first chapter - zirat12 str corrosion mechanisms

7/27/2019 First Chapter - ZIRAT12 STR Corrosion Mechanisms

http://slidepdf.com/reader/full/first-chapter-zirat12-str-corrosion-mechanisms 1/45

ZIRAr 2 PECIAL TOPIC EPOR ORROSION ECHANISMS IN ZIRCONIUM LLOYS 2007

Corrosion Mechanisms

in Zirconium Alloys

Ron AdamsonZircology Plus, Fremont, CA, USA

Fredrch GarzarollErlangen, Germany

Bran CoxUniversity of Toronto, ON, Canada

Alfred StrasserAquarius Services, Sleepy Hollow, NY, USA

Peter RudlngANT International, Skultuna, Sweden

Ron AdamsonZircology Plus, Fremont, CA, USA

October 7

Advanced Nuclear Technology nternational

Krongjutarvgen , SE-73 5 Skultuna

Sweden

[email protected]

www.antinternational.com



ZTr 2 PECIAL TOPIC EPOR 2007 CORROSION MECHANISMS IN ZIRCONIUM LLOYS

DThe information presented in this report has been compiled and analysed by

Advanced Nuclear Technology nternational Europe AB ( nternational®)

and its subcontractors. nternational has exercised due diligence in this work,

but does not warrant the accuracy or completeness of the information.

nternational does not assume any responsibility for any consequences

as a result of the use of the information for any party, except a warranty

for reasonable technical skill, which is limited to the amount paid for this

assignment by each programme member.

Copyright © Advanced Nuclear Technology Inteational Eupe AB, ANT Inteational, . This information is the prpey of Advanced Nuclear Technology Inteational Eurpe AB or is licensed for use by Advanced Nuclear Technology Inteational Eurpe AB by s

customers or paners. The information may not be given to, shared with, or ced to third pay, used for unautho sed purpose, or be copied or reprduced in any form whout the written permission of Advanced Nuclear Technology Inteational Eurpe AB.

(V)




Ay ExANTAOAASTMATRBO BRANLRPXESELSEOESSF

GBGEGNFHBRHFHPAHPFHAEAKKLLHGRLKLL

LRNMANSSSNOPEPTPRRBMK

RTRXASEMSGHRSHESMSSPPSRASSSTEMSTRTEMTMTM

Advanced Nuclear TechnologyAxial Offset AnomalyAmerican Society for Testing and Materia lsAdvanced Test ReactorBeginning of ycleBoiling ater Reactoranadian euterium raniumrud nduced Localized orrosionhalk River nidentified epositsistinctive rud PatternuplexEmergency ore ooling SystemExtra -Low SnEnd Of ycleEnhanced Spacer Shadow orrosionFuel uty ndex

Grain BoundariesGeneral ElectricGlobal Nuclear FuelHalden BRHigh FrequencyHigh Performance AlloyHydrogen Pickp FractionHydrogen ater hemistrynternational Atomic Energy AgencyKernKraftwerk LeibstadtLinear Heat Generation RateLg corrosion (Low orrosion in Swedish)Low Leakage

Light ater ReactorNoble Metal hemical AdditionNuclear Steam Supply SystemNormal ater hemistryOut-nPost-rradiation ExaminationsPressure TubePressurised ater ReactorReaktor Bolshoi Mozhnosti Kanalov ( in English Large Boiling ater hannel typereactor)Room TemperatureRecrystallised AnnealedScanning Electron MicroscopySteam Generating Heavy ater ReactorStandard Hydrogen ElectrodeSecondary on Mass SpectroscopySecond Phase ParticleStress Relieved AnnealedStainless SteelScanning Transmission Electron MicroscopySpecial Topic ReportTransmission Electron MicroscopyTransition MetalThree Mile sland



(V)




TSSTZPVVERYSZZRATZRLO

Terminal Solid SolubilityTetragonal Zirconia PolycrystalVoda Voda Energo Reactor (Russian type W)Y ttria Sta bilised ZirconiaZirconium Alloy TechnologyZirconium Low Oxidation



(V)




TEETE TCEC 27315 K

°C* + x

x. 6 .2

T(K) T(C) T() 1 .

27 3 0 32 5 .2

289 1 6 6 1 10 .39

298 25 77 2 .79

37 3 100 21 2 25 .98

7 3 200 392 25. 10057 3 300 572 100 3.9

633 36 68

673 400 75 277 3 500 93 2

783 51 95

793 52 968

823 55 122 PRESSURE

833 56 1 bar MPa psi

87 3 00 1 1 1 2 1 . 1 1

878 65 1 1 2 1 1 1 1 2

893 62 1 1 8 7 7 995

923 65 122 7. 7. 100097 3 700 1292 100 1 121

123 75 1 382 1 3 1 3 187153 78 136 155 1 5 . 5 223

173 00 172 7 7. 100001 1 3 6 863 1585 1000 1 1 211

1 1 3 87 1598

1 1 7 3 900 1652

1273 1000 1832

133 17 1958

178 12 2200STRESS INTENSITY ACTOR

MASS MPa m ksi inchkg Ibs .91 1

.5 1 1 1 . 1

1 2.2

Radioactivity

1 Sv 100 Rem1 Ci 37 x 1010 Bq 37 GBq1 Bq 1

Copyright © Advanced Nuclear Technology Inteational Europe AB, ANT Inteationa This information is the proper of Advanced Nuclear Technology Inteational Europe AB or is licensed for use by Advanced Nuclear Technology Inteational Europe AB by its

customers or parers. The information may not be given to, shared with, or cited to third pa used for unauthorised purpose, or be copied or reproduced in any form without the written permission of Advanced Nuclear Technology Inteational Europe AB.

V(V)




Acronyms and Explanations

Unit Conversion

Contents

1.11.1.1

.1.·3··5.6

·7.8·92.10

2.11

2.12

3·1.13·1.3·1..1

3·1..3 ·1.33. 3·33·

4.1

4·1.1

4·1.2

4.2

4.2.1

4.2 .2

..3·34.3.1

4.3 .2

333·4.4.1

4.4.2

4+3

Introduction (Ron Adamson)

Effect of temperature (Alfred Strasser and Peter Rudling)ntroduction

Basics of the corrosion and hydrogen pick-up process (Brian Cox)

nitial zirconium surface conditionGalvanic processesRate-determining step for zirconium oxidationStructural growth of zirconia filmsOxide morphology barrier/non-barrier oxide filmsThe oxidation rate transition and its causes

The contribution of H20 to oxide film breakdownLocation of alloying elements and impurities in the metal and oxideHydrogen uptake mechanismEffects of irradiation on corrosion mechanismsNodular corrosion in WsSimulation of nodular corrosion

Key parameters that affect corrosion mechanisms and rate

Material parameters that a ffect corrosion mechanisms and rate(Friedrich Garzarolli)mpact of Material Parameters on orrosionMechanistic aspects of the impact of material parameters on corrosionniform corrosion

Nodular orrosionmpact of material parameters on hydrogen pick-upHydride impact on corrosion (Friedrich Garzarolli)ater chemistry impact on corrosionW Boiling mpact on orrosion

Description of different forms of corrosion (Friedrich Garzarolli)

Accelerated uniform corrosion (Ron Adamson)BackgroundMechanism implicationsShadow corrosion (Ron Adamson)Key overall observations

MechanismsmplicationsEnhanced spacer shadow corrosion (C) (Ron Adamson)ater chemistryMaterialPerformanceMechanismCUD-related failures (Alfred Strasser)ntroductionW CUDW CUDCUD-induced localized corrosion (Ron Adamson)

V

V

-1-1- -3-5-7

-1-11-17--1-1

3-13-3-13 -13-35

3-3-3-73-55

-5-5-1-14-15

-164-19

4-19

4-19

4-19

-4-23

-6-64-27

4-29

4-30

Copyright © Advanced Nuclear Technology Inteational Eupe AB, ANT Inteational, . This information is t he prpey of Advanced Nuclear Technology Inteational Eurpe AB or is licensed for use by Advanced Nuclear Technology Inteational Eurpe AB by s


V(V)




5 Non-classical CC failures (Peter Rudling) -33

Open items (Friedrich Garzarolli) Discussions, implications and summary (Ron Adamson) 6.1 ntroduction (Section I 6-16. Basics of corrosion and hydrogen pick-up (Section ) 6-36..1 Mechanisms 6-36.. Testing 6-36.·3 Key parameters that affect corrosion mechanisms and rate (Section 3) 6-4

6.· escription of different forms of corrosion (Section ) 6-8

References



V(V)




The emphasis of this Special Topical Report () is on the mechanisms of corrosion ofzirconium alloys used in the core of nuclear power plants. The entire area of corrosion (and theaccompanying absorption of hydrogen in the zirconium metal matrix) is of prime interest when

considering performance of the core components and therefore the performance of the entirereactor. For instance, for Pressurised ater Reactors (Ws) a practical corrosion limit exists(a bout 1 m oxide thickness which is ass ociated with a critical amount of hydrogen absorption)that has driven a material change away from Zircalo-.

