ML17340A639

Forwards Bnwl Re Steam Generator Repair Program. Requests Placement in Docket Files.Info Forms Basis for Conclusions Reached in Review of Facility Steam Generator Repair Program

ML17340A639
Person / Time
Site:	Turkey Point
Issue date:	01/21/1981
From:	Grotenhuis M Office of Nuclear Reactor Regulation
To:	Office of Nuclear Reactor Regulation
Shared Package
ML17340A640	List: ML17340A639 ML17340A641
References
NUDOCS 8101230558
Download: ML17340A639 (42)

Text

REGULATORY FORMATION DISTRIBUTION SY M (RIDS)

ACCESSION /BR:8101230558 DOC ~ DATE: 81/01/21 NOTARIZED:

NO FACIL:50-.250 Turkey point PlantE Unit 3P Florida Power and Light C

50 251 Turkey Point Planti Unit QP Florida Power and Light C

AUTH BYNAME AUTHOR AFFILIATION GROTENHUI8 p M, Operating Reactors Branch 1

RECIP ~ NAME RECIPIENT AFF IL'IATION Operating Reactors Branch 1

SUBJECT:

Forwards BN'(iL 801212 l tr re steam generator repair pr ogt am, Requests placement in docket files ~ Info forms basis for conclusions reached in review of facility steam generator repair program.

DISTRIBUTION CODE:

HOOIS COPIES RECEIVED:LTR

(

ENCL j SIZE:

TITLE:

DCD Material NOTES:

DOCKET ¹ 05000250 05000251 RECIPIENT ID CODE/NAME'2 03 INTERNAL: NRC 'PDR EXTERNAL: LPDR

$ 0D:

COPIES LTTR ENCL 1

1 1

1 RECIPIENT ID CODE/NAME EG FILE NSIC 01 00 COPIES'TTR ENCL 1

1 1

1 6 <OTEN ((L((S J/N 2 g 3981 TOTAL NUMBER OF COPIES REQUIRED0 LTTR ENCL

t.

II i F II Ij i

W'It CI I

ij tl II r

I

~

I 1

li MWM 0\\ IWM M g. II

i/~i/g t Y>>

f*

F c

p~~ ~ ~ as~

fD tsJ CD T-l' CDVl C3 Q 'A<A a-~c xx'=~558 4-d ~ AM/.

December 12, 1980 t,b 19)gg~

y p P<OTPvx,.rrzn

'4L 'acific tVorthwest Laboratories P.O. Box 999 Richtand, Mfashington U.S.A, 99352 TcIephone (509)

Telex 15-2874

'alt Pasedag U.S. Nuclear Regulatory Commission Washington, DC 20555

Dear Walt:

This letter is in reference to our telephone conversations of November 26 and December 2, 1980, t.oncerning the potential drop of a steam generator and the subsequent release of radioactive material.

Your assumption that 20/ of the total activity in the steam generator (as calculated prior to decontamination) would be released is, in our view, highly conservative (i.e., it is an overestimate of what would actually be released).

It is, however, the assumption used by PNL in the assessment for moving a Surry steam generator to Hanford.

We believe that the amount of activity released due to such an occurrence would be on the order of a few percent, certainly less than 105, of the amount on the tubes.

It is likely that some material might be removed from the tubesheet by abrasion during the fall, but the radioactive deposits would likely remain attached to the abraded metal.

Our conclusions are based pri-marily on the fact that PWR corrosion films are compact (or tightly adherent) and would not shake or break free during the postulated drop.

It is unlikely that the radioactive deposits would be easily released.

I am enclosing an article on deposits in primary systems.

For further infor-mation on decontamination, you may also wish to refer to NUREG/CR-0968, The Im act of Decontamination on LWR Radioactive Waste Treament Systems and to ppen ix of NURE R-9 now sn pu scat>on

, Radio o ica ssessment of Steam Generator Removal and Re lacement:

U date an Revssson.

If you have further questions or need additional information, please call L. G. Faust (444-3613),

L.

D. Perrigo (444-9024) or me (444-3754).

Sincerely, G.

R.

Koenes Associate Section Manager Dosimetry Technology Section Enclosure xt

54.

Nature of deposits on BAR and PWR primary system surfaces relation to decontamination A. B. Johnson, Jr., B. Griggs and F. M. Kustas, Pacific Northwest Laboratories, and R. A. Shaw, Electric Power Research Institute BHR and PHR crud/oxide layers on out-of-core surfaces differ in composition and morphology.'hey also differ in response to decontamination reaqents.

INTRODUCTION l.

Investigations are underway in several countries to improve decontamination methods for nuclear reactor coolant systems.

An important aspect of those investigations is to better understand the compositions

.and morphologies of the radioactive deposits which must be removed from reactor coolant system surfaces during the decontamination.

6.

The two factors which most stronoly influ-ence the types of films which form are:

the system materials; and the coolant system envi-ronment.

7.

The BHR and PHR reactor coolant systems differ markedly in both materials and environ-ments.

The system characteristics are surmarlzed in the BHR and PHR sections, resoectivelv.

2.

Studies have been sponsored by the'Electric Power Research Institute (EPRI) at'Battelle-Northwest to investigate decontamination methods for boiling water (BWR) and pressurized water (PHR) reactors.

The work has included detailed examination of radioactive films on specimens from components exposed in BHR and PHR reactor coolant systems.

This paper sunmarizes observa-tions regarding characteristics of the radio-active films'etails of reagent studies are given elsewhere (ref. 1-4,9).

INVESTIGATION OF BWR DEPOSITS BWR reactor coolant s stem materials and env>ronment he nrtncinal matertala in the HIIR reactnr coolant system are:

Zircaloy fuel claddinn, stainless steel pipinq and reactor vessel clad-ding, and cobalt-base alloy (Stellite) hard-facing alloys

~

Other deposit sources are:

corrosion products from the feedwater system (carbon steel and stainless steel);

and species Formation of radioactive de osits on reactor coolant s stem surfaces 3.

The need for decontamination of reactor a

ractono e>rc r

reactor coolant.,

The species are carried to the reactor core, undergo neutron activation and are re-dispersed to out-of-core locations.

The radioactive coolant-borne species are transported in both dissolved (ionic) and particulate forms.

which leak throunh the condensate demineralizers 9.

The BWR reactor coolant operates at 290'C, near neutral oH (normal ranqe:

5.5 to 7.5).

Radiolysis generates dissolved oxyqen concen-trations of 0. 1 to 9.5 opm (linuid phase).

Thus, deposits on BHR system surfaces form under conditions which are substantially oxidizino.

4, Thus radioactive films on reactor coolant system surfaces arise from two sources:

10.

The orincipal out-of-core material in BHR reactor coolant systems is 304 stainless steel.

The Principal source of BHR specimens for this study was 3t14 stainless steel recirculation bypass pipe sections from the Vermont Yankee reactor; Specimens from HiIlstone 1 and Ouad Cities 1 reactors were examined for comparison.

oxide growth on the metallic surfaces; radioactive ions carried in the reactor coolant are incorporated into the growing oxide; deposition of radioactive particulates on the outer surface of the oxide.

X Fe.

IBX Cr. 8".Ni. 2" !in, 1% Si.

coolant systems ar>ses because radioactive For example, if the condenser is Admiralty brass, deposits develop on out-of-core surfaces, some zinc and copper carried into the BWR reactor Numerous materials, including iron-base, nickel-vessel will be incorporated into deposits on

base, cobalt-base and coPPer-base alloys, release reactor coolant system surfaces.(5) f i

f th'r osion product to the

II t

0 Decontaminflt Idln 11, Selected examinations and analyses were conducted on the following types of specimens; i>> ad'Xu SEH/EOX inspection of the outer film surfaces, before and after reagent treatments; inspection of film cross-sections, before and after reagent treatments; inspection of particles separated from the film during or after reagent treatments (particles deposited on filters or removed from the surface on tape);

inspection of the as-received

film, separated from the substrate by a methanol-bromine treatment; inspection of par ticles removed by mechanical scraping.

12.

The scope of the paper permits no more than an overview of the results.

Hor hol ies and Com ositions of BWR films Vermont ankee fs ms.

A prominent feature surface is a shallow attack at metal g~ain boundaries.

A similar attack is evident on archive specimens, suggesting that it was caused by post-fabrication pickling.

14.

The Vermont Yankee crud/oxide layer has four features, surrmarized in Table 1

and in Figure 1.

The Vermont Yankee film morphology was more complex than films from the other two reactors.

In fact, the dual film structure on Hillstone and Ouad Cities oipes apnears to be characteristic of films from several BNRs.{5-8)

The Vermont Yankee and Hi)lstone 1 films dif-fered markedly in cnmnosition and morphology.

but the decontanina

.ion rates were very similar, (4) for example, in an NEDTA/oxalic acid/citric acid reagent at 180'C.

In both cases, the compact inner fi'lm dissolved, awhile outer layers

remained, detached but undissolved.

