water treatment methods are 100% suc- protective magnetite film on the boiler steel. cessfulin controlling steam generator deposition and As pressures and temperatures increased. the associated corrosio n and thermal-hydraulic problems. utilization of the high pH-low phosphate treatment chemical cleaning has been adopted as a maintenance procedure was identified as a causative factor in some procedure, first in fossil units and now in nuclear corrosion problems, and the coordinated pH-phos-w its. phate theory became more widely applied. Variations of the latter were suggested and applied, the pnneipal one being congruent pH-phosphate control that employed a mixture of a tri-basic and di-basic INTROD UCTIO N phosphate and ensured the absence of free caustic under boiler water concentrating conditions in "high-Water treatment and water treatment control quality" and/or crevice regions. It has been found have been an evolutionar'y procen in the power that reactions of water treatment chemicals with
' ^
Pocock CllLMISTRY CONSIDERATIONS FOR STEAM GENERATORS indicate a high quality (steam bubble) in the region ALLOY 600 TUBED STEAM GENERATORS adjacent to the inlet tubesheet due to insufficient steam nser capaesty in that location.is Water chemistry requirements for the initial Corrective steps were taken to ensure proper Alloy 600 tubed steam generators resulted from the maintenance of the pH-phosphate relationship in the experience gained with stainless steel tubed nuclear bulk boiler water, and another riser was added at the units and also incorporated that of the fossil boiler tubesheet to eliminate the "high-quality" region. The industry. The initial choice for chemistry control for steam generator had no further difGculties from tube these units in the U.S. was congruent pH-phosphate damage at that location in subsequent operation. control. The initial choice in Germany was all-Destructive examination of the inlet tube tube- volatile treatment (AVT), although subsequent Alloy sheet area when the steam generators were replaced 800 tubed steam generators in that country used a several years later showed many cracks in tubing in form of phosphate treatment at low concentrations the tube-tubesheet crevices. The time of occurrence in the boiler water.2o.2: of this damage is unknown. but it is surmised that it As operational time increased, several types of may ha', c occurred during the initial operating failures occurred under these operating conditions. problem. It is possible that the cracked tubes did not They' included transgranular stress corrosion, inter-cause operating problems because of pluggage of granular stress corrosion. tube thinning and wastage tube-tubcsheet crevice area with corrosion product (phosphate treated units), along with the more debris. The tubes were fully rolled through the tube- mechanical-design-related problems of fretting, wear. sheet. It is beliesed that the seal weld helium test and fatigue. The clearly chemical-related problems of leak path groove in the tube-tubesheet crevice may stress corrosion and tube wastage were again related have contributed to boiler water chemicals reaching to localized chemical concentrating areas in crevices the weld area, conecntrating due to the locally high (tube-tube support plate, tube-tubesheet) and under heat transfer conditions and eventually leading to the porous deposits. Many variations of congruent pH . observed tube damage by stress corrosion cracking phosphate control were attempted to maintain pH in the tube-tubesheet crevice." within the narrow range needed to prevent low-pH The operator's approach at Indian Point I was to (acid-phosphate) tube wastage or high-pH (coustic) utilize high-purity water, volatile amine treatment, stress corrosion problems. While bulk water condi-and depend on very comprehensive monitoring of the tions were carefully monitored, theremas no way to condensate and feedwater by automatic analysis absolutely control local (crevice) conditions. equipment to observe and control condenser leakage Finally, a decision was made to completely and t he re by feedwater contaminants. Automatic change the chemistry to AVT. similar to the initial flame photometers with associated on-line sampling German unit and without solid phosphate chemical devices were used to momtor sodium in the ppb range additives, to reduce as much as possible the total to accomplish this purpose. There were also in-line solids additives available for concentration and re-colorimetnc analyses. No internal phosphate treat- ution under crevice and/or accumulated porous ment was used. Over several years of operation, tube deposit conditions. failures did, however. occur, some of which were A new and more severe type of corrosion damage attributed to stress corrosion." now occurred in these steam generators and was A similarly designed power system on the NSS commonly called " tube denting." The tube dents Sarannah included a water treatment scheme in- resulted from concentrations of chloride containing volving a form af coordinated pH-phosphate and feedwater contaminants under heat transf'r condi-comprehensive control. The water treatment system tions in the tube-tube support plate crevices and the utilized high concentrations of mixed tri- and diso- linear oxide growth rate of carbon steel support plate dium phosphate along with sodium sulGte for oxygen interface in the crevice in the unbuffered bulk water scavenging. The feedwater and condensate system was environment. Laboratory studies conGrmed that a treated with morpholine and also included a hypass low-pH-chloride environment was involved and that dcmineralizer. onset usually occurred after some formation of A very sophisticated and automated water quality porous deposits in the crevice. There are controversies momtonn; system was installed along with an inno- about catalysis effects. but iron. copper, and nickel vatne approach to closely monitor hot-standby and have been mentioned as contributing to acceleration off-line storage control of the steam generators. The of this corrosion problem. It is known to hase power system was operated all during the 1960s both occurred when only 1 ppm of chloride was in the in demonstration cruises and in commercial service. bu k boiler water. Under such hulk water conditions. without a single incident of tube leakage. Condenser local concent, rations as high as 10 000 ppm can occur leakage was fairly common and was esten3ise enough through concentratmn in the crevice. This usually to necessitate pluggage of part of the mam condenser occurs at the metal-oside growth interface. and retubing one complete auNiliary condenscrM Currently. AVT treatment is recommended for
Pocock CHEMISTRY CONSIDERATIONS FOR 5 TEAM GENERATORS Copper removal from drum boilers in instanen CONCLUSIONS where concentrations in deposits were greater than
~10% was by a separate cleaning stage, usually There is still much controversy concerning the employing an oxidizing chemical (commonly bro- proper design of steam generators, the balance-of-mates and/or air) and ammonium hydroxide. At low plant, and the system chemistry. There is, however, a copper concentrations in deposits (~5%), a se- growing recognition of the need to avoid insofar as questering agent (thiourca) was sometimes added to possible material selections that would lead to other the solvent to accomplish copper removal in combina- than the best materials-water chemistry environment tion with corrosion product and feedwater con- and ease of chemical cleaning maintenance. Cost taminant deposit removal.
effectiveness has led to the choice of continual ex-Our view is that maintenance chemical cleaning of pansion of plant size and thereby increased the nuclear steam generators is also required. Some impact of extended downtime on system reliability special considerations are involved in cleaning these from corrosion-related equipment failures or other multiple tube heat exchangers. They are unlike all- problems for that matter." welded fossil boilers with tube side boiling and with, With nuclear plants, these have been especially for the most part, no crevices. They have shell-side costly, as they provided cheaper fuel costs and are of boiling and numerous crevices where localized cor- consistently large size [~1000 MW(electric)] as com-rosion might occur during or as a result of the pared with ever-escalating petroleum and coal prices cleaning process. It is also relatively impossible to for fossil equipment. thoroughly flush looser.ed corrosion products from it has become very important'to examine the large complex tube and shell structures." interrelationships of system component designs and Chemical cleaning solvents investigated for use in materials choices as they relate to the choice of sys-nuclear steam generators, therefore, needed to be tem chemistry control and to carefully evaluate any organic in nature and thermally degradable, without compromises these choices might incur. With very resulting in undesirable thermal degradation products. large system bulk flows invcived. it is important to They also needed to be capable of complete or nearly reduce individual component general corrosion to the complete solutioning of deposits for dependable very minimum and to then prevent this minimum removal from the steam generator. level of corrosion products from being transported to The initial choices were the organic materials the steam generator. Even the most perfectly de-already well developed for fossil steam generator signed and operated system cannot prevent the cleaning. Early work in this field showed that EDTA eventual accumulation of feedwater contaminants in at neutral pH with a reducing agent (hydrazine) was sensitive areas in the steam generators where locali,z_ed effective. Several other cleaning solvent formulations corrosive damage can occur. One part per billion ofl have been developed, and intensive work is still going iron (u Fe)in the feedwater can result in 40 to 50 lb/ ' on. The effect of accumulated sludge (corrosion yr of accumulated corrosion products in a nuclear products) deposits has been well established as a steam generator in a 1000-MW(electric) plant. - precursor to severe localized corrosion damage in Therefore, in addition to optimizing system nuclear steam generators. materials choice and th" related ehemistry control. The addition of hydrazine improved the mag- as well as the utilizat.on of the best feedwater netite solutioning capability of the best solvent found purification techniques, it is necessary to develop and
- in our research (EDTA). The reason postulated was utilize safe chemical cleaning solvents as a periodic the reduction of trivalent iron to divalent iron for maintenance procedure to control the accumulations l
better complexing with the EDTA. Other, reducing of pre-boiler corrosion products in the steam gener-agents such as ascorbic acid accomplished a similar ator to reduce the incidence of deposit-induced effect, but the ones investigated were usually found crevice corrosion. to be less attractive than hydrazine. Neutralization of the solvent was found to reduce corrosive attack of carbon steel materials in high velocity (flow) are:s of REFERENCES the steam generators (principally support plates) when filling and draining and/or arculating the 1. W. C. SCllROEDER and A. A. UERK. "Intererptalline solven t. Cracking of Boder Steel and its Prevennon." Bull. JJ3. Bureau Copper solvents were not part of our early com- '*"'50 W mercial investigations, since a prior decision had been made to recommend against the use of copper alloys 2. A. A. BERK and W. F. WALDEK " Caustic Danger Zone " Orem. Eng. , pp. 235. 236,238 Oune 1950L in B&W nuclear steam supply systems downstream of the condensate polisher in these full flow condensate
- 3. S. F. WillRL and T. E. PURCELL. "Protecuan Amnst polished systems. They are, however, a significant Caustic Einhntilement by Coordinated thphate.pH Con.