The technical literature and many conferences are full of papers dealing with corrosion issues .of particular interest is the series "Zirconium in the Nuclear ndustry: nternational Symposium, STP, American Society o f Testing and Materials, which provides most relevant details.Also reviews of the topic are given in Adamson et aI., ("orrosion of Zirconium Alloys,Zirconium Alloy Technology (7) Special Topics Report, ecember, ) and "atersidecorrosion of zirconium alloys in nuclear power plants , nternational Atomic Energy Agency()-TEO-996, January 1998. The current deals only lightly with general corrosionphenomena, with the aim being to concentrate on corrosion mechanisms. The most recent openliterature review of mechanisms is ox, 5.

orrosion of zirconium alloys i s an electrochemically-driven process affected by themicrostructure and microchemistry of the alloy surface, the nature of the oxide layer that forms,the temperature at the metal/oxide interface, the chemistry and thermohydraulics of the corrodentwater, the effects of irradiation and the effects of time.Specifics of these processes are given laterin this STP; some genera l introduction is given here.

Table I- I gives information on the various types of commercial power reactor systemscurrently being used throughout the world. n comparing Boiling ater Reactor (Ws) withWs, with corrosion mechanisms in mind, the main features are:

• W coolant boils; W coolant does not. This has an important effect at the oxide/waterinterface.

• W coolant contains a high concentration of hydrogen; W coolant does not.

omplementarily, W coolant contains a high concentration of oxygen, W coolant doesnot. This has an important effect on corrosion processes.

• W components generally operate at higher temperatures than W components.orrosion processes are temperature dependent.

• Both reactor types employ chemical additions to the coolant which may effect corrosion andbuild-up of deposits on fuel rods.

Copyright © Advanced Nuclear Technology Inteational Eupe AB, ANT Inteational, . This information i s the prpey of Advanced Nuclear Technology Inteational Eurpe AB or is licensed for use by Advanced Nuclear Technology Inteational Eurpe AB by s


1-1(1-19)



r

ZTr 2 PECIAL TOPIC EPOR 2007 - CORROSION MECHANISMS IN ZIRCONIUM LLOYS

Table 1 - 1 : Design Parameters for water cooled reactors, Adamson et aI., 22.

Parameter Western type B (440/1000) MW

1. Coolant Pressurized Pressurized Pressurized Boiling

HO HO HO. Fuel Materials Zry-4, Zr-alloy E Zry-4 Zry-, Zry-4,

(Pressure tube materials) DUPEX, M5, (Zr.5Nb) Inconel, Inconel,

. verage power rating, (k/I) 8-15 8/18 9-19 4-57

4. Fast Neutron Flux,verage, n/cm.s -9E1 5E17E1 1.5-E1 4-7E1

5. Temperatures, °C

verage Coolant inle t 79-94 7/9 49-57 7-78

verage Coolant outlet 1-9 98/ 9-5 8-

Max Cladding -5 5/5 1 85-5Steam mass conten % 7-14

. System pressure, bar 155-158 15/15 9 7

7. Coolant Flow, ms -' .5/ -5 -5'

8. Coolant Chem istry

Oxygen, ppb <.5 <.1 -4

Hydrogen (),

ppm -4 (-1) .5-.

cc/kg 5-5 -

Boron (as Boric acid), ppm - -14

i (as iOH), ppm .5-.5 .5-. 1

K (as KOH), ppm 5-

NH, ppm -

NaOH, ppm .-.5

Canadian Deuterium Uranium (CNDU), Reaktor Bolshoi Mozhnosti Kanalov (RBMK), Stainless Steel (SS),Voda Voda Energo Reactor (VVER), Zirconium ow Oxidation (

-

BMJ

Boiling

HOZr-alloy E,

(Zr.5Nb)

5

1-E1

7

84

914

7

.7

<

t also sho uld be noted that W zirconium alloys continue to be primarily Zircalo- or slightvariants o f Zircalo-. W zirconium alloys no longer tend to be Zircalo-, for reasons ofinsufficient corrosion resistance (an d hydriding resistance) at high burn-up, but have moved

toward zirconium alloys with Nb additions.

, Variation fro m lower to upper part of the core an d from plant to plant., Variation from lower to upper part of the core and from plant to plant.'" Zn in ppb quantities may be added for BWRs and PWRs; t and h in ppb quantities may be added forBWRs.





I


The type of oxides which form during corrosion in reactor water can be classified into severalcategories. The two most basic are uniform and nodular corrosion. The "uniform category hasan extension "patch or accelerated uniform. The fourth category is " shadow corrosion , whichcan look like thick uniform corrosion but has some characteristics of nodular corrosion. The fifthcategory is crud- related corrosion, which is a temperature driven process induced by poor heattransfer in crud-impregnated corrosion layers. These categories will be discussed later, but will be

introduced here. Table 1- (Garzarolli in Adamson et aI., ) gives a useful summary ofcharacteristics of various corrosion types.

Table 1 -2 : Types of corrosion observed for Zircaloy in-reactor and out-of-reactor, Garzarolli in ZR R "Corrosion in Zirconium Alloys, Adamson et aI., 22.

Irradiation effect i n Irradiation effect i nye of corrosion omment bserved oygenated coolant ydrogenated coolant

Out-oF-pile, in-8WR and,5-10x increased From 2-4x increased after

UniForm Normal mode beginning, low Flux 1 corrosion rate transition,in-PWR

dependency low Flux dependency

Local break down of In-8WRand out-oF-pile Increases almost linearNodular

oxide protectivness>500'C, Zry with large with Fast Fl ux, little temp. not obseedSPP dependency

robably driven byOnly under irradiation

Increases probably almosthadow corrosion and oxidative coolant not obseed

potential differencesconditions (8W�

linear with Fast Flux

Crevice corrosionChange of environment Out-oF-pile, in-8WR and,

obseed Obseedin small gaps in-PWR

Increased corrosion Out-oF-pile, in-8WR and, Increases almost linearIncreases almost linear

eduction of protectivity with Fast Flux, little temp. with Fine SPP in-PWR with Fast Flux

dependency

Increases withIncreased corrosion

eduction of protectivity In-PWR ? increasing Fluence

at high Fluences above a criticalthreshold

Increased corrosionLower corrosion Out-oF-pile, in-8WR and,

lux dependent butat high hydride

resistance of hydride in-PWRobseed less temperature

concentrations dependent

niform corrosion occurs in both Ws and Ws. The oxide itself is uniform in thickness andconsists of several different layers, discussed in Section . For either in or out of reactor, the initialshape of the corrosion-versus -time curve is as shown in Figure I-I in the pre-transition region.The transition point occurs at around 5 m oxide thickness in Ws. The shape of the posttransition curve in Ws depends on several variables: initial Second Phase Particle () size,irradiation, amount of cold work, specific alloy, water chemistry, temperature, local

thermohydraulics, hydride concentration all to be discus sed in Section For Zircalo- theooson hkness a hgh bun-up may eah o exeed m, whle some of he newe alloyshave less than half of that. As noted in Table 1-, only uniform corrosion occurs in Ws undernon-boiling, hydrogenated conditions.





Z Tr 2 PECIAL TOPIC EPOR 2007 CORROSION MECHANISMS IN ZIRCONIUM LLOYS

The study of uniform corrosion in the laboratory at temperatures relevant to reactor operationsuffers from three problems: I irradiation effects are absent, ) corrosion rates are very low at8 (553K) - 36 ( 633K), 3) very long times (hundreds of days) a re required to differentiatebetween material variables. The most widely used test appear s to be 36 ( 63 3K) pressurizedwater for very long times. An example is given in Figure 1-, where a series o f Zircaloy-type alloyswith different Sn contents are compared. Garde et aI., 1994, show that in-reactor results after high

burn-up give the same ranking as the autoclave laboratory tests. Also for Ws, adding Li to thewater is thought to improve comparisons, Sabol et aI., 1994

Figure 1-1:

Figure 1-:

6

40

:

5 Clasicl

! nnQ

-

1

I

2

! A 7

Second

el

./ lsthermal nPWR34·C

4 6

Exposure time [d)

Iohrml outf-p 4'C

B

Schematic of Wcorrosion kinetics, Garzarolli et aI., 199

(mN0

;00

1F · b R

6°K Aocv9 • #

B + #2

• #

0.7. #4

6 • #5

EPRISTD5

4

0.3

2

1

1 2 0 . 4 5 6 7 B 1 11 2 1 14

m, days (thsands

Corrosion weight gain as a function of laboratory autoclave test exposure time for several Zircaloy-4 tubevariants, Garde et aI., 1994

Both Zircalo- and -4 form a protective uniform oxide in typical Ws and Ws. n Ws thevarious types of oxidation kinetic are shown in Figure 1-3.