Apnroxi-mately 80 to 90>> of the radioactivity was

removed, appearing to correspond to dissolution of the compact inner layer.

15.

Activity removal from Duad Cities pipe specimens was less'effective than for specimens from. the other two reactors.

The guad Cities surface film was super imposed on a rouqh metal

surface, having fissures up to 0.01 in.(250um) deep.(4,6)

As with the shallower (up to ~¹m) fissures on other oipe surfaces, the fissures appear to be a consequence of pre-service pickling.

The hypothesis that radioactive species are trapped in the fissures is consis-tent with effects of ultrasonic vibration on pipe surfaces after decontamination.'bout 28>>

of the residual activity was removed from the guad Cities specimen, while the corresnondinq fractions for Hillstone and Vermont Yankee were only 4" and 10", respectively.

Table 1

Summary of BNR Film Characteristics on 304 SS pipe Feature C~om osition a)

A - Vermont Yankee Film Res onse to Reaoents b)

Superficial Particles

~l to 10um Shroud

~0. lum thick Faceted Crystals c)

O.l to lorn a

Compact Oxide Layer up to 2am thick Fe Cr Ni 85-93 3-7 1-2 Unknown 52 18 7

63-76 12-16 5-8 Cu 2n 1

2-4 Undissolved Dissolved or Mashed Away 1

22 1

9-19 Undissolved - Detached llashed Away by Turbulence Dissolved Principal source of activity 8 - Hi llstone 1 Fi lm{7)

Outer Layer Inner Layer C - Ouad Cities 1 Film(6) 86 7

6 61 ll 25 t

1 t

3 Undissolved Dissolved Principal Source of activity Outer Layer Inner Layer 90 5

5 68 18 14 t

= trace NR

• not reported EOX

• Energy Dispersive X-ray NR NR Undissolved NR NR Dissolved (Efficiency for radio-activity removal reduced by rouoh metal surface)

~aBased on EBX analysis - Bot folly ca'librated:

percent metallic species normalized to 100".

b)Based on exposure to EDTA, NTA, HEDTA, etc.

c)Based on analysis of particles detached by reagent treatments.

t

Paper 54:, Johnson ct:

p 'QI'~

~

. ">'i)

~w'

,lg

~

C,i

'Red ~'-C~>>

@V.~~ <<~",

. J+m6~ ".!:

~y~44$

0 Ilc

. 4+tgj Rj&

~5fj "@4N$

~

~..

'pti~Wp'o'4'i, v4'u P

~

~ g'~~i

~ '&

~

1um

~Q

-'g+{'t y5

(,I

~;.i~ o, ~-,.")N..

~

A

~gee A~

I\\

3 3 Nil ~

r.~.

)um

~3 4$

4'.

Fig.

1 Cross Section and Surface Views (SEH) - Vermont Yankee BMR Recircu1ation Bypass Pipe Specimens.

teppYI~

~

l,~

~

~,

~ ~

e ~

'I \\

(~

i~ wP'~,

MA4fg+~y7~"-'.-

IQ um

'Iglp, +~/

Pg' m+ip C.

.f'r.

,'iE" I y

.gM k~'~c 1N9gw-

~rl ft~ 1

'.p'gCl

.tow -,

~

~

10um

'r

'Jt

..r~.

~ g tl, 33um tl'big

" +ri."

=

lacks~~

Fig.

2 Cross Section and Surface Views (SEH) - Indian Point 2

PMR Steam Generator Tube Specimens.

275

t I

Dmcnnrarn nation 16.

Crystallographic analyses by X-ray diffrac-tion suggested that the principal species on the BHR pipe surfaces are Fe203 and spinels of the type:

llix Fe3 04, with radioactive species sub-stituted at cation positions.

The Fe203 remains a~ter the decontamination, and some spinel species also were undissolved.

Studies by general Electric Co. (5-7) suggest that the species which dissolve are nickel ferrites (spineld).

The outer layer, which is higher in iron, remains undissolved.

Our experience suggests that the faceted crystals which do not dissolve are somewhat higher in Cr than the layer which does dissolve, but only mildly so (Table 1). It is possible that the species higher in Cr are interspersed with a lower Cr species which dissolves.

If so, the EOX analyses-may not properly resolve the lower Cr species.

17.

In summary, the BMR films respond favorably to car+on decontamination reagents (e.q.,

EOTA, HEDTA, tlTA), most effectively when they are re-ducing.

The most important aspect of the decon-tamination appears to be dissolution of the compact layer adjacent to the metal surface.

This layer appears to incorporate a larae frac-tion of the radioactivity (e.q.,60Co+2),

probably as metal cations which substitute into the spinel oxide lattice.

~18.

A relatively small f~action of the radio-activity is associated with species not dis-solved by the decontamination

{Fe203 and certain

spfnels, probably relatively high in substituted Cr).

However, these detached particles (O.l to 10am) must be dealt with in a decontamination, by filtration and by assuring flow conditions which will not deposit the particles at unwanted locations.

ItNFSTIGATION OF PWR DEPOSITS PllR Primar S stem Haterfals and Environment 9.

The pr>nc>pa materia s in the PMR primary circuit are:

Zircaloy fuel cladding, Inconel-600 steam generator tubing and stainless steel piping, reactor vessel cladding; cobalt-base alloy hard-facing materials.

20.

The PWR primary coolant has additions of boric acid (~200g ppm 8) and lithuim hydroxide (0.2 to 2.0 'ppm Li).

Hydrogen is added to control oxygen at <0.01 ppm during operation.

Hhile metals continue to oxidize, the system is substantially reducing.

The system temperature is 320-340'C.

21.

The principal out-of-cor e material in most PMR primary systems is Inconel-600 steam gener-ator tubing.

Tubing filmed in two PMR steam generators has been examined, by methods indi-cated for BllR films.

Horoholo ies and Com ositions of PWR Films 22.

he surface of a PWR steam qenerator film had the following characteristics (Fig. 2):

a sparse population of large superficial particles; a layer of fine platlets; a network of dark grafn boundaries.

Table 2 surmarizes compositions of two PMR films.

23.

Fig.

2 also shows a cross section of a crud/oxide layer.

The layer thickness is ~0.5um.

Thin filaments of the layer protrude into the metal along what we shall see are grain boundaries.

24.

Confirmation that the film is keyina alona grain boundaries developed from treatments which separated the film from the base metal.

Figure 3 shows separated film flakes.

Location A is a

flake oriented to show the side adjacent to the metal.

The filaments embedded in the metal are now evident around the grain boundary contours.

Location B shows a flake oriented to show the coolant side.

25.

Responses of the PMR films to reaaent treatments are still being investigated.

To,

date, the following responses have been obser;ved:

flecontaminatfon occurs slowly in the common reagents

{EDTA, HEOTA, HTA)(4)

Addition of H202 to EDTA caused only temporary increases in decontamination rate(<)

Addition of reducing metal ions (e.aes Cr+2) caused Iapid increases in decontami-natfon rate(4>

Certain oxidizing and reducina reaqents dissolved the base metal, undercutting the film, removing it almost completely as flake-like particles.

DISCUSSIOH rd.

Our studtes have clearly demonstrated that BWR and PMR films respond differently to common reagents such as EDTA, HEOTA and llTA.

The BHR films are easier to decontaminate, Early in the studies we beqan to examine the questions:

Do BWR films, formed under oxidizing conditions, respond most effectively to reducing decontamination reagents?

Do PWR films, formed under "reducing" conditions, respond most effectively to oxidizing reagents?

27.

Responses of BWR films to decontamination were in line with the hypothesis:

oxidizing reagents were not effective; reagent effective-ness increased with increasing reducing character.

28.

Beaker studies also suggested that reduc-tion processes promoted rapid decontamination.

BWR pipe specimens which included freshly-cut metal surfaces had higher decontamination rates than specimens where f)q exoosed metal was coated with silicone.<

The observation suggested that corrosion of the fresh metal produced electrons (e.qas Fe'e+

+ 2e) which were available for reduction of species in the radioactive film.

29.

The PllR films showed some mildly favorable responses to oxidizing reaqents, aareeina with the hypothesis nroposed above.

However, the response was not dramatic.

In fact, the PWR films also have shown favorable responses to reducina conditions.

(4) 27fs

bem e

,>pl4 4

I L

~p

~,e, s) r aY

.O~

$ a ~Q

.edm,mp

~

3000X Fig.