part of other and still on-going research. trol." Proc. T/ned . innual li'arer Gm/ . Engmeers' %ciety of 122 NtrLLW TifliNOLOGY VO L. .i 5 OCT.1981
i l -. _ . s Nucl. Enersy.1979. Vol.18, Apr No. 2, !17-132
.I 1
Some ,icstgns of presstiri:ed water reactor (PWR) steam generators Itare experienced a variety of corrosion problems wh include stress corrusion cracking, tube thinning, pitting, fattette, erosion-cnrrosion and support plate corrosion re i l s entmg'. Large international research programmes have been mounted to investigate the phenomena. The operational e penence is reviewed and mechamsms witich have been proposed to explain the corrosion damage are presented. The ' for design development and for boiler and feedwater control are discussed. I Corrosion of PWR steam generators ; R, GarnSey, BSc. PhD. CChem, MRIC* Of the 147 (83 902 MW(e)) nuclear stations of 150 MW(e) development (R&D) programme which is integrated with and over commissioned by the middle of 1978.60(43 363 that of the Electric Power Research Institute.2 France, - MW(e)) were PWRs.8 in the USA alone.46 utilities plan to Sweden and Japan are members of the~ Owners Group,and have a total of 125 PWR units in operation by 1986. PWR there are strong links with Germany, Italy and Canada. problems of any kind are therefore of major interest to the Japan has a large national programme.) The major vendors electricity supply industry. Some designs of PWR plant have their own programmes with international links and have encountered a variety of steam generator problems also have joint programmes with the Owners Group. including contamination, vibration, fretting, waterhammer, Clearly, a utility with a plant which has experienced cracking, wastage, pitting, denting, high cycle fatigue and cracking, thinning and denting has a different perspective erosion-corrosion. There is concern that ptesent practices will not control deterioration sufficiently to ensure a 40 on the problems, and on what is required of R&D pro-year steam generator lifetime without replacement. At least grammes, from one contemplating buying a PWR plant for the first time. tiowever, whether the objective is ameliora-two utilities have decided to replace four umts which will have less than te:t years of commercial service, tion, containment or prevention of damage on existing stations, or to guarantee freedom from such damage ir future Not all the problems listed are encountered on one systems, there is a requirement for ufundamental under-design of plant and for any one design the severity of a particular problem can range from very severe to undetect. standing of the deterioration processes and their sensitivity to design, and operational and water chemistry variables. In able. Secondary circuit corrosion damage, as on any steam some areas there is a broad agreement on causes and implica-generator, is a complex function of boiler design. operating tions though many details and some hypotheses,particularly conditions, boiler water chemistry.and the extent ofingress of air and cooling water contaminants. those which have implications for feed train design, are still under discussion. At least two major changes have been made by Westing-house to the boiler water chemistry specifications in response Design aspects relevant to corrosion to corrosion problems. liigh level alkaline sodium phosphate treatment was changed because of caustic stress corrosion The design details of the steam generator such as tempera-ture, heat Oux, Guid now, methods of fabrication, tube cracking to sodium phosphate with the Na to PO. ratio controlled between 2.8 and 2.0 and then changed again to supports and materials, profoundly innuence vulnerability to corrosion attack. A comprehensive discussion of the
'a:1 volatile alkali'(AVT) because of tube thmning. Denting differences in design is not appropriate but the following oaurred on some stations within 3-6 months of enangm8 to AVT chemistry. The detailed water chemistry specifica. basic differences are relevant to a discussion of corrosion.
There are two basic commercial PWR designs in operation: tions recommended now vary substantially from vendor in .endor and the interpretation of such recommendations a recirculation system which produces dry steam at the saturation temperature (D&S) (a vertical inverted U. tube ur'es from utility to utility. For,cxampic, the Amencan D&S umt is shown in Fig.1), and the once through steam ve .uors recomrnend AVT but Kraftwerk Union (KWU) generator (OTSG) in which all the water entering the advocates low level dosing with disodium hydrogen phosphafe. generator is converted to steam and superheated at the outlet in a smgle pass (see Fig. 2). i The response to these problems and the apparently wide spectrum of operational experience has been the Westinghouse, Combustion Engineering and KWU produce vanants of the inverted U tube design but there are establishment of very large research and development programmes with an impressive degree of collaboration important differences of detail between vendors and between the older and current designs. between governments, vendors and utilities and also between nations. In the USA, a PWR Steam Generator Owners Some Westinghouse and Combustion Engineenng Group has been formed to fund an extensive research and designs have experienced severe corrosion *" which falls into three categones:
- Cen: rat uccincity Research Laborsiones. Leatherhead.
(a) caustic cracking of the tube within the tube sheet i17
~
GARNSEY j ~~~{
/
; \ [ s -
- n t> .
; /.
4ed #
,,'r* ,
\
No. Service No. ', reovired . .- g ,g ,,,in,,
- 1. Pnmary iniet 1 Q ll ,,
2 Pnmary outlet 2 9 'i y. C*"
~
3 Downcomer feeowater 1 l -
*--- 243.75 en 00 y a'
- 4. Steam ovnet 5 Blowdown 2
2 f "~h.* s' e
.H
- 6. Pressure tao 4
O
- 7. t.iovid level 12 s
~
8 Pnmary manway 2 ' ' 9 Secondary manway '~ A @'
- 10. Handhole 2
2 V k ('1 r-
- 11. Main feedwater 2 , q j 3 f ,
82224 in.
', i r
I!
? . #.
I f .
! .' a-189 25 n OD e G 5 %- -
3 : 45' -
, % l'
@T
[lN i .r Q))* [ f A[^x i as i
%Q (:s v d ')
Bottom view /,. - ./ H) , 2 25 d' ) O' *) 00 '4 i Fig. J. A Combustson Engineermt system 80 steam generator or just above the tube sheet in regions submerged in Temperature crud depcaits Secondary circuit temperatures are roughly simdar from
-(b) tube wutage (phosphate thinnm3) in the region just one vendor to another but they have risen from about above the tube sheet 260*C to 289*C as designs have advanced. This increase (c) oxid_e jacking (denting) at tube-tube support inter- could be relevant to phosphate thinning. It is also relevant sections on some plaats which operate with AVT trent- that pressures and hence secondary temperatures ge5. rally'
. m en t,._ . rise when operating at part load. Pnmary circuit tempera.
'Olt stortion di caused by denting has in some cases lead to tures generally vary between 300 and 330*C. Ileat fluxes stress corrosion cracking i of the tubing from the inside of will be higher on the arm of the U-tube where pnmary
,th,e tube. fluid enters 'the hot leg .
With the exception of Obrigheim, the KWU stations have not experienced caustic cracking but thick crud deposits have accumulated on the tube sheet and progressive tube Fluidpow thinnmg has occurred in this region.on some plants. None The point at which feedwater enters the recirculating fluid of these stations has experienced denting. is important because it affects the steam quality of the fluid The Babcock & Wilcox (B & W) OTSG units have not at the bottom. adjacent to the tube sheet. had any of the above problems but have experienced high in the Westinghouse designs, feedwater enters at the top cycle fatigue failures, fretting, tube pitting anJ erosion- by way of a feed ring. In the Combustion Engineering corrosion. ' system 80', shown in Fig.1, feedwater entersjust above the 118
CORROSION OF PWR STEAM GENERATORS
$$,$"""* '" * **" vulnerability to stress corrosion cracking or other rapid
'M N corrosion processes. It would be an indication of too smi'l 1 H "eatment a margin if specification of a specific heat treatment was f _
e 0 5 h al10661121-C/WO essential to ensure freedom from trouble. The important y80 - o 0.5 n at 1010'C/AC design Considerations to have emerged from the researen.qn. ' c
~
2h 49-6 7 A .Y_ ' maYeds~ai U j_ahCa, tion, stresses a m_ust be glnimized i and geometries such as crevices, conducive to
) 60 -
,the generation of concentrated acidic or alkaline solutions,.