Figure 1-:

F [j

6

Characteristics of diferent types of corrosion observed in W. Oxide layer thickness in arbitrary units versusfuel assembly burnu p, A "Corrosion of Zirco nium ll oys.

niform corrosion continues at a very low rate out to high burn-ups. uring reactor exposure themicrostructure of the Zircaloys used in Ws is continually evolving due to irradiation damageand dissolution. n the range 3-5 Md/kg ( 6-x

2 n/cm

2, E> MeV), changes in the

microstructure induce an acceleration of uniform corrosion, as described in Section . First,patches of white oxide appear in the otherwise black or grey uniform background, a s illustrated inFigure -. These patches remain very thin, about the same

Figure 1-4:

54

3

Patch oxide formation on Zircaloy- after an exposure of 8.5x1 n!cm, E> 1 MeV. The oxide is thinnest atuniform black oxide 5 and patch oxide 4 and is thickest at coalesced patches , Huang et aI., 199.

as the black uniform corrosion film. At some point the patches cover % of the surface andoxide thickening occurs at an accelerated rate, labeled "late increased corrosion in Figure -3.

mportantly, at the same or somewhat higher fluences, a marked increase in hydrogen pick-upfraction occurs. This hydriding is potentially a more serious issue than the corrosion increase .

Copyright © Advanced Nuclear Technology Inteational Eupe AB, ANT Inteational, . This information i s the prpey o f Advanced Nuclear Technology Inteational Eurpe AB or is licensed for use by Advanced Nuclear Technology Inteational Eurpe AB by s





Figure 1-5 gives optical micrographs and schematics of various types o f corrosion. Schematics Aand show normal and increased uniform oxide, and the others illustrate nodular corrosion. ZrNb alloys and small- Zircaloys (average size less than about .1 m) in general do not

Figure 1-5:

Te sace wth les

Corrosion morphology for Zircaloy in W.

Metallc ssse

ZO2 0

Z ,

N

-ur

d .

form nodules. n large- Zircaloys, nodules initiate early in life and grow at a decreasing rate

with fluence. n some cases, for aggressive water chemistries and susceptible material, nodules cancoalesce to cover the entire surface. Nodular oxide thickness does not generally cause performanceproblems; however in severe cases spalled oxide can be a source of "grit in control drivemechanisms. Serious fuel failure problems can be induced by a combination of heavy nodularcorrosion and copper-zinc-laden crud, resulting in the halk River nidentified epo sits ( CUD)nduced Localized orrosion L phenomena discussed in Section .

Susceptibility of Zircaloys to in-reactor nodular corrosion can be identified by laboratory hightemperature steam tests. The most e ffective testing procedures are variations of the "two-steptest described by heng et aI., 1987 where Zircaloys are exposed to steam at 15 psi(683K bars) for 8h followed by 5 psi (783K bars) for 16h. n such tests, Zr-Nballoys do not exhibit nodular corrosion.

The fifth type of corrosion, indica ted in Figure 1 -3, is the so-called shadow corrosion,described in more detail in Section . Shadow corrosion is induced on all zirconium alloys when

they are in close proximity to many non-zirconium alloys such as stainless steel or ncone. Theoxde hkness s unusually lage and ofen appeas o be paulaly dense and un-aked Foexample shadow corrosion oxide induced by a stainless steel control blade bundle is shown inFigure 1-6. Shadow corrosion has " always been present in Ws, but not in Ws, primarily



1-6(1-19)




Figure 1-:

Qo

3

(b)

Zirconi um oxides near (b) and away from (a) a stainless steel control blade bund le, damson et aI., .

related to the high W hydrogen concentration which reduces or eliminates galvanic potentialsbetween dissimilar alloy components. n Ws shadow corrosion caused no performance issuesuntil recently when at one reactor fuel failures were induced by unusually severe "enhanced spacershadow corrosion, Zwicky et aI., . More recently, shadow corrosion has been alleged to beinvolved in W channel bow problems, Mahmood et aI., 7. Both issues are addressed inSection 4

The following sections of this will address i n more detail the mechanisms which arebelieved to be most important in describing corrosion in zirconium alloys.

Eect of teperature (Alfred Strasser andPeter Rudling)

Temperature is one of the most important parameters affecting zirconium alloy claddingcorrosion. Yet it is a parameter that can not be measured directly --- it can only be calculated.ncertainties in the calculations, or modeling, are the results of uncertainties contributed bynumerous other parameters that are necessary components of the models, some o f which areequally difficult to measure. Knoledge of the parameters that affect the temperature and an

appreciation of the uncertainties related to each parameter are important in the evaluation of therelationship of temperature to in-reactor corrosion.Temperature measurements of samples in ex-reactor tests or some test-reactor tests

(particularly of unfueled samples) can be made reasonably reliably. The objective of this section isto identify the various parameters that affect the temperature of cladding during in-reactor service,note their degree of importance and provide the reader ith some feeling of the uncertaintiesinvolved.






Ex-reactor corrosion tests of zirconium alloys with initially clean surfaces provide the data thatshow the dramatic increase of corrosion with temperature. n an example of work by Bettis Lab s.shown in Figure 1-7, a 6% increase in weight gai n was produced a s a result of a 5 increase intemperature in a nearly 7-year-long test at a typical cladding temperature of 335. The "typicalmaximum cladding temperatures are summarised in Table 1-3 for Ws and Ws along withthe reactor inlet and outlet coolant temperature ranges. The cladding temperatures are calculated

numbers for clean surface cladding at nominal reactor operating conditions.

Table 1 -3 :

Figure 1-7:

ight Water Reactor (L Coolant and Cladding Temperatures, Nuclear Engineering Intemationa1998.

Coolant Inlet Range

°C

o

Coolant Outlet Range°C

o

Max. Clad Temperature

°C

o

8

E �

272-278

522-532

28-3

536-572

31-32

59-68

a s as wa

': 6 c"

279-29

53-55

313-33

595-626

318-36

6-68

440

267

51 3

296-3

565-572

335-35

635-653

· 8 3 3 3 33 3 3

Zircaloy- and 4 weight gain as a function of temperature, Hillner et aI., .

Copyright © Advanced Nuclear Technology Inteational Eupe AB ANT Inteational . This information is the prpey of Advanced Nuclear Technology Inteational Eurpe AB or is licensed for use by Advanced Nuclear Technology Inteational Eurpe AB by s

customers or paners. The information may not be given to shared with or ced to third pay, used for unautho sed purpose or be copied or reprduced in any form whout the written permission of Advanced Nuclear Technology Inteational Eurpe AB.

1-8(1-19)




The lower pressure and the boiling regime of Ws limit their coolant temperatures and inlet-tooutlet temperature ranges compared to Ws. Greater flexibility in Ws also permitted them toincrease coolant outlet temperatures (and peak cladding temperatures as a result) for the moremodern units in their race to increase efficiency and lower the predicted fuel cycle costs. t isinteresting to note that, due to the high corrosion rate o f cladding at the high temperatures, manyof these newer Ws have modified their system to reduce temperatures and lengthen the life of

their fuel cladding.The Russian 1 units fall within the temperature range of the western Ws, but the

units are lower in temperature and are shown separately for that reason.A brief overview of the relationship of W coolant temperatures and cladding corrosion is

given in Table 1-. At an exposure of 5-8 Gd the oxide thickness on Zircalo- claddingsrange 11-37 m in the lowest-coolant-temperature early W ( Ginna) to 78-1 m in one of themore recent, high-temperature plants (Gsgen), by a factor of about . The other plants cited arein a range of coolant temperatures and oxidation levels between these two extremes.

Table 1-4: Corrosion of Standard Zircaloy-4 as a Function of Plant Parameters, -TECDOC-996, 1998.

IPlant

Ginna

ANO-2

Grohnde

Summer

Noh AnnaGsgen

Ave Core Ave Core

kw/lit Outlet °C ()

89 317 602

97 323 613

95 327 602105 327 620109 327

620

101 325 618

Oxide Thickness, BurnupGWd/mtU

m Rod Assy

11-37 45-48 42.5

20-41 54-56 52

58 ave. 52

21-73 57-60

40-80 60-64

82-92 45-46

76-86 46-48 46

52-96 44-47 46.478-120 40-48

The temperature at the interface beteen the zirconium alloy metal surface and the zirconiumoxide formed governs the post-transition corrosion rate and can be formulated as:

where

is the oxide thicknesst is the time

C exp( T)

is the post-transition activation energyC is an irradiation corrosion enhancement factor is the Gas onstantT is the temperature at the metal/oxide interface





Z Tr 2 PECIAL TOPIC EPOR 2007 CORROSION MECHANISMS IN ZIRCONIUM LLOYS

The Cconstant is very different for zirconium alloys in Ws and Ws. Experiments (see e.g.Stehle et aI., 198) comparing the out-of-pile with the W corrosion rates have suggested avalue of C= t5, while similar tests for Ws have indicated a Cconstant that varies withtemperature. At temperatures of °, C= 1, while at ° the C=. Thus, the out-of-pileand in-pile corrosion rates for Zircaloy can be schematically shown as in Figure 1-8. The bottomline is that a temperature increase at the metal/oxide interface belo a maximum temperature of

400°C results in a larger increase of the corrosion rate in a PWR than in a BWR. This is thereason that there is a licensing limit on the maximum oxide thicknes for fuel rods in PWRs butnot in WRs. Since the oxide has a loer thermal conductivity than that of the metal a largeoxide thickness increase may increase the metal/oxide temperature and thereby increase thecorrosion rate in a PWR leading to a thermal feedback effect.