3A Crud/Oxide Flake Separated From Inconel-600 Steam Generator Tube IP2-3-7 During Decontam-ination Treatment at 180'C 2000X

Fig, 3B Crud/Oxide Flake from IP2-3-7; Oriented to Show Side Exposed to Reactor Coolant Table 2 - Su+nary of PWR Film Characteristics on Feature

~Com ositiooa)

Fe Cr Hi Inconel-600 Si Zn Steam Generator Tubing Remarks A - Tube IPE Superficial Particles 2 to 9um Compact Layer in d~

40-86 4-15 7-40 2-5 11 23 63 1

2 B - Tube I-4-30HL Superficial Particles 2um Compact Layer in d~

A u

Compact Layer detached 18-31 18-26 40-60 2-4 0=2

31 26 40 35 29 31 2

2 4

2 Compact Layer detached 24 46 28 3

Film Removed by Scraping 29-34 25-27 36-38 3-7 1-2 Compact Layer detached 14 55 18 10 3

Compact Layer. detached 9

40

-19 17 t

See Fig. 2,.right-hand views'nalysis of cross section, near film

'uter edge; base metal shine probable Separated from base metal by REPjA/

OXALIC acid reagent, 120-150'C o

Separated from base metal by HEDTA acid reagent, 180'C

>J Film up to 3um thick; minimal base metal shine Film flakes separated from base metal by EDTA/H202 reagent at 180'C b)

Film flakes separated from base metal in methanol-Bromine at 25'C Some base metal possible-appears minimal normalized to 100::.

Reagents also include boric acid and lithium hydroxide, characteristic of PWR system chemistries.

277

l l'

1 Decont~~inat ion 30.

On the hypothesis that the radioactive species are largely present at substitutional positions in the compact oxide, removal of the radioactivity appears to require partial or complete dissolution of the oxide, depending on

--"- - the degree of activity reduction required.-

On ---"

this hypothesis, it does not appear possible to remove the radioactive species (e.g..

60Co) by selective dissolution even though higher ratios of 60Co to removed film were observed.

This conclusion also was reached by other investi-gators. I4) 31.

It is important to minimize post-decontami-nation activity buildup associated with corrosion of the decontaminated surfaces.

Our studies on oxidized laboratory specimens suggest that the post-reagent film growth at reactor temperatures was proportional to the amount pf oxide dissolved during the reagent treatment.'I This suggests

'hat modest DFs'ay leave some protective oxide, minimizing subsequent corrosion.

This supposition has not been confirmed on reactor specimens in our studies.

32.

The BWR films comprise Fe203 and a range of

spinels, some soluble in common reagents; some insoluble.

The particulate species which remain undissolved after reagent treatments include Fe203 and certain spinels, apparently those with relatively high Cr contents

(~18").

However, the major fraction of the radioactivity is soluble and therefore may be removed by ion,

exchange, while particulates can be removed on filters.

This technology has been demonstrated in dilute reagent gecontamination performed in Canadian reactors.(>>)

33.

The PW(t films comprise a variety of spinels and oxides.<>>>

Our analyses indicate that the sparse population of superficial particles tend to be high in Fe. but have a wide range of compositions (Table 2).

The radioactive oxide layer is sometimes less than one micron thick.

The relative insolubility in common reagents probably relates to the high Cr content (30-55

, Table 2).

However, we have been able to achieve relatively effqctive decontaminations by metal ion reduction,<4> or by both oxidizing and reducing reagents which undercut the film, resulting'in a large fraction of the act1vity being removed as particulates.

OF Radiation Level Before Decontamination Rad>at on Leve fter Oecontam>nation REFERENCES 1.

JOHNSON, A.B., Jr.,

GRIGGS B.,

REMARK J.F.

Investigation of Chemicals and Methods of Dilute Reagent Decontamination for Potential Application in Light Water Reactors.

Paper 32 CORROSION/78,

Houston, TX, March 6-10, 1978.

Available from tlACE, P.O.

Box 218340,

Houston, TX 77218.

2.

JOHtlSOt<, A,B., Jr.,

GRIGGS B., Comparison of Decontamination Characteristics of Films from BWR and PWR Primary Systems.

Paper

161, CORROSIO.'l/79, Atlanta, GA, March 12-16, 1979.

Available from llACE, (see above).

3.

JOHNSON, A.B., Jr.,

DILLOtt, R.L., GRIGGS. 8, REMARK, J.F.,

Investigation of Alternate Methods of Chemical Decontamination--Second Progress Report.

January 1,

1977 to December 31,

1977, BN-SA-703-2, Battelle Northwest, June 1978.

~ ~

t1 jt h1

~ 'rte

~ '%

~

~ ~ I

~ I~

4.

JOHtlSON, A.B., Jr.,

GRIGGS, B., DILLON, R. L ~

Candidate Reagents for Activity Reduction in BWR and PWR Pr1mary Systems.

Proceedings ANS Topical Meeting on Decontamination and Oecomnis-

sioning, Sun Valley, ID. September 16-20,1979.

5.

BLOK, J.

Summary Report l<ater Chemistry Program Extension.

NEDC-21550, General Electric

Company, San Jose, CA, June 1977.

6.

ANSTINE. L.O.

BWR Decontamination and Corrosion Product Characterization.

NEOE-12665.

March 1977.

7; Characterization of Corrosion Products on Recirculation and Byoass Lines at Millstone l.

NP-949, Electric Power Research Institute, Project 819-1, December 1978.

8.

IWAHORI, T., MIZUNO T., and KOYAMA, H. Crud Composition on Heat Transfer Surface, Paper 37, CORROSION/78, Houston, TX, available from NACE (see ref.

1 above).

9.

A Study of Dilute Reagent Decontamination for Application in Boiling Water Reactors, NP-Electric Power Research Institute, Project 828-1, (to be published).

10.

LeSURF, J.E.

Water Chemistry of Nuclear Reactor Systems.

Proceedings of BtlES Conference, Bournemouth, October 24-27, 1977.

p. 426.

11.

PETTIT, P.J.,

LeSURF, J

~

ED

STEWART, W.B.,

STRICKERT, R.J.,

and VAUGHAN, S,B.,

Decontami-nation of the Douglas Point Reactor by the CAN-OECON Process.

Paper 39, CORROSION/78, Ma@h 1978 (see Ref. 1).

12.

SHEE, J.L.

Dissolution Characteristics of Metal Oxides in Itater Cooled Reactors.

Proceedings ANS Topical Meeting on Decontami-nation and Oecommiss1oning, Sun Valley, ID.

September 16-20, 1979.

ACKNOWLEOG>IEttTS The following have made significant contributions to this study:

R.L. Dillon, J,L. Daniel, J.E.

Coleman, O.B. Mackey, and R.L. McOoweil.

We are grateful to the Electric Power Research Institute (EPRI) for permission to publish the results of this study.

27S

REGUI.ATJRIIIIFOitMLTIOR l)ISTRIO~II IOi<l SOS'<

(RIOS)

ACCESS IO'I '<<R: R 1 fi123055o

'>uC. I'A E.:

fil /01/?1 FACIL:50 750 Turkey Point r'I ant r Uni t 3F F I or bQ 251 TurF;ey POint r" I ant, Uni t fi F C I Or AUTH ~ ',EAI'IF 'U IT<u'Q AFF Jf JA TIO'>

6ROTE.IVrIU15 i I~.

-OoeratinU Reactors branch

'?EGIN ~ 'Il~"0,

<I;C II Ical'FFTLIATIOf Oner~tin<a reactors EIrafich fIOTARI7LIi:

fv'9 ioa power and LiEiht C

ida "ower anc! Liqnf C

DOCYET 05 0002~

5UUJEC1':

Forwards I"!~L Rulcid I tr re steam generator r er.air procirao} ~

'~eciuests placement in docket files. Ir fo forIIS basis for conc lusions r cached in r eview of faci I i ty stea~

Eienerator r epa i r or os r a ll ~

DISTRISJI IUN C(IOE:

BVO IS

\\,OFIES RECE IVEI):LTR L E'tCL j SIZF.:

TITLE:

DCO fiaterial BIO TF. 5!

HFCIFIC t I'~

CriOE/'A'-*e.

iElf'7 I 5 I TTR EWCL RE C I P IF'IT In CO')l;/~AfsF.

COPTER L TTN e.>JCL Ii>TFHI'IALe l~i(C FOR PXTF f(lsAL; LP()II 1

1 1

1 G FILf 01

>ISTC n~

1 1

1 1

64oTKNHuis Jill 28 1981 TOTAL NUEIBEH UF C'v'PIE5 HKMUJREO:

LTTR ENCL

i/a.i/c t fFL-MW

~~,.+Wc ~

CW

~c ~~ ~c - ~+~M..c

.~> r~z

~-l~= p-~=I M~Rc--W

'2Zw

~<+ac'M~~ M~ ~~a ~

Q44 z-6C4Z~4 ~

~l r

pe

~o~'"

\\

~-CW

. -558

December 12, 1980 fitItt ] 9 lggt P,QTvlpz.,err

~

'4h'acific'Northwest Laboratories P.O. Box 999 Richland, Vvashington U.S.A. 99352 Telephone (509)

Telex 15-2874 Walt Pasedag U.S. Nuclear Regulatory Commission Washington, DC 20555

Dear Walt:

This letter is in reference to our telephone conversations of November 26 and December 2, 1980, t.oncerning the potential drop of a steam generator and the subsequent release of radioactive material.