2 I - should be eliminated where possible. Alternative materials inay oe intr ~oduced when the R & D work demonstrates a h4o . y ; clear increase in the margin of safety.-
~
3 l 4 f 20 -
- o Feed-train design American steam generator vendors are not usually respon-sible for the design of the rest of the secondary circuit.In "g ' g" '* some respects this has been unfortunate because it is clear 0 20 40 60 80 %N' that some of the corrosion problems occurred because the Ahoys 304ss 800 690 600 Na201 necessity of considering the integrity and compatibility of Fig. n. Stress corrosion tests er 316*C using U bend specimens in the steam-water circuit as a whole was not sufficiently de-aerated 10% NaOH; exposure time = 6 weeks (after Sedriks et appreciated.
81 " I The condenser is the major component, which can control the environment, encountered by the steam contamination it is likely to suffer. There is also no reliable . generator,although there are other sources of contamination.
. quantitative relationship .between the compositions of A typical 500 MW(e) condenser might have 300 miles of , solutions generated locally at, say, a tube to tube sheet . , tubing and 50 000 expanded or rolled tube to tube sheet grevice,, the bulk water chemistry and the thermal-hydraulic,, _ joints exposed to an aggressive environment so that parameters. , occasional leaks are inevitable. Detection of salt ingress is The data on caustic stress corrosion illustrates the 'not a particular problem but location of leaks less than difficulty of assessing the relative merits of candidate about 0.51/h can be extremely time consuming, not least materials when the range oflikely environments is bothlarge because they tend to block up with debris when the unit is ,
and uncertain. It can be seen from the international Nickel brought off load. Condenser integrity problems very widely data" in Fig. 6 th'at in 507o and 107, sodium hydroxide the from one plant to another. On.some of the best. plants,, cracking rates decrease withincreasing nickel concentration. thlo_tidt.Jevels in..the steam generator are a few tens of ppb These experiments were performed at the rest potential (p. arts per 10'), on the worst, levels of 100 ppm have been which for such solutions will have been low. If however encountered. experiments are performed at higher potentials the cracking rates of 3I6 and incoloy 800 are much slower and can be less than for Inconel 600. Higher potentials correspond to De aerated caustic soda solution 350 C
008 for more dilute caustic solutions (see Fig. 7) and caustic 3MO solutions containing sulphate ion or copper oxide.s' These latter solutions are likely to be more realistic simulations of ,
the environments generated within riserwater cooled units $ Stainless steel Anoy 800 with copper alloys in the feed train. e T 316 KWU steam generators use incoloy 800 tubing. The y matenal most widely used is inconel 600, although the heat I2000 treatment conditions vary from vendor to vendor. Early o Westinghouse and all Combustion Engineering plants use ; mill. annealed material. B & W heat. treat the complete tube S bundle assembly at between 594*C and 612*C" to stress l reheve the tube to tube sheet joints which sensitizes the j inconel material. Westinghouse now recommend a thermal j1000 h Ahoy treatment, where the tubing is treated at the tube mill for i 690 12 h at 700*C in a vacuum furnace.The small radii U bends 2 D/f p
/ g-';f are stress relieved at 700*C for a further two hours after ,
bending. This treatment minimizes residual stresses and 4 improves metallurgical structure. A semi continuous carbide l ., a precipitate is produced at the gram boundary without an t 4 10 40 100 500 associated chromium depleted layer. This improves resis- WI tance to stress corrosion cracking in dilute environments. Fie. 7 Cornoarsson of the resistance to stress corrosion cracAing of From the steam generator usets point of view. there is a llov 600, ollbv 800and TIUl6 staunicss steetin causne solunons ond requirement for a considerabic margin of safety from effect of causne soda concentration (after Bette and Donan** ) 121 9 -_
CARNSI Y Table 4. KWU specification of the steam generator water and of the feedwater
/*I s
j Steam generator Feedwater f pil at 25'C 8.8 - 9.6 9.0-9.5 [ Conductivary : uS/cm <50 < 0.2
- O, mg/kg <0.005 .
N,il. mg/kg <0.01
. . . . . PO. mg/kg 2.0 - 6.0
,, Mole rauo Na/PO. < 2.6
, f r *. CI mg/kg < l .0 SiO, mg/kg <4.0
! ait specimen esposed l'e mg/kg <0.02 i .
,,,se n. ,,, 3. Cu mg/kg <0.003 l 4M sod 6um hydrea 6de
! * * "8 **d i" * ** * * *
- steels (sensitized in weld heat affected zones) are much j' 4* asse more susceptible to caustic stress corrosion cracking, this 4 .'
difference in operational experience requires some explana-tion. A tubular header design is necessary so that. crud accumulation around the tubes is less likely and the tubes
~
.are bore welded to upstands machined on the header (Foster Wheeler weld) so that there are no crevices. Concentration
'(
~
efficiencies, aitherefore axely to"Te~ili'uTh. iess: the als specimen exposed average secondary side metal temperature i'sals0 IUCely to to, a n.u,e 'a have been 'ess than on commercial units. Another importan_t 4n sodium hyd,e ide difference is that a sulphitt.rather.than hydrazine is used as
=see an. oxygen scaveriger so that submarine boilers invariably Fig. 8. Inhibition of caustic stress corrosion of 316 specimen by operate with, some sulphate present. It has been shown
addition of sodium sulphate -
,that the presence of a sulphate substantially inhibits the caustic striss c6rrosion of 300 series stainless steefs (see when concentrated.) Westinghouse implemented a solution [ Fig. 8).
control range in which the Na-PO. ra tio was within 2.0-2.6. The' reduction in alkalinity resulting from either a There was sufficient feedback from their plants by mid reduction of the Na-PO. ratio or the change to AVT could 1973 to conclude that caustic cracking had been suppressed. explain a reduction in the evidence of caustic stress However, failures attributed to stress corrosion cracking corrosion. However, the continued occurance .ofjailures (SCC) have continued to occur at Beznau I and Point _ attributed to3C~a~t' Beznau l',~and Poin't B_each requires __ Beach. some expjanation because although residual alkalinity may Tube inspections towards the end of 1973 revealed iot_ have been completely eliminated, the caustic con. extensive tube thinning in nearly all plant inspected. In centrations must have been reduced _by_ the proTo'nied~ December 1973 Westinghouse revised the phosphate regime __operati.on witn av i chemistry.In addition to the arguments _ , l recommendation to an Na-PO 4 ratio of 2.3 to 2.6 and concerning_the high concentration efficiencies in these unjtt. l' instituted laboratory and field inspection programmes to it should be noted that they are contaminated with copper evaluate phosphate treatment and alternatives. They estab- oxide and, being riverwater cooled, could also become lished to their satisfaction that phosphate treatment was .contam.inated with sulphate. Both copper oxide and sulphate not controllable for stations with severe crud and condenser leakage problems and probably not viable in the long term . for any PWR steam generator.