Figure 1-8:

p

\\\\P WR

\\\

\ - R

\\\,\

Schematic diagram showing the impact of metal/oxide inteace temperature on post-transition corrosion rate.

Figures 1-9 and 1-1 depict typical PWR fuel clad temperatures and the resulting fuel rod axialoxide thickness profile. Higher corrosion rates occur in the upper part of the fuel rod, due tohigher temperatures at the oxide/metal interface. The fuel ro d oxide thickness suppression at thegrid locations is due to lower fuel clad temperatures at these locations. This is caused by:

) increased turbulence due to the spacer res ulting in better cooling and) lower reactivity due to para sitic thermal neutron ab sorption of the grid material

However, it i s important to note that the PWR corrosion acceleration, due to thermal feedbackeffect, occurs at a much larger oxide thickness (about 1 m) than the corrosion acceleration dueto hydride formation, see Figure -.



1-1(1-19)




Figure 1-9:

Figure 1-1:

360

0

w

Sb-coold bog�

A . " , " ' }' ," I ,

Typca sfac / i!

ta

: :' : e 320

, : " .'

�,

300

/�

i

!TYPca v factoII

/1 \----�-----�----- OC hght ]

Schematics showing typical Wfuel clad suace temperature as well as bulk coolant temperature as afunction of fuel rod elevation. In the upper part of the fuel rod, the coolant temperature may locally at the fuel rodsurface exceed the saturation temperature leading to local boili ng (called "sub-cooled boili ng since the bu lkcoolant temperature is below the saturation temperature).

6

i

�

Z

N

Win-pile oxide profile. Modified figure according to Sabol et aI., 1994.

6



1-11(1-19)




Figure 1-11:

E QC� :Q>� 40Q·x� 20r

00

DXb* DX DXtb b

COITO. iouacceratn e 10hyde fOl

40

+ +

Fl r brnp /kgUSiemens Wdata. Modified figure according to Seibold et aI., .

n Ws the coolant and fuel rod temperature axial profiles are much more constant than that ofthe Ws. The more constant axial temperature in the Ws is due to the boiling that occurs atabout the lowest spacer location. The fuel rod oxide profile is NOT correlated to the fuel rodtemperature (except in those case s when the fuel rod is overheated - temperatures ranging from and upwards - due to a thermally insulating CUD layer) , see Section

Figure 1-1:

E

:aCisv=

G

4

B _

i i;

W In-p ile Oxide Profile, Garzarolli et aI., 1979.

'\

The variables that determine the temperature at the metal/oxide interface are shown pictorially inFigure 1-13 and are listed in Table 1-5. As shown on the figure, the interface temperature inquestion is determined by the power or heat flux (Q) generated in the fuel rod and the efficiencyof the coolant in removing this heat. The thermal resistance of the oxide and the crud layersbetween the coolant and the metal surface need to be taken into account in the calculation of themetal oxide interface temperature.



1-1(1-19)




lrl Tv

Q

cds >,

,-Op

k r

ta ei , mu

pef

Figure 1-1: Factors that aect cladding temperature, AA-TECDOC-99, 1998.

Table 1-5: Variables in Determining Metal-Oxide Inteace Temperature

• Bulk fluid T• Laminar boundary layer

Single phase

bulk fluid Theat fluxheat transfer coe fficient ( ittus-Boelter, etc. )

hydraulic diameter fluid roerties: density, viscosity, secific heat, thermal conductivity, velocity Nusselt number

Two phase

ditto, but for nucleate boilinWall °T = °T",;o + °Twheat transfer correlation ens-Lottes, Tho, hen, etc.): heat flux, ressure

• CUD layerheat fluxcrud-bulk fluid interface Tcrud thicknesscrud thermal conductivity

• Oxide layerheat fluxoxide-crud interface Toxide thicknessoxide thermal conductivity






The thermal-hydraulics of the coolant and the thermal properties of the metal are reasonably wellknown. The crud thermal characteristics are the least well-characterized and the differencebetween the deposits observed a fter shutdown and those present during operation are also notalways known with certainty.

The heat ux or power, generated in the fuel rod is a key parameter in each one of thevariables that affect the metal/oxide interface temperature as listed in Table 1 - 5 The temperature

gradient produced by the heat flux increases with the increased thermal resistance (decreasedthermal conductivity) of the material it pas ses through. The resulting increased temperatureincreases corrosion rate. The effect of heat flux alone on the corrosion rate is difficult to a sses sbecause other parameters are invariably involved as well. The heat flux can cause secondaryeffects that increase corrosion rate . As an example, corrosion resistance can be lowered due to thepresence of hydride surface layers on the cladding. Also, increased subcooled boiling at the Wfuel rod surface may enrich water-soluble species in the zirconium oxide that potentially couldresult in degradation of the oxide protectiveness .

Different fuel rod poer histories can affect the oxidation rates in different manners. n thiscase, the effects are interactive with the oxide and possibly the crud thickness as these developwith exposure . An example of a theoretical, calculated case is given below.

The power histories for a Low Leakage LL fuel management scheme by a W vendor arecompared to his out-in scheme OJ at the same exposure, Figure 1-14 The high power early in

life in the

LLscheme is advantageous from an oxidation perspective, since then the oxide layer is

thin, the interface temperature is low and in addition the corrosion rate i s low. The OJ schemeexperiences higher power at the end of life when the oxide thickness/interface temperature ishigher, the oxidation rates and the resulting increases in oxide thickness are greater. Thecomparison o f oxide thickness shown in Figure 1-14 indicates a corresponding higher peak oxidethickness at end-of-life of the OJ scheme, Strasser et aI., 199






Figure 1-14:

1 . £0

5

0

v. 0n

�0. '

0 8

00. 700 0

0 0

040 0

cv

tI d 12

Sn

9

0

0

-I . -- -

-VD201

!

_ I

'

0 IS 2

Cor p q up /

w

U V S. . VL SS.

RABU g Bum

. _1

,

10 O

v� �

O Ox

£0

Comparison of the efects of out-in (VD1 � and low-leakage (VD1LL) fuel management schemes on Wcladding oxidation, Strasser et aI., 199.






Properties of zirconium oxide and deposits

The temperature of the oxide-metal interface can be calculated from the relationship:

where,

T temperature of the oxidemetal inteface (Kelvin)Tw temperature of the oxidewater interface (this as sumes no crud) (Kelvin) heat flux (/m2)K oxide thermal conductivity (/m K) oxide thickness (m)

The he at flux and the oxide thermal conductivity are key parameters in determining the interfacetemperature. Figure 1-15 compares some of the unirradiated and irradiated ZrO2 thermalconductivity data and Table 6.9 on Page 168 of -TEO-996, 1998, gives an additional,extensive database with a wide range of values. The actual values of K used by modelers are in therange .8 . m K, which is quite wide. The temperature drop acros s a 7m oxide layerwith a heat flux of 7 /cm

2is 31 using a K of 1.8 and 7 with a K of .8. The difference in

K can cause a 4X increase in post-transition corrosion rates in some predictive models.

3

Oul-ofpi e

it +

x +

x x x x

E 2

x?

In uC0UOxd Sampethckss: Jm

342°C/58 x

PW·A 5 PWA PWD30 5/3 h u

2 40 6Tre

Figure 1-15: ZrO Thermal Conductivity, Stehle et aI., 1984.

A comparison of predicted oxide thickness as a function of time using a smaller difference inthermal conductivity of 1. and 1.8 is shown in Figure 1-16. This can still result in significantdifferences in oxidation predictions for high burn-up performance.



1-16(1-19)




Figure 1-1:

200-34--.r-

H F l u x 0 7 M\ m-2�150�_r

. 1 I I I I I I hem cod W m l 00+A� hem co = 1 .B l K V J//� ! :t>/o 0

I i � /�! -- �=�=-++}-

00 200 300 400 500 600 700BO

900 1 000 00 1 200

(d)Examp le of effect of different oxide thermal co nductivity values on predicted oxide thickne sses (Nu clear Electric ZI RC Code predictions), AA-TECDO C-99, 1998.

The thermal conductivity of irradiated Zr02 is slightly lower than that of the unirradiated oxide.Other changes in the conductivity value could be induced by penetration of the porosity by thecoolant and its components, cracking and spalling. These added uncertainties will be magnified asthick oxide films are developed at extended burn-up or on materials with poor oxidationresistance to the coolant conditions.