Your assumption that 20K of the total activity in the steam generator (as calculated prior to decontamination) would be released is, in our view, highly conservative (i.e., it is an overestimate of what would actually be released).

It is,

however, the assumption used by PNL in the assessment for moving a Surry steam generator to Hanford.

We believe that the amount of activity released due to such an occurrence would be on the order of a few percent, certainly less than 105, of the amount on the tubes.

It is likely that some material might be removed from the tubesheet by abrasion during the fall, but the radioactive deposits would likely remain attached to the abraded metal.

Our conclusions are based pri-marily on the fact that PWR corrosion films are compact (or tightly adherent) and would not shake or break free during the postulated drop.

It is unlikely that the radioactive deposits would be easily released.

I am enclosing an article on deposits in primary systems.

For further infor-mation on decontamination, you may also wish to refer to NUREG/CR-0968, The Im act of Decontamination on LWR Radioactive Waste Treament Systems and to Appen ix of N

RE CR-now m pu ication

,= Radio o ica ssessment of Steam Generator Removal and Re lacement:

U date an Revision.

If you have further questions or need additional information, please call L. G. Faust (444-3613),

L.

D. Perrigo (444-9024) or me (444-3754).

Sincerely, r',/

G.

R.

Hoenes Associate Section Manager Dosimetry Technology Section Enclosure

I AI

54.

Nature of deposits on BWR and PWR primary system surfaces relation to decontamination A. B. Johnson, Jr., B. Griggs and F. M. Kustas, Pacific Northwest Laboratories, and R. A. Shaw, Electric Power Research Institute 6.

The two factors which most stronoly influ-ence the types of films which form are:

the system materials; and the coolant system envi-ronment.

BllR and PWR crud/oxide layers on out-of-core surfaces differ in composition and morphology'.

They also differ in response to decontamination reaqents.

INTRODUCTION 5.

In sunmary, the radioactive film is a

l.

Investigations are underway in several composi te of the deposited crud laver and the countries to improve decontamination methods for growinq oxide on the metal substrate.

nuclear reactor coolant systems.

An important aspect of those investigations is to better understand the compositions arid morphologies of the radioactive deposits which must be removed from reactor coolant system surfaces during the decontamination.

2.

Studies have been sponsored by the'Electric Power Research Institute (EPRI) at'Battelle-llorthwest to investigate decontamination methods for boiling water (BWR) and pressurized water (PWR) reactors.

The work has included detailed examination of radioactive ff1ms on specimens from components exposed in BWR and PWR reactor coolant systems.

This paper sumaarizes observa-tions regarding characteristics of the radio-active films.

Details of reagent studies are given elsewhere (ref. 1-4,9).

Fortration of radioactive de osits on reactor coo ant s stem surfaces 3.

The need for decontamination of reactor coolant systems arises because radioactive deposits develop on out-of-core surfaces.

'fumerous materials, including iron-base, nickel-

base, cobalt-base and copper-base alloys, release a fraction of their corrosion product to the reactor coolant.

The species are carried to the reactor core, undergo neutron activation and are re-dispersed to out-of-core locations.

The radioactive coolant-borne species are transported in both dissolved (ionic) and particulate forms.

4.

Thus radioactive films on reactor coolant system surfaces arise from two sources:

oxide growth on the metallic surfaces; radioactive ions carried in the reactor coolant are incorporated into the growing oxide; deposition of radioactive particulates on the outer surface of the oxide.

7.

The BllR and PltR reactor coolant systems differ markedly in both materials and environ-ments.

The system characteristics are summarized fn the BWR and PllR sectfons, resoectfvelv.

INVESTIGATIO!< OF BWR DEPOSITS BWR reactor coolant s stem materials and env ronment he nrlncloel materlelt ln the RHR reector coolant system are:

Zircaloy fuel cladding, stainless steel pipinq and reactor vessel clad-ding. and cobalt-base alloy (Stellfte) hard-facing alloys.

Other deposit sources are:

corrosion products from the f'eedwater system (carbon steel and stainless steel);

and species which leak throunh the condensate demineralizers.

For example, if the condenser is Admiralty brass, some zinc and copper carried into the BWR reactor vessel will be incorporated into deposits on reactor coolant system surfaces.(5) 9.

The BWR reactor coolant operates at 290'C, near neutral oH (normal ranqe:

5.5 to 7,5).

Radiolysfs nenerates dissolved oxyqen concen-trations of O.l to 0.5 opm (linuid phase).

Thus, deposits on B!IR system surfaces form under conditions which are substantially oxidfzino.

10.

The orfncipal out-of-core material in BllR reactor coolant systems is 304 stainless steel.

The principal source of BHR specimens for this study was 304 stainless steel recirculation bypass pipe sections from the Vermont Yankee reactor.

Specimens from lfillstone 1 and f)uad Cities 1 reactors were examined for comparison.

<<rrX Fe, 18K Cr, 8"tli, 2" !fn, lX Sf.

Decontamination Selected examinations and analyses were conducted on the following types of specimens:

/>> ainu SEH/EOX inspection of the outer film surfaces, before and after reagent treatments; inspection of film cross-sections, before and after reagent treatments; inspection of particles separated from the film during or after reagent treatments (particles deposited on filters or removed from the surface on tape);

inspection of the as-received

film, separated from the substrate by a methanol-bromine treatment; inspection of particles removed by mechanical scraping.

12.

The scope of the paper permits no more than an overview of the results.

Nor hol ies and Com ositions of BWR tflms ermont ankee i ms.

prominent feature surface is a shallow attack at metal grain boundaries

A similar attack is evident on archive specimens, suggesting that it was caused by post-fabrication pickling.

'l4.

The Vermont Yankee crud/oxide layer has four features,.surrmarized in Table 1

and in Figure 1.

The Vermont Yankee film morphology was more complex than films from the other two reactors.

In fact, the dual film structure on tfillstone and Ouad Cities oipes apnears to be characteristic of films from severa)

BllRs.(5-8)

The Vermont Yankee and lfllstone 1 films dif-fered markedly in cnmnosition and morphology, but the decontamfna ion rates were very similar, (4) for example, in an I!EOTA/oxalic acid/citric acid reagent at 180'C.

In both cases.

the compact inner film dissolved, while outer layers

remained, detached but undissolved.

Apnroxi-mately 80 to 90" of the radioactivity was

removed, appearing to correspond to 4isso1ution of the compact inner layer.

15.

Activity removal from Ouad Cities pipe specimens was less effective than for specimens from. the other two reactors.

The tluad Cities surface film was superimposed on a rough metal

surface, having fissures up to 0.01 in.(250um) deep.(4,6)

As with the shallower (up to ~20um) fissures on other oipe surfaces, the fissures appear to be a consequence of pre-service pickling.

The hypothesis that radioactive species are trapped in the fissures is consis-tent with effects of ultrasonic vibration on pipe surfaces after decontamination.

About 285 of the residual activity was removed from the guad Cities specimen, while the corresnonding fractions for tlillstone and Vermont Yankee were only 4" and 10", respectively.

- Table 1

Summary of BWR Film Characteristics on 304 SS pipe Feature

~Com ositian a)

A - Vermont Yankee Fflm Res onse to Reaoents b)

Superficial Particles

~1 to 10um Shroud

~0.lum thick Faceted Crystals c) 0.1 to lum Compact Oxide Layer up to 2um thick Fe Cr Ni 85-93 3-7 1-2 Unknown 52 18 7

63-76 12-16 5-8 Cu Zn 1

2-4 Undissolved Dissolved or Washed Away 1

22 1

9-19 Undissolved - Oetached Washed Away by Turbulence Oisso'lved Prfncina) source of activity 8 - Millstone 1 Film(7)

Outer Layer Inner Layer C - Ouad Cities I Film(6) 86 7

6 61 11 25 t

1 t

3 Undissolved Dissolved Principal Source of activity Outer Layer Inner Layer 90 5

5 68 18 14 t

trace tlR

• not reported EOX

• Energy Oispersive X-ray tlR tlR Undissolved NR NR Oissolved (Efficiency for radio-activity removal reduced by rouoh metal surface)

~a Based on EBE analysis - llot folly ca'librated:

percent metallic species normalized to 100"..

b)Based on exposure to EOTA, NTA, HEOTA, etc.

c)Based on analysis of particles detached by reagent treatments.

1

PcpcL 54

Johnson ct PV

~

rQ'3 lun.~~i:[~~c~q'i

%f~

l IC fp?

Q ~

CC@,

lpm.~'7+i'~

lum Qc

~tn Fig.

1 Cross Section and Surface Views (SEH) - Vermont Yankee BWR Recircu1ation Bypass Pipe Specimens.

\\

wl Fpz

~

~

lap

'Sgla. +P

~+'" '~

\\tip

~ 1r*

J io,s (ii'

.j

~

~

~1 lg

'Hf~g

~ lOym

~.%

'<v~

l

~~ g0Pj+4 f -+<$49 yy,>sf, s, Jj Fig.