/- soi c comoin 1
On 18 August,1974, AVT was recommended to all is . / onosonate Na P p / Westinghouse customers. Previous phosphate chemical con- - trol specifications and the current Westinghouse AVT /M Na2 HPO, specifications are given in Tables 2 and 3, respectively. 12 . gp KWU stations operate a phosphate regime where the -
.W -
l Na-PO. ratio is close to 2 an-d the phosphate concentration is between 2 and 6 ppm. Their specification is giten in 88 ' I f /* ,
/ NE2.1P Table 4. These stations continue to experience phosphate f--
/ /
thinning but with the exception of Obrigheim have not had
/ / n, p caustic cracking problems. Stations since Obrigheim have been tubed in incoloy 800; there are other relevant d ~
/
'/ / ci l
Jifferentes from Westmghouse plant, such as the main. tenance of a high standard of control of coobng water and 0 g/ 2 4 e a oxygen ingress into the feedwater. : gpo,g moi ng l Submanne steam generators operate with a trisodium phosphate regtme. In v1ew of the fact that the 300 senes Fig. 9. Phase d6tgram for aqueous sodium phosphate at 230*C 124 , I
. CORROSION OF PWR STEAM GENERATORS leakage. Rapid failure occurred at Point Beach: however, will depend on the acidity generated locally when salts inspection procedures and water chemistry control were present in the boiler water are concentrated. Phosphate revised as a result of the incident and the more recent residues remaining from previous boiler treatments can en-designs are much less vulnerable to this type of attack.The hance attack resulting from the ingress of seawater but are circumferential failures at Oconee could have opened up not an essential prerequisite for denting. Dere is evidence .
rapidly under accident transients but in fact they were to suggest that copper ingress to the steam o-- detected, and this type of failure mode should not occur on enhances, denting. Cupric oxide ingress has been shown to latter designs of OTSGs which do not have the vacant promote denting in experimental systems but this may inspection lane. Circumferential failures have not been simply be because CuO acts u an oxidizing agent. Air ingress observed on U-tube systems;also, because the tube is always could have a similar effect. subject to an axialtensile load,a detectable leak will develop Goncentration of chloride and sulphate salts alone is not before substantial circumferential propagation has occurred. sufficient to explain denting on riverwater cooled stations _TthThistory w of alkaline boiler water. The presence of in' i Conclusions oxidizing agent, such as excess air, can produce _ acid Severe corrosion problems have been encountered on some , solutions in crevices by anodic dissolution even when the PWR steam generators and large intemational research and b'ulk w'ater is neutral or slightly alkaline. development programmes have been mounted. In all cases ~ ~ Avl'~5oilir water treatments reduce the risk of caustic the corrosion has been associated with some particular cracking but provide no protection against ingress of acid design feature, salts or the development of acid solutions in crevices caused The most vulnerable areas within the units are where by oxygen potential gradients. Rigorous exclusion of salts boiling occurs with restricted Guid recirculation because from cooling water and air are therefore necessary if AVT is acidic or alkaline solutions can be generated. Dryout within employed. sludge deposits and crevices at both tube-tube sheet and The addition of a solid alkali to the boiler reduces tube-support plate intersections are examples. vulnerability to both acid salts and excess oxygen.There is Areas of high residual stress or those which become sub- . also evidence to suggest that boric acid inhibits acid chloride ject to high tensile strains as a consequence of distortions ' attack at least for limited periods of time. , produced by corrosion damage have been shown to be sus-ceptible to stress corrosion. Acknowledgements in some OTSG units, high Guid velocities and cross._09.w_.. The Author is indebted to many colleagues on the Corrosion have induced high cycle fatigue, fretting and ero_sion: Advisory Committee of EPRI. This Paper was written at corrosio.n problems. CERL and is published by permis1 ion of the CEGB. Damage in crevices and beneath sludge piles is critically dependent on boiler water chemistry and the control of References ingress of cooling water salts and air. Comparisons between the various plants demonstrated the importance of treating 1. IlOWLES L.R. Nuclear station achievements. 2nd quarter, the whole secondary water-steam circuit as an integrated 1978. Nuct Entnr. Int. 1978. Vol 23 Sept.. No. 276,56-57,
- 2. MARTEL LJ. et al TPRI steam generator programs". Amencan entity. Pow r Conference, Illinois Institute of Technology. Chicago.
Tube failure by caustic stress corrosion which has been A pr.1977, Vol. 39. 825 -839. experienced on some units initiating in the tube-tube sheet 3. NATIONAL ASSOCIATION OF CORROSION ENGINEERS. crevices and sludge deposits should not occur on newer The first US-Japan joint symposium on corrosion problems in light water reactors. NACE. Fuji institute, Japan. 28 May - designs because of changes made to the method of 2 June,1978. Vol. I.To be published. fabncation, now patterns, and water chemtstry. The
- 4. INTERNATION A L ATOMIC ENERGY. AUTilORITY.
Elimination of copper alloys from areas of the feed train ' Operating experience with nucicar power stations in member which could result in cupric ion transport to the boiler,and states in 1972'. international Atomic Energy Authority, Vienna, the control of sulphate ion contamination, could further 5 . 11 0 r I. e al ' Steam generator reliabihty: the Canadian reduce the risk of caustic SCC in inconel tubed boilers. . approach'. Atomic Energy orCanada Ltd,1974. AECL 4771. Phosphate thinning continues to occur on some units 6. KurFER K. ' Operating expencnce with nuclear power plants dosed with phosphates. The nsk of attack increases with in Switt.criand'. ENS /ANS cnnference on nuclear energy. increased concentration in the boiler water, the concentra. European Nuclear Society /American Nuclear Sociery, Pans, tion efficiency of boiling sites, and the temperature. If the 7, f,Pg3.f[f, 3 . Steam generator tube failures: world expenence mechanism presented, which postulates that thinning is in water-cooled nuclear power reactorsin 1974'. Atomic i{nergy caused by separation of a phosphate rich liquid phase, is of Canada Ltd.1975. Al{CL 5242 ' Steam generator tube fail-correct, attack should not occur below about 275*C. ures: world expenence in water-cooled nuclear power reactors in 1975'. Atomic Energy of Canada Ltd.1976, ALCL 5625. The.dentmg. damage which has resulted in tube crushing,
- 8. FLETCllER W.D. and M ALINOWSKI D.D. Operatmg ex-suEEort.pl. ate distortion and stress corrosion in areas of high perience with Westinghouse steam generators. AucL Technol, distortion, is .a coasequence of rapid acid chloride attack of 1976, Vol. 28,356-373.
mild steel in the annular crevice between the tubes and 9. M A LINOWSKI D.D. and l LETCllER W.D. ' Update of opera-
, drilled, hole support plates. Changes in material and the tron with Westanchouse steam generators *. ANS Transactions Annual Meetine New York,1977. Vol 26,425.
design of the tube support will substantially reduce vulner. & RQ and DONATI LR. An evaluanon of PwR steam 3 ability to this type of attack. In vulnerable stations,i.e. those generator rubing alloys.Nuct Energy,1978, Vol 17,0ct., No. with dnlled mild steel support plates, the extent of attack 4.291-299. 131
EFFECTS OF COPPER AND NICKEL --- 0 COMPOUNDS ON THE CORROSION OF PRESSURIZED WATER REACTOR
~
STEAM GENERATOR MATERIALS EARL L. WHITE and WARREN E. BERRY Battelle Columbus 1.aboratories. 305 King Avenue, Columbus. Ohio 43201 Received October 30,1980 Accepted for Publication April 3,1981 E ymt w a w a.w.w > a?"M" - INTRODUCTION Copper corrosion products often are found in Pot boiler tests have been conducted to study the sludge at tubesheet and tube support areas of' steam effects of copper and nickel compounds on the tube generators in pressurized water nuclear reactors wastage and support plate denting phenomena ob- (PWRs). The source of the corrosion products is the served in steam generators of pressuri:ed water re- small amount of corrosior. that occurs on copper actors. The results of these tests revealed that copper alloy tubes in the condensers and feedwater heaters compounds produced denting when chloride was in the steam systein circuit. Questions have been present, but were not a necessary ingredient since raised as to the role of these copper corrosion prod-NiCl2 produced even more severe denting in the ucts on the corrosion performance of the Inconel 600 absence of copper. The pot boiler consisted of seven tubing in the steam generators. Two of the major steel umbrellas mounted on a heated Inconel Alloy corrosion problems encountered tirdate are wastage 600 tube under boiling conditions at 288'C. Six tests. (general thinning) of Inconel 600 tubing at tube sup-each of 30-days duration, were conducted with all port crevice areas and beneath deposits on the tube-volatile treatment (morpholine at pH 9.0 to 9.2) m sheet, and accelerated attuck of the steel support each test. Water chemistry of 50 ppm phosphate (as plate in the tube-to-support plate crevice area that in Na2HPO4) to 15 ppm chlorine (as Nacl) produced extreme cases has led to collapse (denting) and wastage on the inconel tube but no denting (fast cracking of the Inconel 600 tubes and to cracking of linear magnetite growth) of the steel umbrellas. the support plate itself. Adding CuO sludge and substituting CuClfor NaClin the phosphate system reduced the wastage attack on g,egg7,ung the tubing and produced only incipient denting on the steel umbrellas. Water chemistries of CuO sludge- Considerable wastage (general thinning) has been CuCl (15 ppm chloride) or Fe304 sludge-ViCl 2 (15 observed on inconel 600 tubing at such restricted ppm chloride) produced extensive denting, but no t]ow areas as support plates, tubesheet sludge de-wastage, with the attack by NiCl being2 more severe. posits, and anti-vibration clamps in PWR steam gener-The Nacl alone or American Society for Testing and ators operated with coordinated phosphate boiler Materials sea salt phis NiFe:04 shedge (15 ppm chlo- watcr treatment."* Inconel 600 tubes have exhib-eide in both testsi produced no denting of steel um- ited considerable wastage, particularly in those cases brellas nor wastage of Inconel tubes, perhaps because where units were operated at low sodiu m-to-PO. tests were not conducted for sufficient time to de- ratios after caustic stress corrosion cracking (SCC) relop acid-chloride conditions beneath the umbrellas. problems developed in several plants that had been Microprobe examination revealed that the chloride operated at high sodium-to-POa ratios. The wastage concentrated at the steel surface in the umbrella- phenomenon seems to be best explained by the tube crevices of those specimens that exhibited retrograde solubilities of the di-sodium and tri-sodium denting. For the most part, nicAel(hur not copper / phosphate salts. Under localized flow restriction and was associated with the chloride except ut the steel concentrating conditions. the solubilities of the di-surface. and tri-sodium salts are exceeded. these salts precip-itate out of the solution, and the adjacent solution becomes acidic as the result of the low of alkalinity. NLILLAR TLCllNOLOGY VOt. 55 OCT.1981 0029 5450/M1/0010 0135502.00/0 019M1 ANS 135
l
- l White and Berry EFFECTS OF COPPER AND NICKEL ON THE CORROSION OF PWR GENERATORS j Thus, wastage appears to be the result of acid phos- on the wastage and denting performance of in-phate attack. conel 600 and carbon steel.