The sources of crud on the fuel are the corrosion products from the reactor primary coolantsystem that deposit primarily on the hot heat transfer surfaces i. e. the fuel cladding. The effect of

crud deposits on the cladding metal-oxide interface temperature are subject to even greateruncertainties than the effect of the oxide layer, because the crud can be present in a ide varietyof morphologies. The crud layers can have physical structures that vary from dense layers toporous ones. The porous layer types can vary as well and be fluffy, loose deposits; densestructures with pores and cracks or relatively loose deposits with convective coolant chimneys(wicks). Each of these forms will have a different thermal conductivity. The conductivity of theseheterogeneous deposits depend partly on the crud composition and structure and partly on thecomposition and thermal state of the coolant that penetrates its voids.

The measurement of the thermal conductivity of the crud layers, particularly the loose or nonadherent ones, is extremely difficult since the preservation of the layers for subsequentmeasurement may not be possib le. A further, more basic question is what the crud layer wasduring operation and whether the crud deposited on the fuel during shutdown; this questionapplies primarily to the loose crud types. Examples of crud thermal conductivity measurementsand estimates are shown in Table 1-6 and are compared to the thermal conductivities of waterand steam that can infiltrate the crud. he conductivity of the various crud compositions isgenerally higher than that of Zr02, higher than that o f water and significantly higher than ste am -- that's the good news; unfortunately the variability and uncertainty among the various crud typesis greater.






Table 1 -6 : Thermal Conductivity of R Deposits on Fuel Cladding, Wikmark & Cox, 21 .

LayerThermal Conductivi ty

RemarkW·mK"

Water .57 288, 7MPa

Steam .63 288, 7MPa

"Porous magnetite crud 9% density

Porous hematite crud >6 1 % density

R zinc crud 1 . 1 At 35

CuO 3.3 At 1

Compressed hematite . 5 At

Dense copper-iron crud 2 At 220

The "normal crud that results from normal reactor operation rarely results in excessive claddingtemperatures and the increas ed corrosion observed is due to a combination of increasedmetal/oxide? temperature and Li concentration that is increased significantly above it s nominallevel in the coolant within the crud and the oxide film.

mpurities such as AI, a, Mg, Mn and Si tend to densify the crud and have the potential toraise the cladding temperature a s shown in early ex-reactor experiments, Figure 1-17 andconfirmed by fuel failures due to such crud in the Steam Generating Heavy ater Reactor(W) in the K. Since then, no adverse effects have been identified in-reactor to datealthough recommendations are to keep these impurities at a low level in the coolant. Measuringthe effect of crud on cladding temperature has been and and is a current topic of investigations inthe RENE loop in adarache, France.



1-18(1-19)




Sf

J

�"� � �S_("T

e

'-

CI0

�O�" ) I. ) ) I II O. !.�CROO) ."

0 1 0

Figure 1-17: The effect of magnetite crud on maximum wall temperature elevations, Macbeth et aI., 1971.

The injection of Zn for the reduction of activity transport or steam generator tubing corrosion hasresulted in the formation of zinc ferrites in a thin, dense layer. A smal l, but to date nondetrimental increase in corrosion has been noted due to the resulting temperature increas e.

Heavy crud deposits that raise the interface temperatures significantly can result in fuelfailures, see Section

As a n example a he at flux of 1 /cm2, about the maximum for Ws, the effect of thermal

conductivity is as follows:

1) a 1m-thick crud layer with a conductivity of m rai ses the clad temperature by6,

2) a 15m steam-filled crack with a conductivity of .7 m ra ises it by 58.

The small steam gap would raise the nominal 95 clad surface temperature to over 55.Oxidation at this temperature level would fai l the clad in significantly less than 1 days. At 6-65 the clad would fail in less than 5 days.

Thermal Hydraulics. The temperature of the coolant entering the bottom of a reactor core isuniform and increases as it pa sses through the core a s a function of core power and coolant flowrate. For example, an increase in the coolant inlet temperature in a W will result in higher fuelrod temperatures that in turn can increase corrosion rate.

Copyright © Advanced Nuclear Technology Inteational Eupe AB, ANT Inteational, This information is the prpey of Advanced Nuclear Technology Inteational Eurpe AB or is licensed for use by Advanced Nuclear Technology Inteational Eurpe AB by s

customers or paners The information may not be given to, shared with, or ced to third pay, used for unautho sed purpose, or be copied or reprduced in any form whout the written permission of Advanced Nuclear Technology Inteational Eurpe AB




B y k B C

nitial zirconiu surface conditionZirconium and its alloys are highly reactive metals and in oxygen containing atmospheres alwayshave at least a thin oxide film on their surfaces, ox, 1976. This film can be dissolved in amixture of HN0 (High Frequency) during electropolishing, but will always reform when themetal is exposed to air or water at Room Temperature (RT) Hence its description as the "airformed oxide film . This film can also be removed by dissolution into the metal by heating at hightemperature in a vacuum (or near vacuum). This occurs because oxygen is more stablethermodynamically when present as a dilute solution in the metal than it is as a Zr02 surface film.

At intermediate temperatures ( > �) the air-formed oxide will increase in thickness as aresult of oxygen diffusion through the air-formed oxide to produce more layers of Zr02 at theoxide metal interface. f the oxidation environment is water then water molecules are dissociatedat the oxide/environment interface by the electrons from the oxidation process diffusing out (in

the oppo site direction to the oxygen ions diffusing in) . The hydrogen atoms res ulting from thisprocess may recombine to form hydrogen molecules in solution in the water but some of themmay manage to diffuse into the metal (s ee hydrogen uptake mechanism later) . However, there isno evidence that hydrogen in the atomic form can diffuse through the Zr02 crystal lattice. Theonly mobile species in the oxide film at normal reactor operating temperatures are oxygenions/oxygen ion vacancies and electrons/electron holes .

Galvanic processesThe oxidation process should be considered as an electrochemical cell reaction. Figure -1,ox, 3 and 5.

Figure -1:

Aic af ctn at t xde /ma tc:-

Zr ) " ' . . . . ( :Cthi hll t 8 [h onvt t s

° 4 . " (Eq. 3)

S t h yg ccs \ gt u hg t 0 h z st ,

tb cts w at v fe\ c ss h sct wth acks x tgg y c ph pc t thc hc acti shl cs

A W sin

F . . . . . . . . . . . . . . . . . . . . . . q

pt chrg s

+ . . . . . . . . . . . . . . . . . . . . . . . q

"Corrosion as an E lectrochemical Process.






The oxygen vacancies will diffuse out from the oxide/metal interface in the opposite direction tothe oxygen ions, which have been shown to diffuse preferentially via the oxide-crystalliteboundaries, ox & Roy, 1966; ox & Pemsler, 1968. The hydrogen atoms released at theoxide/environment interface by reduction by the electrons can recombine to form hydrogenmolecules (which will disso lve in the water) or can diffuse inwards to the oxide/metal interface ifthey can find a suitable path through the oxide film. The direct diffusion route through the oxide

film is not available to them, since all claims to measure such diffusion coefficients have beenshown to be due to OH ions migrating down cracks or pores in the oxide and were not H

diffusion through the Zr02 Lattice, Ramasubramanian, 1996 and Elmoselhi et aI., 199. e willgo into more details about this process, since the proportion of hydrogen atoms successfullydiffusing into the metal gives the value for the "Hydrogen Pick-p Fraction

'' .

Ratedeterining step for zirconiu oxidationOnly in a very few conditions, far removed from those in an operating reactor has Zr 4+ migrationin the oxide been observed. n the range of normal reactor operating conditions there are only twomobile species in the ZrO2 lattice: oxygen ions/vacancies and electrons/holes. Although oxygendiffusion is a slow process, it is the essential mechanism by which the oxide thickens. The

migration of electrons is even slower. Zirconia films are better insulators than they are oxygendiffusion barriers. The only competitor in the electrical insulator field is hafnium oxide, and bothare being considered as insulators in micro- electronic devices because of their excellent resistivityand high band-gaps (-5 eV).

Because of this very high electron resistivity electron transport and not oxygen ion diffusion isthe rate determining step in the zirconium oxidation process . This fact has been demonstratedmany times by the measurement of the electrical potential that develops across the oxide filmduring corrosion. The potential of the oxide/metal interface relative to the potential of theoxide/environment surface can be measured in any non-aqueous environment ( surface potentialsconfuse the measurement technique in water but it can be measured in dry steam, oxygen, air andfused salts . The results are always the same: the oxide/metal interface rapidly develops a negativepotential after oxidation starts. The potential increases � 1. I to -1. V, roughly equal to the half

cell electrochemical potential. Such a negative potential develops to accelerate the electronmigration process and retard the oxygen-ion diffusion process, until both processes operate at thesame rate and further increases in space charges cease, Figure -, Bradhurst et aI., 1965. Thus,the electron transport is the more difficult process since it needs an accelerating potential toequalise the two processes. Since the rate controlling process is not oxygen ion diffusion throughthe Zr02 lattice, the pre-transition corrosion kinetics are not pa rabol ic. Electron conduction-therate determining step para bolic obeys either Poole-Frenkel or Mott-Schottky kinetics, which tendsto give kinetics nearer to a cubic, but they can vary over quite a range of exponents.