2 Cross Section and Surface Views (SEH) - 1ndian Point 2

PMR Steam Generator Tube Specimens.

275

D..d:dtdteani natiOn 16.

Crystallographic analyses by X-ray diffrac-tion suggested that the principal species on the 8'HR pipe surfaces are Fe203 and spinels of the type:

Nix Fe3 x04, with radioactive species sub-.

stituted at cation positions'he Fe203 remains

- a<t r the decontamination, and some spinel species also were undissolved.

Studies by General. Electric Co. (5-7) suggest that the species which dissolve are nickel ferrites (spinels).

The outer layer, which is higher in iron, remains undissolved.

Our,experience suggests that the faceted crystals which do not dissolve are somewhat higher in Cr than the layer which does dissolve, but only mildly so (Table 1).

It is possible that the species higher in Cr are interspersed with a lower Cr species which dissolves.

If so, the EOX analyses may not properly resolve the lower Cr species.

17.

In su+nary, the BWR films respond favorably to common decontamination reagents (e.q.,

EDTA, HEDTA, tlTA), most effectively when they are re-

ducing, The most important aspect of the decon-tamination appears to be dissolution of the compact layer adjacent to the metal surface.

This )ayer appears to incorporate a large frac-tion of the radioactivity (e.q.,60Co+2),

probably as metal cations which substitute into the spinel oxide lattice.

18.

A relatively small fraction of the radio-activity is associated with species not dis-solved by the decontamination (Fe203 and certain

spinels, probably relatively high in substituted Cr).

However, these detached particles (O.'1 to

)Ourn) must be dealt with in a decontamination, by filtration and by assuring flow conditions which wi11 not deposit the particles at unwanted locations.

INYESTIGATION OF PWR DEPOSITS PHR Primar 5 stem Materia)s and Environment 9 ~

The principa mater)a s in the PMR primary circuit are:

Zircaloy fuel cladding, Inconel-600 steam generator tubing and stain'less steel piping, reactor vessel cladding; cobalt-base alloy hard-facing materials.

20.

The PWR primary coolant has additions of boric acid (~2009 ppm 8) and lithuim hydroxide (0.2 to 2.0 'ppm 'Li).

Hydrogen is added to control oxygen at ~0.0) ppm during operation.

Mhi)e metals continue to oxidize, the system is substantially reducing.

The system temperature is 320-340'C.

21.

The principal out-of-core material in most PWR primary syst;ems is Inconel-600 steam gener-ator tubing.

Tubing filmed in two PWR steam generators has been examined, by methods indi-cated for BliR films.

Horoholo ies and Com ositions of PWR Films 22.

The surface of a PWR steam generator film had the following characteristics (Fig. 2):

a sparse population of large supe~ficial particles; a layer of fine platlets; a network of dark grain boundaries.

Table 2 surmarizes compositions of two PWR films.

23.

Fig.

2 also shows a cross section of a crud/oxide layer.

The layer thickness is ~0.5um.

Thin filaments of the layer protrude into the metal along what we sha))

see are grain boundaries.

24.

Confirmation that tbe film is keyino a)onq grain boundaries developed from treatments which separated the film from the base metal.

Figure 3 shows separated film flakes.

Location A is a

f)ake oriented to show the side adjacent to the metal.

The filaments embedded in the metal are now evident around the grain boundary cont:ours.

Location 8 shows a flake oriented to show the coolant side.

25.

Responses of the PWR films to reagent treatments are still being investigated.

To,

date, the following responses'ave been observed:

Ilecontamination occurs slowly in the common reagents (EDTA, HEDTA, HTA)(4)

Addition of H202 to EOTA caused only

/

temporary increases in decontamination rate(<)

Addition of reducing metal ions (e.ges Cr+2) caused Iapid increases in decontami-nation rate(4>

Certain oxidizing and reducing reaqents dissolved the base metal, undercutting the film, removing it almost completely as flake-like particles.

DISCUSSION Vd.

Our studtes have clearly demonstrated that BHR and PMR films respond differently to common reagents such as EDTA, HEDTA and HTA.

The BHR films are easier to decontaminate.

Early in the studies we began to examine the questions:

Do BMR films, formed under oxidizing conditions, respond most effectively to reducing decontamination reagents?

Do PMR filns, formed under "reducing" conditions, respond most effectively to oxidizing reagents?

27.

Responses of BMR films to decontamination were in line with the hypothesis:

oxidizing reaqents were not effective; reagent effective-ness increased with increasing reducing character.

28.

Beaker studies also suggested that: reduc-tion processes promoted rapid decontamination.

BWR pipe specimens which included freshly-cut metal surfaces had higher decontanination rates than specimens where g)e exoosed metal was coated with silicone.l The observation suggested that corrosion of the fresh metal produced electrons (e.q.,

Fe' Fe+

+ 2e) which were available for reduction of species in the radioactive film.

29.

The PMR films showed some mildly favorable responses to oxidizing reaqents, aqreeino with the hypothesis proposed above.

However, the response was not dramatic.

In fact, the PMR films a)so have shown favorable resnonses to reducing conditions.

(4)

burne:

~P$~

e ~

t

~ m iform fq

$)

.A~

m

~ Lm 3000X Fig.

3A Crud/Oxide Flake Separated From Inconel-600 Steam Generator Tube IP2-3-7 During Decontam-ination Treatment at 180'C 2000X Fig.

38 Crud/Oxide Flake from IP2-3-7; Oriented to Show Side Exposed to Reactor Coolant Inconel-600 Steam Generator Tubin9 Remarks Si Zn 2-5 1

2 See Fig. 2,.right-hand views Analysis of cross section, near film outer edge; base metal shine probable Separated from base metal by RSpjA/

OXALIC acid reagent, 120-150'C o

Separated from base metal by HEDTA acid reagent, 180'C b) 55 18 10 3

'40 19 17 t

Compact Layer detached 14 Compact Layer detached 9

Table 2 - Suranary of PWR Film Characteristics on Feature

~Com os itiooa)

Fe Cr Ni A - Tube fp2 Superficial Particles 40-66 4-15 2-40 2 to 9um Compact Layer in a~

11 23 63 8 - Tube I-4-30HL Superficial Particles 2Um Compact Layer ~n d~

2 2

4 2

31 26 40 29 31 46 28 Compact Layer detached 35 Compact Layer detached 24 3

Film Removed by Scraping 29-34 25-27 36-38 3-7 1-2 18-31 18-26 40-60 2-4 0-2 Film up to 3Um thick; minimal base metal shine Film flakes separated from base metal by EDTA/H202 reagent at 180'C b)

Film flakes separated from base metal in methanol-Bromine at 25'C Some base metal possible--appears minimal normalized to 100::.

Reagents also include boric acid and lithium hydroxide, characteristic of PWR system chemistries.

277

i

Occon Cawi'L Oiled'I oA 30.,=

On the-.bypothesfs that, the radioactive species, are largely present at substitutional positions fn the compact oxide, removal of the radioactivity appears to require partial or complete dissolution of the oxide, depending on

.""-.'he degree of activity reduction required.-. On--

this hypothesis, it does not appear possible to remove the radioactive species (e.g.,

6OCo) by selective dissolution even though higher ratios of 60Co to removed film were observed.

This conclusion also was reached by other investi-gators.i4) 31.

It fs important to minimize post-decontami-

, nation activity buildup associated with corrosion of the decontaminated surfaces.

Our studies on oxidized laboratory specimens suggest that the post-reagent film growth at reactor temperatures was proportional to the amount. qf oxide dissolved during the reagent treatment.<91 This suggests

'hat modest OFs* may leave some protective oxide, minimizing subsequent corrosion.

This supposition has not been confirmed on reactor specimens in our studies.

32.

The BWR films comprise Fe203 and a range of

spinels, some soluble in common reagents; some insoluble.

The particulate species which remain undissolved after reagent treatments include Fe203 and certain spinels, apparently those with relatively high Cr contents

(~18").

However, the major fraction of the radioactivity "is soluble and therefore may be removed by ion

exchange, while particulates can be removed on filters.

This technology has been demonstrated in dilute reagent ffecontamination performed in Canadian reactors.(>>

)

33.

The PWg fjlms comprise a variety of spinels and oxides.<12I Our analyses indicate that the sparse population of superficial particles tend to be high in Fe, but have a wide range of compositions (Table 2).

The radioactive oxide layer is sometimes less than one micron thick.

The relative fnsolubflity in common reagents probably relates to the high Cr content (30-55%, Table 2).

However, we have been able to achieve relatively effqctive decontaminations by metal ion reduction,l"> or by both oxidizing and reducing reagents which undercut the film, resulting"in a large fraction of the activity being removed as particulates, Radiation Level Before Decontamination Rad>at on Leve ter Decontamination REFERENCES 1.

JOHNSON, A,B., Jr.,

GRIGGS B.,

REHARK J.F.

Investigation of Chemicals and Hethods of Dilute Reagent Decontamination for Potential Application in Light Water Reactors.