Accordingly, most PWRs have switched to all-volatile treatment (AVT), which consists of hydrazine Test Matrix for oxygen scavenging and a volatile alkaline pH agent Six 30-day tests .were conducted to study the ] such as ammonia or rnorpholine. Thus, most PWRs i effects of various water chemistries and contaminants j now in service are bemg operated with AVT ,m the on denting and wastage. The test environments are steam side, but with residual phosphate still m the described in Table !. steam generator from prior phosphate treatment m many units. Some recent plants and new plants just Test I was conducted with Na:HPO., chloride (as . coming on linc have been using AVT from inception. Nacl), and AVT (morpholine to maintain a pH of The change from the phosphate water treatment 9.0 to 9.2) to establish that the pot boiler was capa-I l to the so-called AVT 1hydrazine alone or with an ble of producing wastage. amine) has led to yet another corrosion phenomenon Test 2 was performed with AVT plus 15 ppm - termed denting. The features of this type of attack chloride (as Nacl) to establish whether AVT and neu-are heavy magnetite buildup m support plate tube tral chloride produce denting. ) holes, collapsed tubes, plastic stram, of support plates, and, in severe cases, cracking of support plate hole-to- Test 3 was performed with AVT plus CuO sludge I' hole ligaments and SCC on the inside (primary side) and 15 ppm chloride (as CuCl) to establish that cop- ! of short radius-bend Inconel 600 tubes.s The forces to per compounds will produce denting when AVT is accomplish these phenomena are generated by corro. contaminated with chlorides. I , sion products (magnetite-Fe30.) of the carbon steel Test 4 was performed with phosphate, AVT, CuO j tubesheet that grow at rapid (hnear) rates primarily at sludge, and CuCl to establish whether phosphate in-i tube support plate crevices. The accelerated corrosion hibited denting. is attributed to acid chloride attack in the crevice area. The source of chloride is condenser in-leakage. Test 5 was performed with AVT plus Fe30. and The concentrating conditions are probably a com. NiCl2to demonstrate that contaminants other than bination of localized boiling because of restricted copper can produce denting. (All equipment was flow and the occluded cell phenomenon that results flushed with HNO 3to remove copper prior to Tests 5 in negative chloride ions migrating into the crevice to and 6.) maintain charge neutrality when positively charged Test 6 was conducted with AVT plus NiFe:04 ferrous ions form m the crevice. . and synthetic sea salt to provide a simple simulation in service, attack has been most severe m steam of service conditions. generators that operated on phosphate chemistry i prict to changeover to AVT, are on brackish or sea- Apparatus water condenser cooling, and have a high surface area A m del boiler was designed and built to use of copper alloys in the feedwater and condenser por- , umbrella-type steel specimens similar to those used tions of the system. However,it should be noted that successfuHy m studies conducted at Combustion En-denting has occurred to some degree in units that , gmeermg.' A sketch of the boiler portion is shown in ! have operated only on AVT, and in units that use freshwater for condenser cooling, and has not oc-curred in some units that have copper alloys in the TABLEI secondary circuit.' The factors that contrjbute to denting have not been fully evaluated. Environments for Pot Boiler Tests Thus, the effects'of copper corrosion products on Environmental Conditions l corrosion performance of ineonel 600 tubing need to be investigated unde,r phosphate water treatment Test Na:HPO.. Chloride. conditions and under post phosphate water treatment Number ppm AVT' Studge ppm conditions with AVT chemistry and chloride contam-50 Yes None 15 (Nacl) ination. Accordingly, the International Copper Re. 1 search Association has rettuested Batte!!e's Columbus j (,, "'
]5 y,
N c]' j Laboratories to undertake such an investigation 4 50 Yes CuO 15 (CuCl) (INCRA Project No. 290). This paper presents the 5 None Yes Fe30 15 (NiCli )b 6 None Yes NiFe:0 15 (ASTM ses salt)" results of this research.
'All volatile treatment-pil maintained at 9.0 to 9.2 by morpholine EXPERIMENTAL PROCEDURES additions.
Pot boiler tests were performed to study the
- ASTM = Amenean Society ror Testing and Matenals.
etTeets of phosphates. copper oxides, and chlorides ' Synthetic ses salt per ASTM D.1441. NLTLEAlt TLCilNOLOGY , VOL SS OCT.1981 136
MATERIALS PERFORMANCE IN t --tm a NUCLEAR PRESSURIZED WATER ! REACTOR STEAM GENERATORS STANLEY J. GREEN and J. PETER N. PAINE Electric Power Research Institute. P.O. Box 10412 Palo Aito. California 94303
~
Received October 20.1980 Acc:pted for Publication May 12.198: md11 J2LdfE2EZZETH2d Early in 1977, industry leaders recognized that solutions to the growing number and extent of steam The nuclear industry has had a rariety of re. generator problems would not be forthcoming with-liability problems with pressurised water reactor out a concerted effort on the part of affected utili-steam generators. Most of these problems have been ties. The Steam Generator Owners' Group (SGOG) associated with corrosion and mechanicallv induced (see Table 1) was established in 1977 under the damage, including secondary water intergranular cor. auspices of the Elcetric Power Research Institute for rosion and stress corrosion cracking (SCC), primary the specine task of developing optimum soluti9ns and water SCC. wastage, high cycle fatigue, and frettin'g facilitating their implementation by the sponsoring and wear of the inconel 600 or incolay 300 tubes, utilities. plus accelerated corrosion of carbon steel tube sup. There are 43 utilities in the U.S. that have or port structures in crevice regions. Corrosion and plan to have a total of 89 PWR units in operation by trachanically induced damage are caused by complex the end of 1986 (Ref. l).-Of these 43 utilities. 23 interactions of water chemistry, thermal-hydraulic have contributed to the formation and operation of design, matcrials design choices. fabrication methods. the SGOG. In addition. 4 non-U.S. organizations and secondary plant materials, design, and operations _ representing S7 PWR units are u(filiated with the Corrosion has affected ahnost 90% of steam genera. SGOG. This broad selection of sponsoring utilities, tors operationalprior to I977. resulting in forced and representing steam generators that are operating. scheduled outages to plug or sleere tubes and repair under constructien. and on order, has ensured that or replace generators. Utility operators hare begun to operational or reliability problems and design options respond rigorously with improved operating and receive attention fre n a broad range of professional maintenance procedures that reduce air and cooling disciplines, n.at a reasonuble statistical basis is avail-yyater inleakage; with installation of fadiflow con- able to determine th. extent of and rate of growth or densate polishers, titanium or stainless steel con- progression of the various iwues that have arisen. and densers, retubed feedwater heaters, and moisture that many options for resolution or correction are separater reheaters: and with modifications to Inakeup expiored. water and blowdown systems. The Stcam Generator Operating difficulties reported by the various Owners' Group continues to provide a focus for sponsoring utdities have been cuteyorit.ed, as shown dere/opment work to understand damage mecha- in Table 11. Also listed in Table 11 are potential msms, proride remedialactions, and effect transler of problems that have been observed only in accelerated technology to the utility operators. laboratory model boiler testing but that may become problems to the operating steam generators following an extended incubation or mitiation period. Table 111 provides a current plant-by-plant listing of various steam generator problems caused by corro-INTRO D UCTION sion attack.h"' The extent of corrosion as shown in Table ill suggests that no operatmg plant can be con-The nuclear power generation industry has ex- sidered immune. F.ach year, a high percentage of perienced a variety of reliability problems with opera ting reac tors req uires outage time to repair pressurized water reactor (PWR) steam generators. corrosion attack in their ste.un generators. as shown 10 NUCLLAR TLCilNOLW;Y VOL. 55 OCT.1951 0029 5450ml/nul0 00luin100/0 Ol9%lANS
l . f* Green and Paine MATERIALS PERFORMANCE IN PWRs TABLE 111 . . 7 Plant Listing of Steam Generator Problems Corrosion Attack Vibration
- ~ .