, HPUF






Figure -

1 .6,-I.

5 1

z: 1.0zou 08

E ALST CONTDURNG N

(7-3 M IUT)

ti, lnu

V

IFF PUMP NT

2' • 6

60

Behaviour of potential developed d uring oxidation of zirconium at

Suggestions that the rates of the surface reactions are rate controlling, Bossis et aI., , do notbear inspection, since the initial growth rates of oxide films in oxygen, air, steam, water, H2S04« .5 molar ), HNO « .5 molar ) , NH40H, and even LiOH « .1 molar) are all nearly identicalat temperatures in the 3-3 5 range. The big differences between these environments onlyappear when the oxide films' protective layers begin to break down (see later). The rates ofsurface reactions in this wide range of environments must vary by many orders of magnitude.Arguments for rate control by the rate of the surface reaction usually ari se because the authorshave chosen to ignore the cathodic half-cell reaction (i.e. electron transport).

Structural growth of zirconia filsThe initial " air-formed oxide present on all Zr alloy surfaces exposed to an oxygen-containingenvironment, consists of an ar ray of roughly equiaxed nano-crystals of Zr02, with many differentorientations relative to the orientation of the Zr grain on which they form. These crystallites are� nm diameter and are either cubic or tetragonal ( ) zirconia, Ploc, 197. t is very difficult tostrip and determine the crystallography of these crystallites without doubts, Ploc, 1968 and 1969.Such a determination may not be important for the structure of thicker oxides. At elevatedtemperatures some of these crystal orientations grow preferentially to develop into arrays ofcolumnar crystallites. The preferential crystallite orientations are those that will minimise stressdue to the volume change on going from Zr to Zr02 (the Pilling-Bedworth ratio of �.56).

The result is an oxide film formed of columnar Zr02 grains with a preferred orientation,Figure -3, Yilmazbayhan et aI ., 6. The extent to which these columnar crystallites growwithout breaking down is a function o f the alloy composition and the oxidation environment.Adjacent to the oxide/metal interface, where new oxide is forming, there is usually a layer oftetragonal Zr02 crystallites. This has been established by Raman Spectroscopic studies of theoxide, Godlewski et aI., . At the oxide/environment interface these crystallites transform tomonoclinic ( ) -zirconia. This -zirconia transformation will have inevitably generated microcracks in the oxide because of the volume change associated with the phase change. Possiblemechanisms for this transformation will be discussed later (see Section .7).

Copyright © Advanced Nuclear Technology Inteational Eupe AB ANT Inteational . This information i s the prpey o f Advanced Nuclear Technology Inteational Eurpe AB or is licensed for use by Advanced Nuclear Technology Inteational Eurpe AB by s

customers or paners. The information may not be given to shared with or ced to third pay, used for unautho sed purpose or be copied or reprduced in any form whout the written permission of Advanced Nuclear Technology Inteational Eurpe AB.




Figure -: Diferences in extent of columnar oxide crystallite formation, Yil mazbayhan et aI. , .

The growth of columnar crystallites of -Zr02 (?) and their breakdown can show up as repeatedcycles in the oxidation kinetics and as "layers in the resulting oxide, Figure -4 and, Figure - 5 ,Bryner, 199 an d Motta et aI. , 5

Figure -4:

Q"

N� 8

iz G�� 4�

2 3

(Y)

Sho-time corrosion weight gain of Zircaloy-4 in water at K.

4 5





ZAT2 PECIL TOPIC EPOR 2 - CORROSION MECHNISMS IN ZIRCONIUM ALLOYS

3 Key parameters that affect corrosion mechanismsand rate

Material parameters that affect corrosion mechanismsand rate (Friedrich Garzarolli)

The major material parameters that can affect corrosion are:

• lloying additions and impurities• olute content• PP characteristics (the influence being their size and composition)• egree of cold-workrecrystallisation

The effect of all these parameters on corrosion rate differs in different environments, such ashydrogenated, oxygenated, and LiH-containing coolants. The effect also varies with temperature and

system pressure.s pointed out in ection 2 and shown in Figure 3 the initial corrosion rate follows a parabolic

law in hydrogenated water. In pure zirconium a corrosion breakaway occurs resulting in an unstableoxide growth at some point. In corrosion-resistant zirconium alloys a rate transition occurs at athickness of about 2 m, depending on material. In pressurized water and high-pressure steam, repeatedcycles with corrosion rate transitions can be seen for highly corrosion-resistant alloys, e.g. with >3 %Fe. However, in low-pressure steam the corrosion rate remains constant after the transition and is, onaverage, similar to that in high-pressure steam at the same temperature. In alloys with very fine PP orwith a high-r alloy content the corrosion rate accelerates after a certain oxide thickness: at least if tested in hydrogenated environments at high pressures.

Figure 3-1:

/

I•"-

I/u. Zr,Wr hh pes sem

/ /•

//

';� :�I-

,l�oW

Ft high Cr or.m a SPPiernd gh p.. .. m

I

I�ol R. Ao , , ILow p."0

, .--

.�o es s loy

IWr a hgh rtm

rinciple corrosion behaviour of diferent materials in different environments.

-

In PWRs the oxide film is of the uniform type and its growth rate follows out of pile behaviour up to anoxide thickness of about m, after which it is enhanced by irradiation, Figure 3-2. The irradiationenhancement factor increases primarily with increasing n content. t very high fluences the corrosionrate often increases further probably due to irradiation-induced dissolution of intermetallic particleswithin the metal matrix and other phenomena.

4 owever, the corrosion rate follows a cic law in the retransition region

Copyri ght © Advanced Nuclea Technology Inteatonal Eupe AB, ANT Inteatonal, 007. Ths nfomaton s the ppey of Advanced Nuclea Technology Inteatonal Eupe AB o s censed fo use by Advanced Nuclea Technology Inteatonal Eupe AB by s customes o panes.

The nfomaton may not be gven to, shaed wh, o ced to thd pay, used fo unauthorised pupose, o be coped o epduced n any fomwthout the wten pemsson of Advanced Nuclea Technology Inteatonal Eupe AB.




Figure 3-2:

:

c'i�

EoQ

E:c'<C:

1

1,0

8

6

2

1 r

Noiz ime It*x(RT, h)

a.) Corrosion of Zircaloy-4 in PWRin deionised water and in LiOH.

�

!II

3 -!

6

r �

,i

,.

0C

5

4 -

3

2

I Normali me -ORn.

'W

2.5

b.) Corrosion of Zr-2.SNb in PWRin deionised water and in LiOH.

5b

_A.�-J·/ !J -.:- -G f �T

z Q/ h

c.) Corrosion of Zircaloy-2 and Zr-2.SNb in BWR and out of pile in water.

Effect of irradiation under hydrogenated conditions (PW and oxygenated conditions (BW on corrosion of Zircaloy-2/4

and Zr2.SNb, Garzarolli et aI., 1991.

Copyri ght ©Advanced Nuclea Technology Inteatonal Eupe AB, ANT Inteatonal, 007. Ths nfomaton s the ppey of Advanced Nuclea Technology Inteatonal Eupe AB o s censed fo use by Advanced Nuclea Technology Inteatonal Eupe AB by s customes o panes.





In BWRs, corrosion rate is increased particularly in the early exposure period and saturates at highexposure times. Zr-2.5Nb is, in PWRs, much more corrosion resistant than ZircaloY-4. However, inBWRs, much higher oxidation than ZircaloY-2/4 is observed.

In BWRs, besides uniform corrosion, shadow corrosion, starting at <10 days, can appear at positionsclose to Inconel or components and nodular corrosion, starting after an exposure of 10 to 100 days.Furthermore, a late increased uniform corrosion is observed (at least with ZircaloY-2) at high burn-ups

and long exposure times. Figure 3 3 shows the time dependency of the different types of in BWRcorrOSOn.

Figure :

E.U

£>

�>

l0r----�-----,-

1+-1

xpu m (d)

Time dependency of the different types of corrosion in BWR.

The pressure tubes of CAD U reactors show an early corrosion enhancement for ZircaloY-2 and

Zr-2.5Nb, somewhat similar to BWRs, but no differences between ZircaloY-2 and Zr-2.5Nb. However,in CADU reactors a late acceleration of corrosion at long exposure times (>2000 days) is observed inZircaloY-2 but not Zr-2.5Nb.

ome fraction of the corrosion hydrogen Hydrogen ickp Fraction (PU) is picked up by themetal. The PU is usually lower in oxygenated than in hydrogenated environments. In BWR at verylong exposure times the pick-up of corrosion hydrogen increases to very high levels (from 5-20% toalmost 100 %) somewhat before the late increased oxide buildup rate starts. It is interesting to note thatthat, also in CAD U reactors, a significantly late increase of the hydrogen pick-up is observed (at >2000days) for ZircaloY-2 but not for Zr-2.5Nb, rbanic et a., 1987.