Paper 32 CORROSION/78, Houston, TX, March 6-10, 1978.

Available from NACE, P.O.

Box 218340,

Houston, TX 77218.

2.

JOHNSON, A.B., Jr.,

GRIGGS B., Comparison of Decontamination Characteristics of Films from BWR and PWR Primary Systems.

Paper

161, CORROSION/79, Atlanta, GA, Harch 12-16, 1979.

Available from MACE, (see above).

3.

JOHNSON, A.B., Jr.,

DILLON, R.L.. GRIGGS, 8, REHARK, J.F.,

Investigation of Alternate Hethods of Chemical Decontamination--Second Progress Report.

January 1,

1977 to December 31,

1977, BN-SA-703-2, Battelle Northwest, June 1978.

~

I~

~ P ~

6

~

~ IH

~ 1

~ ~

I

~

~ ~

4.

JOHNSON, A.B., Jr.,

GRIGGS, B., DILLON, R.L.

Candidate Reagents for Activity Reduction in BWR and PWR Primary Systems.

Proceedings ANS Topical Meeting on Decontamination and Decomnis-

sioning, Sun Valley, ID. September 16-20,1979.

5.

BLOK, J.

Summary Report Water Chemfstry Program Extension.

NEDC-21550, General Electric

Company, San Jose, CA, June 1977.

6.

ANSTINE, L.D.

BWR Decontamination and Corrosion Product Characterization.

NEOE-12665.

Harch 1977.

7:

Characterization of Corrosion Products on Recirculation and Byoass Lines at Millstone l.

NP-949, Electric Power Research Institute, Project 819-1, December 1978.

8.

IWAHORI, T., HIZUNO T., and KOYAMA, H. Crud Composition on Heat Transfer Surface, Paper 37, CORROSION/78,

Houston, TX, available from MACE (see ref.

1 above).

9.

A Study of Dilute Reagent Decontamination for Application in Boiling Water Reactors, NP-Electric Power Research Institute, Project 828-1, (to be published).

10.

LeSURF, J.E.

Water Chemistry of Nuclear Reactor Systems.

Proceedinos of BNES Conference, Bournemouth, October 24-27, 1977.

p. 426.

ll.

PETTIT, P.J.,

LeSURF, J.E.,

STEWART, W.B.,

STRICKERT. R.J.,

and VAUGHAN, S.B.,

Decontami-nation of the Douglas Point Reactor by the CAN-OECON Process.

Paper 39, CORROSION/78, Ma@h 1978 (see Ref. 1).

12.

SHEE, J.L.

Dissolution Characteristics of Hetal Oxides in 1later Cooled Reactors.

Proceedings ANS Topical Meeting on Decontami-nation and Decommissioning, Sun Valley, IO.

September 16-20, 1979.

ACKNOWLEOG>fENTS The following have made significant contributions to this study:

R.L. Dillon, J.L. Daniel, J.E.

Coleman, D.B. Hackey, and R.L. HcDowell.

We are grateful to the Electrfc Power Research Institute (EPRI) for permission to publish the results of this study.

278

ACCESS'IU>>'I i'ICIER: R F4C IL:5 T

b0-251 A

TH ~, IAI i GROTE,iuHU I S r N.

RECIPE u REGULATOR,

'uFOHIolATIQN DISTHIbUI'TUtu "

'E'~

(R tOS) 101230556 I)'uL ~ DATE:

8 1/01/21 iIATAP.I7EO:

WO DOCKET urkey po>nt Plantr

<Init 3F Flnrida Power and Liant C

05000250 rkey po> nt i'1 ant r rjni t li F F I or i da Power and Liqn f C

05000251 AUTHu'Q Al.FIl IA TION-Qoeratrno Reactors Hranch 1

irEC IP Ic.1~I 1'EFF JLIATIOIu Oneretincj ReaCtOrS BranCh 1

SUBJECT:

Forwards iE liL 8Vlc1c!

1 tr re steam qener ator r enair pr ograrrr ~

"e'~uests>>

acement sn docket f i les. Ir fo forns Lasis for conclus>ons reacned in review of facility steam regenerator repair pronr attr ~

OTSTRTI)UT ION CIIOE:

HOOIS COF1ES RECEIVEI):LTR L EIICL,. J

'SIZF.'

ITLE:

OCO I>aterial t'IO TF.S:

HEC IF I Err I' 3 COBE/" ArlE

<rlP TES LTTR EiRCL REC IPIEI'<T IO CODE/WAE~E COPIES LTTR E'4CL IiuTEHIIIAL'rl(C POR EXTPHIIIAL: LPi)R 1

1 1

EG F ILE 0 1 NSIC 1

1 1

1 64omN Huis

)Cjjg gpss TOTAL NVEIBFH OF COPIES REQUIRED; LTTR ENCL

~ '/wi/s l Q~ A.s /

/,<~ Mm J'>

+A J74 ~AM g~ pZC~g

~C

~~~~< ~Cn'c ~

W~r = ~~~..e 3 ~-(

~

3Zw 6=

py~

n==:.P s ~ way Z~E'WC~

+< L4ACW~~~

~c~~~~~

r

<7

~

(o'q r

December 12, 1980 t

rs f98~

RQYt'an~ Tr Pacilic Northwest Laboratories P.O. Box 999 Richland, Washington U.S.A. 99352 Telephone (509)

Telex 15-2874 Malt Pasedag U.S. Nuc'lear Regulatory Commission Washington, DC 20555

Dear Malt:

This letter is in reference to our telephone conversations of November 26 and December 2, 1980, t.oncerning the potential drop of a steam generator and the subsequent release of radioactive material.

Your assumption that 20Ã of the total activity in the steam generator (as calculated prior to decontamination) would be released is, in our view, highly conservative (i.e., it is an overestimate of what would actually be released).

It is,

however, the=assumption used by PNL in the assessment for moving a Surry steam generator to Hanford.

We believe that the amount of activity released due to such an occurrence would be on the order of a few percent, certainly less than 105, of the amount on the tubes.

It is likely that some material might be removed from the tubesheet by abrasion during the fall, but the radioactive deposits would likely remain attached to the abraded metal.

Our conclusions are based pri-marily on the fact that PMR corrosion films are compact (or tightly adherent) and would not shake or break free during the postulated drop.

It is unlikely that the radioactive deposits would be easily released.

I am enclosing an article on deposits in primary systems.

For further infor-mation on decontamination, you may also wish to refer to NUREG/CR-0968, The Im act of Decontamination on LWR Radioactive Waste Treament S stems and to Appen sx of NUR CR-9 now sn pu

>cation Ra

~o o >ca ssessment of Steam Generator Removal and Re lacement:

U date and Revssson.

If you have further questions or need additional information, please call L. G. Faust (444-3613),

L.

D. Perrigo

{444-9024) or me {444-3754).

Sincerely, r.i G.

R.

Hoenes Associate Section Manager Dosimetry Technology Section Enclosure

54.

Nature of deposits on HVAR and PWR primary system surfaces

. relation to decontamination'.

B. Johnson, Jr., B. Griggs and F. M. Kustas, Pacific Northwest Laboratories, and R. A. Shaw, Electric Power Research Institute BllR and PMR crud/oxide layers on out-of-core surfaces dfffer fn composition and porohology.'hey also differ in response to decontamination reaqents.

INTROOUCTION 5.

In su+nary, the radioactive film is a

l.

Investigations are underway in several composite of the deposited crud laver and the countries to improve decontamination methods for qrowinq oxide on the metal substrate.

nuclear reactor coolant systems.

An important aspect of those investigations fs to better understand the compositions and morphologies of the radioactive deposits which must be removed from reactor coolant system surfaces during the decontamination.

6.

The two factors which most stronaly fnflu-ence the types of films which form are:

the system naterfals; and the coolant system envi-ronment.

7.

The BllR and PMR reactor coolant systems differ markedly in both materials and environ-ments.

The system characteristics are sumnarfied fn the BMR and PllR sections, resoectfvelv.

2.

Studies have been sponsored by the Electric Power Research Institute (EPRI) at'Battelle-

, llorthwest to investigate decontamination methods for boiling water'(BWR) and pressurized water (PllR) reactors.

The work has included detailed examination of radioactive films on specimens from components exposed 1n BMR and PMR reactor coolant systems.

This paper summarizes observa-tions regarding characteristics of the radio-active films.

Oetails of reagent studies are given elsewhere (ref. 1-4,9).

INVESTIGATION OF BMR OEPOSITS BWR reactor coolant s stem naterials and env ronment e princioal materials in the iillR reactor coolant s stem surfaces 3.

The need for decontamination of reactor coolant systems arises because radioactive deposits develop on out-of-core surfaces.

'lumerous materials, including iron-base, nickel-

base, cobalt-base and copper-base alloys, release a fractfon of their corrosion product to the reactor coolant.

The species are carrfed to the reactor core, undergo neutron activation and are re-dispersed to out-of-core locations.

The radioactive coolant-borne species are transported in both dissolved (fonfc) and particulate forms.

9.

The BMR reactor coolant operates at 290'C, near neutral oH (norma) ranqe:

5.5 to 7.5),

Radiolysfs nenerates dissolved oxyqen concen-trat1ons of 0.1 to 0.5-opm (liquid phase).

Thus, deposits on BHR system surfaces form under conditions which are substantially oxfdfzfno.

4.

Thus radioactive films on reactor coolant system surfaces arise from two sources:

10.

The orfncipal out-of-core naterial fn BllR reactor coolant systems is 304 stainless steel.~

The principal source of BHR specimens for this study was 304 stainless steel recirculatfon bypass pipe sections from the Vermont Yankee reactor; Specimens from H11'Istone 1 and Ouad Ci t1es 1 reactors were examined for comparison.

oxide growth on the metallic surfaces; radioactive ions carried in the reactor coolant are incorporated into the growing oxide; deposition of radioactive partfcu'Iates on the outer surface of the oxide.

5 Fe, 18% Cr. 8"lli, 2

.'fn, IX Sf, coolant system are:

Zfrcaloy fuel claddfnn, stainless steel pipinq and reactor vessel clad-ding, and cobalt-base alloy (Ste)lite) hard-facing alloys.

Other deposit sources are:

Formation of radioactive de osfts on reactor corrosion products from the feedwater system (carbon steel and stain'less steel);

and species which leak throuoh the condensate demineralizers.

For example, ff the condenser fs Admiralty brass, some zinc and copper carried into the BW!? reactor vessel will be incorporated into deposits on reactor coolant systen surfaces.{5)

Decontaninati nn ll.

Selected examinations and analyses were conducted on the following types of specimens:

i >> adieu SEMIEOX inspection of the outer film surfaces, before and after reagent treatments; inspection of film cross-'sections, before and after reagent treatments; inspection of particles separated from the film during or after reagent treatments (particles deposited on filters or removed from the surface on tape);

inspection of the as-received

film, separated from the substrate by a methanol-bromine treatment; inspection of particles removed by mechanical scraping.

12.

The scope of the paper permits no more than an overview of the results.

Mor hol ies and Com ositions of BWR tilms Vermont ankee i ms.

prominent feature surface is a shallow attack at metal grain boundaries.

A similar attack is evident on archive specimens, suggesting that it was caused by post-fabrication pickling.

14.

The Vermont Yankee crud/oxide layer has four features, summarized in Ta6le 1 and in Figure 1.

The Vermont Yankee film morphology was more complex than films from the other two reactors.

In fact, the dual film structure on Millstone and Ouad Cities oipes apnears to be characteristic of films from several BHRs.(5-8)

The Vermont Yankee and Millstone 1 films dif-fered markedly in cnmfyosi tion and morphology, but the decontanina

.io.. rates were very similar, (4) for'example, in an I!EDTA/oxa'1ic acid/citric acid reagent at 180'C.

In both cases, the compact inner film dissolved, while outer layers

remained, detached but undissolved.

Apnroxi-mately 80 to 90>> of the radioactivity was

removed, appearing to correspond to dissolution of the compact inner layer.

15.

Activity removal from Ouad Cities pipe specimens was less effective than for specimens from; the other two reactors.

The (}uad Cities surface film was superimposed on a rough metal surface.

haying fissures up to 0.01 in.(250Ltm) deep.(4,6)

As with the shallower (up to ~20um) fissures on other oipe surfaces, the fissures appear to be a consequence of pre-service pickling.

The hypothesis that radioactive species are trapoed in the fissures is consis-tent with effects of ultrasonic vibration on pipe surfaces after decontamination.

About 28>>

of the residual activity was removed from the Ouad Cities specimen, while the corresnonding fractions for, Mi1,1stone and Vermont Yankee were only 4" and 10", respectively.

Table 1

Summary of BWR Film Characteristics on 304 SS pipe

'eature

~Com ositiona)

A - Vermont Yankee Film Res onse to Reagents b)

Superficial Particles

~l to 10ttm Shroud

<O. lorn thick Faceted Crystals c) 0.1 to lum Compact Oxide Layer up to 2um thick Fe Cr Ni 85-93 3-7 1-2 Unknown 52 18 7

63-76

'12-16 5-8 Cu Zn 2-4 Undissolved Dissolved or Washed Away 1

22 1

9-19 Undissolved - Detached tlashed Away by Turbu'lence Dissolved Princinal source of activity 8 - Millstone 1 Film(7)

Outer Layer Inner Layer C - Ouad Cities 1 Film(6) 86 61 ll 25 t

1 t

3 Undissolved Dissolved Principal Source of activity Outer Layer Inner Layer 90 5

5 68 18 14 t

~ trace NR tt not reported EDX te Energy Dispersive X-ray NR NR Undissolved NR NR Dissolved (Efficiency for radio-activity removal reduced by rouah metal surface)

~a Based on EBX analysis

- slot fully calibrated; percent metallic species notomtited to 100".

b)Based on exposure to EOTA, NTA, HEDTA, etc.

c)Based on analysis of particles detached by reagent treatments.

Paper 54: Johnson er al

~

~

i ~

g'>>

~

)

A

. Prt~

~ ~

p pw~~ia c>'4w~ '

'L

'~Zc', ~~'y 4f a

'cs>

i

~

~qN;A@

gt p'

jn 1 um qr, Fig.

I Cross Section and Surface Views (SEH) - Vermont Yankee BWR Recirculation Bypass Pipe Specimens.

'54~%+

Q<l o'

4QVa +'p

~Mj'fp-ir~ >

OqP gA

.f'r,"

cia' I

\\

I

~ n ~4 <<ii'pe,.,; ~ aga4~,j,i~

jpg~<pg(+'

]Q pm

~c 4

p'w.

I V\\

~

rj Pl. ~

V 3.3 pm "c ~-

~gg~t+Z ~ ~

la4i~a Fig.

2 Cross Section and Surface Views (SEM) - Indian Point 2 PWR Steam Generator Tube Specimens.

2?5

D;Cdsnt prtt'natian d

16.

Crystallographic analyses by X-ray di ffrac-tion suggested that the principal species:

on the; BWR pipe surfaces are Fe203 and spinels of the-type:

Nix Fe3 x04, with radioactive species.sub-stituted at cation positions.

The Fe203 remains a<ter the decontamination, and some spinel species also were undissolved.

Studies by General, Electric Co. (5-7) suggest that the species which dissolve are nickel ferrites (spinels).

The outer layer, which is higher in iron, remains undissolved.

Our experience suggests that the faceted crystals which do not dissolve.are somewhat higher in Cr than the layer which does dissolve, but only mildly so (Table 1).