e 7 :;
;g .: 3 =
I E 5Y E.$ llt h '8 8k I I 21 u; $3
] E,1 di .E 5 -g _ -
- t= ~3 ms i Es i Approximate ,y 0] {j .] g t, jy j q% q "-y? E g Date of *ag .5 t .5 E to 'E j 8 .= 5 o .5 *- jd -5 o Unit Commercial Start NSSS' 5 ': 34 3E 3b Cm CS 5
C eXi OCU =0 [X 15 - U 55 %$ C1 0D n 3 0 E Yankee Rowe 6/61 W X Seini 6/64 W ScNA (Chooz) 4/67 A X V C )nnecticut Yankee I/68 W X X X V San Onofre 1 1/68 W X X X X V. X X Obrigheim 3/69 K X X X Zorita 8/69 W X V
- Beznau I 12/69 W X X X
- Ginna 3/70 W X X X X Mihama 1 11/70 W/C X Point Beach 1 12/70 W X X X -
Robinson 2 3/71 W X X X U i Palisades 12/71 C X X Beznau 2 3/72 W X X X- X Stade 5/72 K X X
~
Mihama 2 7/72 W X X Point Beach 2 10/72 W X X X Maine Yankee 12/72 C X Turkey Point 3 12/72 W X X X X
. Surry I 12/72 W X X X X Surry (original tubing) 2 5/73 W X X X X X X Surry (retubed) 2 9/80 W X Zion 1 6/73 W X X Ft. Calhoun 6/73 C Oconee 1 7/73 B X X X X X Turkey Point 4 9/73 W X X X X X Borssele 10/73 K X Prairie Island i 12/73 W , X X X Zion 2 12/73 W X Kewaunee 6/74 W
. Indian Point 2 7/74 W X X X X Oconee, 2 9/74 8 X X X X Three Mile Island i 9/74 B X X Takahama 1 11/74 W X X X V Oconee 3 12/74 B X X X X X
'K = KWU (All Alloy 800 except Obrigheim) LWCHW = Light W. iter Cooled. Heavy Water Moderated Reactor W = Westinghouse FiC = Framatome/Creusot-Loire C = Combustion Engineering H = Hitachi B = Babcock & Wilcox M = Mitsubishi Heavy Industries. Ltd.
U = Unidentified Signal A = Ateliers de Constructions Electriques de Charleroi S. A. I V = AVB Wear with Westinghouse Belgium /Framatome l (Continued) 12 , _ NUCLEAR TLCilNOUX;Y VOL. 55 OCT.1920
7 Green and Paine MATERIALS PERFORMANCE IN PWRs TABLE !!! (Continued) Corrosion Attack Vibration
=~ :
^
3
==
g= I NI $$ Mi Eh di j EN di in 3 $$
*= Ts^ 6 o
s 5e e
.M. e e > < " - g Approximate ,,e ;y
.g --,
a pw .p u q'g p= , y Date of fj .c t c ." c j y 3gy jg go ;- git .g .c " j, g Commercial c* - 8g- 53 58 E 000 90 5 CE OD 3 0
- - c.t Unit Start NSSS*
12/74 B X X X ANOl 2 12/74 W X X
- Praine Island Doel i 2/75 W X X Biblis A 3/75 K X ,
X X Rancho Seco I 4/75 B C X X Cahert Cliffs 1 5/75 W ,X X X Ringhals 2 5/75 Cook I 8/75 W X e Tihange 1 9/75 W X X X . Genkai i 10/75 W 2 11/75 W X X Doel Takahama 2 11/75 W - Sfillstone 2 12/75 C X .. 5/76 W X Trojan Neckarwestheim GKN I 10/76 K s Indian Point 3 8/76 W X Mihama 3 12/75 W ' Salem 1 6/77 W X St. Lucie i 12/76 C X Biblis B 12/76 S X Crystal River 3 3/77 8 X Beaver Valley 1 4/77 W Calvert Cliffs 2 4/77 C Ikata 1 9/77 MHI Davis.Besse i 11/77 B X W X X Farley 1 12/77 Fessenheim i 12/77 F/C Fessenheim 2 3/78 F/C Ko.Ri I 6/77 W X 6/78 W X X North Anna 1 Cook 2 6/78 W Three Mile Island 2 12/78 B X X Bugey 2 2/79 F/C X X Bugey 3 2/79 F/C OHI I 3/79 W
*K = KWU(All Alloy 800 except Obrigheim) LWCHW = Light. Water Cooled, Heavy Water. Moderated Reactor W = Westinghouse F/C = Framstome/Creusut.Loire C = Combustion Engineering H = Hitachi M = Mitsubishi Heavy Industries, Ltd.
B = Babcock & Wilcox A = Ateliers de Constnictions Electnques de Charleroi S. A. U = Unidentified Signal V = AVB Wear with Westinghouse Belgium /Framatome (Continued) 13 NL' CLEAR TECilNOLOGY VOL. 55 OCT.1981
Green and Paine MATERIALS PERFORMANCE IN PWRs TABLE !!! (Contmued) Corrosion Attack Vibration
- U 5=
E 5 j;:s e,2 E _!. ::t*g g
- 2 _
M o
.E ,4 s m f3 4 =ll = ,o , =
v- s 7 D
= > a . x=
- * *
- H *J o < . * .= m
, Approximate Date of
,y y5 y .d.,
a
.5 'E
=
P j(( ,jj i y
.E 2 y-{
=-
~- g
.5 2 'E g 8 -3 j 5 -E ?
Commercial i5 E EE ~5 3 X ~5 X '5 5% Y' 5 s E Unit Start NSSS* 3O 5.g '3 5 *- E CCC OC 0 C1 0D k 0 Fugen (LWCHW) 2 3/79 H/M Bugey 4 7/79 F/C Esensham 10/79 K OH1 2 12/79 W Bugey 5 II/79 F/C Salem 2 1/81 W Gravelines Bl 12/79 F/C Tricastin 1 12/79 W Almarez 1 12/79 W
- North Anna 2 8/80 W X McGuire 1 8/80 W Sequoyah 1 10/80 W
, ANO-2 3/80 C Dampierre 1 7/80 F/C ( Farley W 2 11/80
*K = KWU(All Alto. ,,00 except Obrigheim) LWCHW = Light. Water Cooled. Heavy Water Moderated Reactor W = Westinghouse F/C = Framatome/Creusot4aoire C = Combustion Engireering H = Hitachi B = Babcock & Wilcox M = Mitsubishi Heavy Industries. Ltd.
U = Unidentified Signal A = Ateliers de Constructions Electriques de Charleroi S. A. V = AVB Wear with Westinghouse Belgium /ltramatome clearly define the failure mode and (b) ditTerences Figure 1 presents a histogram of the industry and lack of precise terminology in plant-to-pLmt history of tube plugging from all causes. (Data are reporting. It is useful, however, to review operating taken from Refs. 2 through 8.) The tube plugging plant tube plugging history because patterns and rate has varied fro.n 0.2 to 0.7';F of tul'es in service trends can be discerned and used as general guides, plugged each year. Since steam generators can operate even though specifie application has normally proven at full power with as much as 20's of their tube fruitless. , bundle plugged. this overall plugging rate would not TABLE IV Cumulauve Tube Defects Versus EFPD* to December 31,1977 Resetors Tubes Percent Percent EFPD in Survey With Defects with Defects In Survey With Defects with Defects
<500 16 3 19 253 970 799 O.31 500 to 1000 2S 15 54 385 622 6S37 1.77
>t000 35 28 80 370 301 9239 2.51
*Elfective full power days (from Ref. 3).