Copyri ght ©Advanced Nuclea Technology Inteatonal Eupe AB, ANT Inteatonal, 007. Ths nfomaton s the ppey of Advanced Nuclea Technology Inteatonal Eupe AB o s censed fo use by Advanced Nuclea Technology Inteatonal Eupe AB by s customes o panes.




ZRT 2 SPEC TOPC REPO 2007 COOO MECM ZCOUM OY

4 Description of different forms of corrosion(Friedrich Garzarolli)

During corrosion of zirconium alloys in water and steam three different types of corrosion

features can be seen in principle, namely:

• Uniform corrosion (in all types of LWRs), see Figure 4-A• Nodular Corrosion (almost only in BWRIRBMK), see Figure 4-• Shadow Corrosion (almost only in BWRIRBMK) that can appear as nodular or as increased

uniform corrosion, see Figure 4-C

Uniform corrosion is the normal mode of zirconium alloy corrosion. Nodular corrosion ischaracterized by much thicker, local oxide patches appearing as white spots. Shadow corrosion isan increased zirconium oxide thickness (either nodular type or a thick uniform oxide) at positionswhere or Ni-base components are close to the zirconium alloy material, appearing as a shadowof these components.

. _

.

1bAN T IN2Zr02

A) NomalUnifor

NularB)

-Inceased

C Unifor

Figure 4-1: Types of oxide films formed in pressurized water and steam on zirconium alloys.

The burn-up (or time) dependency of the corrosion rate varies for different types of corrosion. Thetime dependency of uniform corrosion usually has at least two different regimes both out of pileand in pile. nitially, corrosion follows an approximately cubic law and later after a rate transitiona more or less linear law. During the initial period of oxidationcorrosion of zirconium alloys, athin, protective, black oxide is formed. As the zirconium oxide grows in thickness, the outer partof the oxide (that which faces the water/steam phase) is transformed into a greyish, porous oxide.The transformation to porous oxide starts at isolated, localised regions and then extends over thesurface as can be observed from the development and growth of grey spots e.g. Figure 4-2. Thesegrey regions can be seen under in-PWR corrosion conditions to form at an oxide thickness of about 5 m, representing the thickness at which the regime where irradiation enhanced corrosionbegins. This critical thickness is reduced to about m in out-of-pile tests at high concentration of LiOH. No such grey dots are seen in pressurised water or steam out of pile, probably because thecolor change at the transition is very moderate. The dense to porous oxide transformation occursat a certain oxide thickness and is usually cyclic (and is not enhanced), at least if corrosionproceeds in high-pressure water and steam. EM examinations have shown that the oxide has alayered structure e.g. Figure 4-3 with bands of fine pores separated by dense layers with athickness that is often reported to decrease with increasing corrosion rate.

Copyright © Advanced Nuclear Technology Inteational Eupe AB. ANT Inteational. This information is the prpey of Advanced Nuclear Technology Inteational Eurpe AB or is licensed for use by Advanced Nuclear Technology Inteational Eurpe AB by s

customers or paners The information m ay not be given to. shared with. or ced to third pay, used for unauthosed purpose. or be copied or reprduced in any form whout the written permission of Advanced Nuclear Technology Inteational Eurpe AB



ZRT 2 SPEC TOPC REPO 2007 - COOO MECM ZCOUM OY

Figure 4-2:

Figure 4-3:

Appearance of corroded Zircaloy tube exposed to high-pressure water containing 70 ppm LiOH at 350°C for 140days, Beie et aI., 1994.

--,

SEMmicrograph on cross section of the M5 oxide layer formed in a PWRafer 6 cycles at span 4 in backscatered electron mode, Bossis et aI., 2006.

At low temperatures, e.g in a BWR, corrosion may remain in the pre-transition regime for long

time periods. Nodular corrosion and shadow corrosion start in-BWR after a few to 100 days afterstart of operation and usually saturate at higher exposure times. Figure 44 shows the appearanceof nodular corrosion and a typical metallographic cross-section. t should be mentioned that theshape of nodular corrosion can vary. The growth rate of nodules decreases at high burn-ups, atleast for ZircaloY-2.

Figure 4-4: Typical appearance of nodular corrosion in visual inspection and metallographic examination.


customers or paners The information m ay not be given to, shared with, or ced to third pay, used for unauthosed purpose, or be copied or reprduced in any form whout the written permission of Advanced Nuclear Technology Inteational Eurpe AB



ZRT 2 SPEC TOPC REPO 2007 - COOO MECM ZCOUM OY

Figure 4-5 shows a typical example of shadow corrosion on a BWR fuel channel due to theproximity of a control rod handle. The oxide thickness in the area of the shadow appearsmostly as a much thicker, uniform oxide but sometimes as a nodular oxide (see Figure 4-4).

Figure 4-5: hadow corrosion

oth in BWRs and PWRs/VVERs, late-in-life corrosion acceleration may occur due to the impactof zirconium hydride rim formation, PP dissolution and other effects causing a degradation of the zirconium oxide barrier layer (see Figure 1-3).

Uniform corrosion exhibits a strong temperature dependence, Figure 4-6. However, thenodular corrosion rate temperature dependence is on the contrary very small. Thus, extremecorrosion rates that may lead to failures are always due to uniform corrosion.

Figure 4-6:

F-O.l

,

,'

e'

jl��10'

1

In hih srewa(tm

\yNoU (xa)

\ \ n-

' wa

>"\ A'� \' / VNou

'. " \, , 0 ' Uio

:nt/"1"-'t, ul �"(h p e "f l ) Ff HBWan

Kd

W;'2--I'------ ----.0pro t: 10" K-I!

50 4 s Tu: "

Hypothetical Zircaloy corrosion map for an exposure of 500 days in reactor under oxygenated and hydrogenatedwater conditions, and outside the reactor in high pressure water and steam, Garzarolli et aI., 1979.

Copyright ©Advanced Nuclear Technology Inteational Eupe AB, ANT Inteational, This information is the prpey of Advanced Nuclear Technology Inteational Eurpe AB or is licensed for use by Advanced Nuclear Technology Inteational Eurpe AB by s

customers or paners The information m ay not be given to, shared with, or ced to third pay, used for unauthosed purpose, or be copied or reprduced in any form whout the written permission of Advanced Nuclear Technology Inteational Eurpe AB



IRAT EL TO REO 7 OOON MEHNM N ONUM ALLOY

5 Open items (Friedrich Garzarolli)

The corrosion mechanism of zirconium alloys depends on the environmental conditions and isgoverned by several very complex and often related processes and features. In this report we havedealt with many questions but more understanding is yet to be evolved.

To clarify the mechanisms there is a need for co-operation between scientists from severaldifferent technological areas. A better understaning of the corrosion mechanisms will help:

• Zirconium alloy development for P and P• Development of water chemistry specification• Root cause analysis of accelerated corrosion phenomena.

In Figure 5-1 the boundary conditions, rate-governing mechanisms, as well as the processes andfeatures affecting corrosion are summarised. The corrosion rate depends on the oxygen iondiffusion rate and on the barrier thickness. Temperature effects on corrosion rate are mostlygoverned via oxygen ion diffusion (although there is also some effect from the barrier layerthickness as pointed out by eie et a., 1994). However, the environment, zirconium alloy

chemistry and microstructure affect both barrier layer degradation and oxygen ion diffusion.Oxygen ion diffusion may by governed, under certain conditions, by electronic conductivity. Moreor less always, oxygen ion diffusion through the outer part of the barrier layer occurs along oxidegrain boundaries. Close to the metal/oxide interface, where the oxygen potential is very low andthe oxygen vacancy concentration is very high, a rather fast bulk diffusion occurs. This bulkdiffusion is affecting very important processes for the barrier layer degradation such as:

• mode of crystallisation of the oxide layer,• fraction of tetragonal oxide• stresses created by the larger volume of the oxide than the metal,• oxide crystal alignment.

The authors of the report think that more effort should be put into the study of the followingaspects:

• Improved characterisation and understanding of the structure of the barrier layer at themetal/oxide interface

• Quantification of the thicknesses of the grain boundary and bulk anion diffusion zoneswithin the barrier layer and how are they affected by the material and environment?

• Development of an understanding of how anion diffusion in the innermost part of the oxidelayer affect the metal/oxide interface, the barrier layer thickness, the character of sub-oxides,the metal/oxide interfacial roughness and finally the crystallisation process within the barrierlayer?