It is possible that the s'pecies higher in Cr are interspersed with a lower Cr species which dissolves.

If so, the EDX analyses may not properly resolve the lower Cr species.

17.

In sugary, the,BWR films respond favorably to common decontamination reagents (e.q.,

EDTA, HEDTA, tlTA), most effectively when they are re-ducing.

The most important aspect of the decon-tamination appears to be dissolution of the compact layer adjacent to the metal surface.

This layer appears to incorporate a laroe frac-tion of the radioactivity (e.g.,60Co+2),

probably as metal cations which substitute into the spinel oxide lattice.

18.

A relatively small fraction of the radio-activity is associated with species not dis-solved by the decontamination (Fe203 and certain

spinels, probably relatively high in substituted Cr).

However, these detached particles (0.1 to

)OLm) must be dealt with in a decontam1nation, by filtration and by assuring flow conditions which will not deposit the particles at unwanted locations.

INVESTIGATION OF PWR DEPOSITS PWR Primar S stem Nateria)s and Environment 9.

The principa mater>a s in the PWR pr1mary circuit are:

Zircaloy fuel cladding, Inconel-600 steam generator tubing and stainless steel piping, reactor vessel cladding; cobalt-base alloy hard-facing materials.

20.

The PWR primary coolant has additions of boric acid

(~200()

ppm B) and lithuim hydroxide (0.2 to 2.0 'ppm 'Li).

Hydrogen is added to contro) oxygen at <0.0) ppm during operation, While metals continue to oxidize, the system is substantially reducing.

The system temperature fs 320-340'C.

21.

The principal out-of-core material in most PWR primary systems is Inconel-600 steam gener-ator tubing.

Tubing filmed 1n two PWR steam generators has been

examined, by methods indi-cated for BI/R fi)ms.

Noroho)o ies and Com ositions of PWR Films 22, The surface of a PWR steam generator film had the following characteristics (Fig. 2):

a sparse population of large superficial partfc)es; a layer of fine platlets; a network of dark grain boundaries.

Table 2 surnwrizes compositions of two PWR films.

23. 'ig.

2 also shows a cross'ection of a crud/oxide layer.

The layer thickness is ~0.5um.

Thin filaments of the layer protrude into the metal along what vie shall see are grain boundaries.

24.

Confirmation that the f1lm is keyinq a)onn grain boundaries developed from treatments which separated the film from the base metal.

Figure 3 shows separated film flakes.

Location A is a

f)ake oriented to show the side adjacent to the metal.

The filaments embedded in the metal are now evident around the grain boundary contours.

Location B shows a flake oriented to show the coolant side.

25.

Responses of the PWR films to reagent treatments are still being investigated.

To,

date, the following responses have been observed:

Decontamination occurs slowly in the common reagents (EDTA, HEDTA, NTA)(4)

Addit1on of H202 to EDTA caused only temporary increases in decontamination rate(4)

Addition of reducing metal ions (e.gss Cr+2) caused Iapid increases in decontami-nation rate(4>

Certain oxidizing and reducinn reagents disso)ved the base metal, undercutting the film, remov1ng it almost completely as flake-like par ticles.

OISCUSSIOll 26.

Oor stvdies have clearly demonstrated that BIN and PWR films respond differently to common reagents such as EDTA, HEDTA and NTA.

The BWR films are easier to decontaminate.

Early in the studies we began to examine the questions:

Do BWR films, formed under oxidizing conditions, respond most effectively to reducing decontamination reagents?

Oo PWR films, formed under "reducing" conditions, respond most effectively to oxidizing reagents?

27.

Responses of BWR films to decontamination were in line with the hypothesis:

oxidizing reaqents were not effective; reagent effective-ness increased with increasing reducing character.

28.

Beaker studies also suggested that reduc-tion processes promoted rapid decontamination.

BWR pipe specimens which included freshly-cut metal surfaces had higher decontamination rates than specimens where

$bg exoosed metal was coated with si)icone.<">

The observation sugqested that corrosion of the fresh metal produced electrons (e.qss Fe' Fe+

+ 2e) which were available for reduction of species in the radioactive film.

29.

The PWR films showed some mildly favorable responses to oxidizing reaqents, aareeina with the hypothesis proposed above.

However, the response was not dramatic.

In fact, the PWR films also have shown favorable resnonses to reducinn conditions.

(4) 27fi

)

Vapor 54:

Johnson ct al s

~I

~rl'

~

A

~ 0

~

m 3000X

Fig, 3A Crud/Oxide Flake Separated From Inconel-600 Steam Generator Tube IP2-3-7 During Decontam-ination Treatment at 180'C 2000X Fig.

38 Crud/Oxide Flake from IP2-3-7; Oriented to Show Side Exposed to Reactor Coolant Table 2 - Sulunary of PWR Film Characteristics on Inconel-600 Steam Generator Tubing Feature

~tom osltfoo Remarks Fe Cr Ni Si Zn A - Tube IP2 Superficial Particles 2 to 9Mm Compact Layer in a~

40-86 4-15 7-40

'2-5 t

23 63 1

2 55 18 10 3

40 19 17 Compact Layer detached 14 Compact Layer detached 9

8 - Tube I-4-30HL Superficial Particles 2Mm Compact Layer in a~

Compact Layer detached 35 Compact Layer detached 24 29 31 4

2 46 28 3

Film Removed by Scraping 29-34 25-27 36-38 3-7 1-2 18-31 18-26 40-60 2-4 0-2 31 26 40 2

2 See Fig. 2,.right-hand views Analysis of cross section, near film outer edge; base metal shine probable Separated from base metal by HEPjA/

OXALIC acid reagent, 120-150'C o

Separated from base metal by HEDTA acid reagent, 180'C ~)

Film up to 3Mm thick; minimal base metal shine Film flakes separated from base metal by EDTA/H202 reagent at 180'C b)

Film -flakes separated from base metal in methanol-Bromine at 25'C Some base metal possible-appears minimal normalized to 100::.

Reagents also include boric acid and lithium hydroxide, characteristic of PWR system chemistries.

277

Dccont urination 30.

On the hypothesis that the radioactive species are largely present at substitutional positions in the compact oxide, removal of the radioactivity appears to require partial or complete dissolution of the oxide, depending on

.-" - - the degree of activity reduction required.<< On--

this hypothesis, it does not appear possible'o remove the radioactive species (e.g.,

60Co) by selective dissolution even though higher ratios of 60Co to removed film were observed.

This conclusion also was reached by other investi-gators.<4) 31.

It is important to minimize post-decontami-nation activity buildup associated with corrosion of the decontaminated surfaces.

Our studies on oxidized laboratory specimens suggest that the post-reagent film growth at reactor temperatures was proportional to the amount qf oxide dissolved during the reagent treatment.L9>

This suggests that modest OFs'ay leave some protective oxide, minimizing subsequent corrosion.

This supposition has not been confirmed on reactor specimens in our studies.

32.

The BWR films comprise Fe203 and a range of

spinels, some soluble in cowmen reagents; some insoluble.

The particulate species which remain undissolved after reagent treatments include Fe203 and certain spinels, apparently those with relatively high Cr contents

(~18">>).

However, the major fraction of the radioactivity is soluble and therefore may be removed by ion

exchange, while particulates can be removed on filters.

This technology has been demonstrated in dilute reagent Jecontamination performed in Canadian reactors. Ill) 33.

The PWg films comprise a variety of spinels and oxides.<>>>

Our analyses indicate that the sparse population of superficial particles tend to be high in Fe, but have a wide range of compositions (Table 2).

The radioactive oxide layer is sometimes less than one micron thick.

The relative insolubility in comon reagents probably relates to the high Cr content (30-55", Table 2).

However, we have been able to achieve relatively effl;ctive decontaminations by metal ion reduction,<4> or by both oxidizing and reducing reagents which undercut the film, resulting'in a large fraction of the activity being removed as particulates.

OF Radiation Level Before Decontamination adiat on Leve fter Decontamination REFERENCES 1,

JOHNSON A B

~ Jr.

~

GRIGGS 8 ~

~

REMARK J.F

~

Investigation of Chemicals and Methods of Dilute Reagent Decontamination for Potential Application in Light Water Reactors.

Paper 32 CORROSION/78, Houston, TX, March 6-10, 1978.

Available from NACE, P.O.

Box 218340,

Houston, TX 77218.

2.

JOHNSON, A.B., Jr.,

GRIGGS B.. Comparison of Decontamination Characteristics of Films from BWR and PWR Primary Systems.

Paper

161, CORROSION/79, Atlanta, GA, March 12-16, 1979.

Available from NACE, (see above).

3.

JOHNSON, A.B., Jr.,

DILLON, R.L., GRIGGS, B,

REMARK, J.F.,

Investigation of Alternate Methods of Chemical Decontamination--Second Progress Report.

January 1,

1977 to December-31,,

1977, BN-SA-703-2, Battelle Northwest, June 1978.

~

~ ~

0 W

I~ I'f h 0 P

~ 8 1

I ~

~ P g1 IE 4.

JOHNSON, A.B., Jr.,

GRIGGS, B., DILLON, R.L.

Candidate Reagents for Activity Reduction in BWR and PWR Primary Systems.

Proceedinqs ANS Topical Meeting on Decontamination and Decormis-

sioning, Sun Valley, ID. September 16-20,1979.

5.

BLOK, J.

Summary Report Water Chemistry Program Extension.

NEOC-21550, General Electric

Company, San Jose, CA, June 1977.

6.

ANSTINE. L.O.

BWR Decontamination and Corrosion Product Characterization.

NEDE-12665.

March 1977.

7.'haracterization of Corrosion Products on Recirculation and Byoass Lines at riiilstone l.

NP-949, Electric Power Research Institute, Project 819-1, December 1978.

8.

IWAHORI, T., MIZUNO T., and KOYAMA, H. Crud Composition on Heat Transfer Surface, Paper 37, CORROSIOIl/78, Houston, TX, available from NACE (see ref.

1 above).

9.

A Study of Dilute Reagent Decontamination for Application in Boiling Water Reactors, NP-Electric Power Research Institute, Project 828-1, (to be published),

10.

LeSURF, J.E.

Water Chemistry of Nuclear Reactor Systems.

Proceedings of BNES Conference, Bournemouth, October 24-27,

1977,

p. 426.

ll.

PETTIT, P.J.,

LeSURF, J.E.;

STEWART, W.B.,

STRICKERT, R.J.,

and VAUGHAN, S.B.,

Decontami-nation of the Douqlas Point Reactor by the CAN-DECON Process.

Paper 39, CORROSION/78, Ma@h 1978 (see Ref. 1).

12.

SHEE, J.L.

Dissolution Characteristics of Metal Oxides in Water Cooled Reactors.

Proceedings ANS Topical Meeting on Decontami-nation and Decommissioninq, Sun Valley, ID.

September 16-20, 1979.

ACKNOWLEDGMENTS The following have made siqnificant contributions to this study:

R.L. Dillon, J.L. Daniel, J.E.

Coleman.

D.B. Mackey, and R.L. RcDowell.

We are grateful to the Electric Power Research Institute (EPRI) for permission to publish the results of this study.

278