14 NUCLLAR TI CIINOt.UGY. VOL 55 'OCT.1981 l
- __ _ _ A Green and l'.nne MATERIALS PERFOl01ANCE IN PWRs !
LL Uccer scan OD surface lane region .y ctreurr femntial % Pnmary enlet Upper lp cracks y p y / tubesheet , .= . . . . g h, ) / ,( ~M I ..#..
- h , , ,,
- i It M
M' \ Fiftssnth support - -- p. plate wear N < s Aujuliary '
, , v ,l f tee ater . e.h;*;;,*;
f~ (, g
- l., ~.
,.;3 k'
.... . . ,- p5-1
__g f[ . 7 Fourtssnth tube / ?. -- * ' .* * - ,. '.
- 4. I r ~~
- support . alate _/ - . '
n nonerv - S am .c ... erosion. corrosion , l_ . . .-r ' :. .
* . : ?. -f . . . . ..
, Feecwater . . .;.* ;* * '
}.
M" h s- - . - - "., e ., b~ l i .L u..' _ ,,; s, U .&. % , . 4-~. N Aspirating I surfsee 50X [
.rt D;ngs at ninth -/,, I steam Fig. 8. Intergranular corrosion of Inconel 600 in the tubel q
support plate . / p;,3, 1g n tubesheet crevice. 3
,r- support otate
, - sw Distorted y eddy current / \ Incidences of OD initiated cracking. unrelated to RO
9"
-\ ( ) Lower
' tubesheet crevice corrosion, have been observed. Examination of a tube sample removed from the region near the
( i .s ll upper tubesheet of an OTG5 rnealed SCC. Sulfur $ *p t Pnmary outlet / was present and may have been involved. Sources of # sulfur compounds can include ion exchange resin il beads, resin fines, and sulfuric acid regen 7 ration of d Fig. 7. Sketch of OTSG with illustrated problem areas. full-flow condensate polishers. Additional sources are 3 sulfuric acid treatment of cooling tower water and the natural occurrence of sulfur compounds in sea-M water and freshwater. ** l with the components of the sludge pile are in com- { munication with a fully wetted crevice. Plant startup +5 followed by crevice dryout woulJ leave all the im- . h-l purities in the crevice, and these would be added to %. Tube gD y I or refreshed at. the next @utdown/startup cycle. It is F ~ tL { f. ~ Y q.S.Th possible that concentration of deposits from this p' '[r"" w - IT mechanism can Se supplemented by those that could n ' d, M
} ,
4 . ., ; [ l enter as concentrated species from the sludge pile r under the influence of gravity and capillary llow. - Jh 3 * ~* A major effort is getting under way to define the * *
- j causative species. the thermal-hydraulic factors. and
.jy
. .. .f- -
i the mechanisms of corrosive attack. The mechanisEI ' il
"",'~,
responsible for general mtergranular attack and its
$(
relationship to previously ex perienced caustic-in . - - duced, stress-assisted cracking is not yet clear. Labo ,' , i ! ratory data have shown unequivocably that sodium ; :- !~ . fg i hydroxide will induce stress corrosion crac king ,' (,'. - (SCC)." but tests with mixed corrodents-sodium' ' - Q hydroxide plus sulfates ulicates, sludge components.k }. ,, '
~
? L etc.-have not demonst rated the t ransition t rom i *'
YubeID 100X cracking to generd intergranular attack of essentially :
- h..!
every grain boundary.1: is this latter condition that ; Fig. 9. Intergranular anack of inconel 600 with stress crack- [ I-warrants considerable further study. j ing m the rubc.tubcshcr t erevice. NUCLEAR TLCilNOL(x.Y VOL. 55 (M1.19M I 19 l
Green and Paine MATERIALS PERFORMANCE IN PWRs hole or tube hole to llow hole had failed. Recently. a Oxygen. Copper. am/ Nickel preliminary inspection of a steam generator removed The roles of copper, nickel, and oxygen in the from service at Surry 2 was conducted. It was noted denting process are not fully understood. In labo-that a small piece of support plate had broken loose ratcry testing, copper, nickel, and iron ions have by failure of all surrounding ligaments. In an operat- acted as accelerators of the denting corrosion process. ' mg dented steam generator, compressive forces on the One explanation for the role of these species is that corroding steam generator bundle arc Sufficient t they increase the generation of crevice acidity. A hold the support plates m place in spite ofligament slightly different mechanism has been proposed in
~
which a half-cell reaction is set up, which can drive s the corrosion product oxide is formed and the the crevice reaction by taking up the electrons support plates grow, the flow slots in the center of produced b} the oxidation ofiron. the plate between the hot leg and cold leg tend t in samples removed from actual steam generators, become smaller. One of the better mdicators of small beads of copper have been found in the dented dent,mg progression is the closure of the flow slots. In region. Similar observations have been made in the Surry 2 steam generator referred to above, it laboratory model boiler dents. The contribution of was found that at the support plate closest to the c pper, nickel, and iron ions to the pH of the crevice tubesheet, closure was essentially all from the hot leg in a enloride solution continues to be evaluated. side (highest heat flux). At the upper support plates. the motion from the hot leg and cold leg sides was Tube Mechanical 0arnage approximately proportional to the relative heat Mechanical damage that affects steam genera' tor fluxes. This pattern is not unlike that found cise-where. In other SGOG projects, the progress of tubes generally fa!!s into the categories of tube wear, denting in steam generators is being followed by due to either fretting or impingement, and tube changes in flow slot dimensions. fatigue cracking. The forcing function for tube Denting has been shown in the laboratory to be fretting and fatigue is tube vibration induced by flow. - caused by the deposition of a high concentration of This topic, plus its relationship to steam generator acidic species in the crevice. Work has focused on design and operation, has been addressed in a work-chloride species, although in principle sulfate may si p sponsored by the SGOG and reported in Ref.21. react similarly. A discussion of denting reactions is orrosion is an additional factor that complicates gieen in Refs.19 and 20. In this paper, we present evaluation of mechanical tube damage ~When coupled, caly a synopsis of the mechanistic studies. mechanical damage and corrosion are often syn- , ergistic. This ef fect varies from the erosion of passi-vating films, which allows further corrosion, to the Chloride accelerating effects of some aggressive environments Denting has been successfully reproduced m. the on metal fatigue. laboratory with acid chlorides and has been arrested in the laboratory by the addition of a base. indicating Anti-Vibration Bars that denting is an acid reaction. In addition, the Fretting has caused extensive tube damage at the amount of superheat available at the tube / tube 3 anti-vibration bars in several steam generntors (see support plate crevice is known to cause sufficient Fig. 21). This damage has been attributed to the concentration of acid chlorides to result in observed inadequate design of the anti-vibration bars utilized accelerated corrosion rates. Also, the morphology of in the steam generators and has been corrected by the dents reproduced in the laboratory with acid their replacement. conditions (e.g., acid chlorides) is similar to the morphology of plant dents. Broached Support Plates D # Recently. tube fretting has been postulated to have occurred in the vicinity of the 15th support There are recent data that indicate that steam generators at one unit dented after one cycle of plate of OTSGs. An affected tube removed from the operation. Denting was presumably initiated after the steam generator evidenced rectangular wear patterns inadvertent addition of ~l35 kg (300 lb) of resin correspnnding to the shape of interfacing land areas into the steam generator. At the operating conditions. of the 15th support plate. Figure 22 is a magnified the resin will decompose and form sulfurie acid.The photograph of one sueli tube. Metal loss varied. but in some regions approached a 0.25-mm (0.010-ina
; unit was operated at load following this occurrence.
and cleanup was attempted by blowdown. Available reduction m tube wall thickness from the OD. The test data give conflicting results on the role of sulfates pattern of metal low suggested that there was or had on denting. Furthet testing with sulfates n present!y b,;en some deviation f rom the desired normal inclina-tion between the tube and tubesheet, perhaps due to undet way or planned to clarify their role.
~
l l l ) Green and Paine M ATLRIALS PLit00RM ANCL IN P% Rs 6 bream generator. Potential sources of foreign objects i ETCHED PHOTOMICROGRAPHS ! nclude i devices used to aid in the fabrication of steam. l THROUGH D AMAGE AREA ' generJtors, such as tube-nose guides used during tubc l l InstJllation, objCett Carried in and lost by Workers during fabrication. maintenance. or inspections, ob-l that became dislodged along the secondary l ,,
'^ 0432 in. !jectssystem outside the steam generator and t I inte the steam generator, etc.
In one case. a wire spring ~200 mm (8 in.) long l g and 5 mm (0.2 in.) in diameter was found in the i p secondary side between the tubes in the steam gen-erator. The result was a ruptured tube. In addition. wall thickness reductions of 62 and 20'7e were found i by eddy current inspection in two adjacent tubes at' the same elevation as the rupture. l i 0 O' 'n Fatigue Cracking , Fig. 21. Fretting wear dsmage st anti-vibration bars. Circumferential tube er.:cking has occurred in! I OTGSs (Ref. 22). In 1977 several tubes were removed l I from service. and examination revealed that they hadl l bowed tubes, misabgnment forces or unusual flow- incurred several types of distress (see' Figs. 23 and 24).l mduced forces. Cracks observed at the upper tubesheet and highest There is concern that because of design changes tube support plate along the open inspection lane (i.e.. geometry and/or matenal) to the support plates . were OD initiated and were oriented circumfer-utilized in recirculatmg steam generators. trettmg and entially. Large through wall eracks as well as small. possible crack tmtiation in the tretted regmn can partial-wall cracks were found. All identified cracks become a problem. especially when one considers the were located at defects (i.e. at regions of localized 40-yr design life of the steam generator. wall thinning) on the OD surface of the tubes. It wa indicated that the regions of erack initiation e:dtibite Foreign Objects transgranular structure-sensitive fructographi e char-
- Fretting type titbe damage can be caused by the aeteristics. It has been theorized that the initiation of I pre ence of foreign objects in the secondary side of a these cracks may have been environmentally assistedq Powible initiating mechanisms f or these cracks in; clude fretting, fatigue and localized sertosion. En vironmental fatigue testing completed to date for 101 i .-
- -rqf 73 3; ;
7 fl . . . . " . . y'
. .s -:.x .
- u i
{$,, , (e,
. 3 , .; ;
.i c ..
?. 'o nt ,Y. ' . * .
' ' : ; f ,.
- l
' ~
'.:bi
- . . y s _ g ,: ..
jw ..- np ll g ,, .s r w $ rf '* \ g d,, a n. l'.f/ $[:j)gQ I d l'. . \ !!l ~
'; j {W& -
I h ' p' ,W N 7[ #~ p d,
- gi
.I 3 ((
*g..
w
.u l 1. \ ;.
.4 .
*L -
yi , _ .; . y '-r- A; s ,.tne w, : T_ . I
~ '
- \ N" ^_ , , ,g._g}
-pj w, -
k -
- c. . Y d -
itu -5 I Fig. 22. Frettmg. wear at tube support plate m Imd snuet Fig. 23. Fati;ue crack vi inconel o00 tube matenal at upp area ot OTSC tube. tubesheet ut OTSG
-e
. 1 l
._ _ _ _ _ _ _ _ =
l
. j
< Green and Paine MATERIALS PERFOltWtNCE IN PWRs
*y . W!"' -
.r._'~.--., '"**W.l' :2
, .. . V._ -
. ., g
- I a -
, t, w- --1
~ . -qc-e --- - -ag 7
N -
~)" Va'~N~ '.
'~
$h ~~~T
. .1, 7. . .".% _h..
,;a +_----,--y~~
M i hx l #:.M Wy v. '-
"5
- n 4 '.~.3 nW k:$_h=5d n d, A l
g
~
a4 M R s'. 1 * % o, y f, .. -*a- -vi.*:C%,-
-r 3 t. q y* *
. :. U. .;' m 4.".
Y '~ . ( ) . . i, ' ' Fig. 24. Ci,cumferential transgranular fatigue crack of in-
.3
* '-p -
. m .p N." r.
conel 600 tube material at upper tubesheet of OTSG.
\~ %,,,_
.- N~ -. --;.
'C**
- ,2 cycles with 288'C (550*F) air, 238'C (550*F) water. ' ' 44 e + ~
E.45-15J GEM ~ ' A : / % * ' 8N and 288'C (550*F) steam showed no difference in '""7 .
~
fatigue properties. There was no reduction in fatigue streneth with cavitation-induced damage, but there
~
Fig. 25. " Candle. flame" erosion. corrosion near tube support was a 30'7c reduction in fatigue strength caused by an in land contact area of OTSG tube, electrochemical notch. Other factors affecting these failures may ir, elude entrained water (flow loading), uneven flow distribu- systems-related actions to the steam ge'nerator and tion, and/or bowing of lane tubes (bending stress). the secondary plant. Major mechanical modification Possible aggravating events that could cause cyclic to an operating steam generator is, of course, not tube stresses include turbine stop valve testing, possible. A variety of chani;es to secondary plant l auxiliary feedwater injection, and waterhammer. design. maintenance, and operation can be made to I reduce the ingress of impurities to the steam genera-
/mpingement Effects tor. This section briefly outlines the acticas being it is believed that the impingement of solid o lua r in some instances implenwnted in operat.
ing plants. particles on steam generator tubes under operating The principal dil,h,eutties encountered to date conditions can cause mecham. cal tube damage (see F.ig. 35). Ispingement effects were judged to be one relate to the corrosion of the inconel tubes and car-of the possible causes for tube defect indications in bon or low a'loy steel tube support plates and tube-two OTSGs. Potential sources of solid particles sh ut. .Mo&eations to tk conoshe envimnment include any particles entering the steam generator e ns ute a major appmad to bring about insprove-nwnt. De M has shown dat n is not capable ut f*om the feedtrain and particles entrained from ! within the steam generator. Sources of internally pmMng bunering capa@ to pmwnt mmsn s l entrained particles include sludge previously de- I cations having high concentrations of impuritieg posited. corrosion products generated willu.n the But plant data show that neid adherence to an exact-steam eenerator e.g.. magnetite at tube support plate ine program for maintaining me
- . . h-purity A\,T water crevices. and materials remaining from the manu-
- E ' "E *"Y *"W""""E.'""'*"'
fa. cturmg procen. Coupling AVT with periodie 11ushing oi the steam generator and/or improved layup technittues or en-sironments to remove hideout themica!s appears to l REMEDIAL ACTIONS / CORRECTIVE MEASURES ba a prudent preventative treatment for those steam generators not experiencing significant corrosmo to Operating steam generators have proven su s- date. Penodic chemical cleaning prior to the time t ceptible to a variety of reliability problems. Potential when a steam generator has experienced m.nor dif fi-I solutions to these problems must include a variety of culties also appears to offer pronose.1 or those steam chemistry-related actions as well as mechanical- or genera tors in which denting or tube crackmg has i wucumu.cmasuuo_wu m wt.- m ______w
OQED u:.xnnCATE T SERVICE % Ag 2 sw C
'Ihe undersigned, a m$mber of DAME / SAFE, certifies that on this' served a copy of the attached hwnt on each member of the Service. List by United States regular mail, Special Delivery, or by other means as ap- ,
propriate. Dateck: July 30,1982 } %g Diane Chavez, DAARE/ SAFE Subscribed and sworn t:o before me en this Sohtday of July,1982. hm k Y ()btary Public .
' y cemn;an4k Y5 r _
SERVICE LIST s 1 '
, Morton B. Margulies, Esq. Alan Bielawski Administrative Judge and Chairman Isham, Lincoln & Beale Three First National Plaza Atomic' Safety and Licensing Board US Nuclitar Regula. tory Commission Chicago, Il. 60602 Washirston, D.C. 20a55 Joseph Gallo Dr. Richard F. Cole Isham, Lincoln & Beale Atomic Safety and Licensing Board 1120 Connecticut Ave., N.W.
US. Nuclear Regul'atory Commission "
' Washington, D.C. 20555 , , (,s n on, D.C. 20555 Dr. A. Dixon Callihan Douglass W. Cassel, Jr.
Union Carbide Corporation Jane Whicher P.O. Box Y. Oak Ridge, Tennessee 37830 [ute1300 Atomic Safety.and Licensing Board 109 N. Dearborn US. Nuclear Regulatory Commission Chicaao, Il. 60602 Washington, D.C. 20555 Dr. Bruce von Zellen N Dept. of Biological Sciences Secretary Attn Chief, Docketing and ;NIU DeKalb, Il. 60115 Service Section US. Nuclear Regulatory Com:Pission Myron Cherry Washington, D.C. 20555 Cherry & Flynn Mr Steven Gcidberg i Three First National Pla::a Ms. Mit:1 Young ' Suite 3700 Office of the Executive Legal Director Chicago, II. 60602 US. Nuclear Regulatory Commission Ms. Betty Johnson Was hington , ' D .C . 20555 1907 Stratford Ln.
.\- FJcc@Rcg#tn XL @33@7_ ,1}}