More and more reports conclude that the type and width of a barrier layer is important forcorrosion resistance. It is not clear how the barrier layer affects corrosion rate. The majorquestions are:

• Motta et a., believe that crystallite alignment of barrier-layer tetragonal oxide is animportant influence on corrosion because optimally-aligned crystals minimise compressivestresses in the oxide. Can this very interesting idea be confirmed by stress measurements andcan it be explained how PP and Nb affect this process?

• Is the tetragonal oxide at the barrier layer stabilised by high compressive stresses, as somescientists believe, or by a fine grain size?


customers or paners. The information m ay not be given to, shared with, or ced to third pay, used for unauthosed purpose, or be copied or reprduced in any form whout the written permission of Advanced Nuclear Technology Inteational Eurpe AB.



IRAT EL TO REO 7 OOON MEHNM N ONUM ALLOY

• Often a high fraction of tetragonal oxide is observed in samples with low corrosionresistance. Is this high fraction of tetragonal oxide the consequence or the cause of a highcorrosion rate?

• Under which conditions can amorphous oxide form at the metal/oxide interface? This canprobably only be examined in samples cooled very quickly from the corrosion temperature(e.g. by removing the autoclave from the furnace during cool-down).

It is very questionable as to which process initiates the corrosion rate transition. In particular thefollowing questions arise:

• Does the tetragonal to monoclinic oxide transformation occur preferentially at the corrosionrate transition as is often stated or is the transformation a more a continuous process?

• What is the hydrogen/proton concentration at the oxide/metal interface, which parametersaffect it, and, if it is an important parameter, which concentration causes break-down of thecolumnar grains and nucleation of new equiaxed grains at the metal/oxide interface?

Relatively good-quality information exists, from in-situ electrochemical experiments (impedancemeasurements and polarization tests), on corrosion potential, electrical conductivity and oxide

characteristics under out-of pile, hydrogenated-water conditions. However, similar tests undermore real conditions are needed, such as:

• In-situ electrochemical tests under B and P conditions with samples exhibiting bothhigh and low corrosion resistance, nodular corrosion and with different deposited crudtypes. Such measurements should also be able to give reliable answers to questions regardingthe effect of irradiation on corrosion including why its effect differs with environment andzirconium alloy.

• In-situ electrochemical tests on different alloys in different water chemistries to assess theimpact of water chemistry on corrosion mechanisms.

As far as the action of PPs on corrosion and hydrogen pick-up is concerned, the oxidation and

dissolution characteristics of PPs in the oxide layer grown in different environments on differentzirconium alloys is probably an important process step. A good database exists, e.g. from TE of oxide layers, regarding the oxidation and dissolution characteristics of Laves-phase PP in theoxide layer.

• However, almost nothing is known on the behaviour of the Zintel-phase PP.

Furthermore, it is suggested that more detailed studies are required to answer the followingquestions:

• Why is corrosion linear in I bar steam (does migration of H20 through fine pores controlcorrosion in low-pressure steam?)

• Why, in contrast, is corrosion cyclic in high-pressure water and steam (is it probably becausecorrosion is controlled by anion diffusion through a barrier)?

• Is there any difference in the long-term corrosion behaviour between ZircaloY- or Zircaloy-4 with optimum PP size and with fine PP size?





IRAT EL TO REO 7 - OOON MEHNM N ONUM ALLOY

For nodular corrosion, several widely-varying mechanistic models have been proposed. Nodularcorrosion starts at a critical oxide thickness and grows with a widely-varying rate usuallyexhibiting a more or less pronounced saturation. Obviously, nodular corrosion starts earlier inZircaloY- than in ZircaloY-4 but exhibits a more pronounced saturation in ZircaloY-. Tounderstand what are the important parameters which influence nodular corrosion, the followingquestions need to be answered:

• What are the important mechanistic processes?• Should we differentiate between initiation and later growth?• Why do ZircaloY- and ZircaloY-4 behave differently? Does the relatively large Zintel phase

PP in ZircaloY- enhance nodule nucleation and promote saturation of nodule growth?• In aggressive environments, why does nodular corrosion start earlier and why is nodular

corrosion rate higher? Does a high concentration of monovalent species, such as Na+, Li+,F-, etc., in positions with slightly increased porosity initiate and/or accelerate nodule growth?

The rod power history influences corrosion rate not only in Ps but also in Bs. Why is therea power history dependence in Bs keeping in mind that the impact of temperature oncorrosion rate is low and not as strong as it is in Ps? It is well known that gadolinia rods in

Bs were more prone to nodular corrosion and CILC failures (compared to U02 rods) in earlieryears. It is also well known that gadolinia rods have a very different power history compared tothat of U02 rods.

Why do corner rods in B assemblies often show much higher oxide thickness and Pvalues than rods in other locations?

What is the real mechanism of shadow corrosion?

Bounda

d afecting

processes d t:

TempeatueOxyge paial

pessue,evlome

(Wate eamoxen pessueedox ,adiatio puties

·Baler aye egradaUO

i unde:petansitpostansit }acceleated coosi

Mode of crystaizaonlnterlaye, aliovalet elemets

ctal oietati ad ate)

·Facton eag. Zr02'Sesses (crystal entaionSPP tea. ZO?

Figure 51:

'Z aloy (cem andmicost

nula cosionsad coospos beakawa

Oxygen on usooed y-GB d tside Bulk di at M/Ointe

Mechanism of zirconium alloy corrosion.

I·E of c ysalze H-enichm &igai vs p gadiantLi cceati iati{ G area (Gan sze)Sn oxde a GSPPG-uso ayer hckEeconc codvity·Coroslon oeia

The pick-up of corrosion hydrogen by zirconium alloys depends on the environmental conditionsand is governed by several very complex and often related processes and features. Clarification of this situation needs the co-operation of scientists from several different technological areas. It isconsidered that the major goals for future research should be:

• Zirconium alloy development for P and B.• Development of water chemistry specification.• Root cause analysis for accelerated hydriding phenomena.





6

ZIRATr 2 SPECL TOPC REPO 2007 COOON MECHNM N ZCONUM ALLOY

D, ml mmR Am

Various aspects of corrosion and the mechanisms of corrosion have been discussed in the

preceding sections. A brief summary of this report is given here:

6.1 Introduction (Section

The emphasis of this R is on the mhasms of corrosion of zirconium alloys used in the coreof nuclear power plants. The entire area of corrosion (and the accompanying absorption of hydrogen in the zirconium metal matrix) is of prime interest when considering performance of core components and, therefore, the performance of the entire reactor.

Corrosion of zirconium alloys is an electrochemicallydriven process affected by the microstructure and microchemistry of the alloy surface, the nature of the oxide layer that forms,the temperature at the metal/oxide interface, the chemistry and thermohydraulics of the corrodent

water, the effects of irradiation and the effects of time. Specifics of these processes are given in

various sections of this ST.In comparing BWRs with PWRs, with corrosion mechanisms in mind, the main features are:

• BWR coolant boils; PWR coolant does not. This has an important effect at the oxide/waterinterface.

• PWR coolant contains a high concentration of hydrogen; BWR coolant does not.Complementarily, BWR coolant contains a high concentration of oxygen, PWR coolant doesnot. This has an important effect on corrosion processes.

• PWR components generally operate at higher temperatures than BWR components.Corrosion processes are temperature dependent.

• oth reactor types employ chemical additions to the coolant which may affect corrosion andbuildup of deposits on fuel rods.

It also should be noted that BWR zirconium alloys continue to be primarily ZircaloY2 or slightvariants of ZircaloY2. PWR zirconium alloys no longer tend to be ZircaloY4, for reasons of insufficient corrosion resistance (and hydriding resistance) at high burnup, but have movedtoward zirconium alloys with Nb additions.

The type of oxides which form during corrosion in reactor water can be put into severalcategories.

) Uniform2) atch or accelerated uniform3) Nodular4) Shadow5) Renhanced

The two most basic are uniform and nodular corrosion. Uniform corrosion has an additional formof corrosion, - "patch or "accelerated uniform. The fourth category is "shadow corrosion, which can look like thick uniform corrosion but has some characteristics of nodular corrosion.The fifth category is crudenhanced corrosion, which is a temperaturedriven process induced bypoor heat transfer in crudimpregnated corrosion layers. These categories are discussed in detail invarious sections of this R.



6(614)




Uniform corrosion occurs in both PWRs and BWRs. The oxide itself is uniform in thickness andconsists of several different layers, discussed in Section 2. The shape of the corrosion versus burnup (fluence, time) curve in PWRs depends on several variables: initial PP size, irradiation,amount of cold work, specific alloy, water chemistry, temperature, local thermohydraulics,hydride concentration - all discussed in Section 3. For ZircaloY4 the corrosion thickness at highburnup may reach or exceed Jm, while some of the newer alloys have less than half of that.

As was noted in Table 2, only uniform corrosion occurs in PWRs under nonboiling,hydrogenated conditions.

In BWRs, uniform corrosion continues at a very low rate out to moderate burnups. Duringreactor exposure, the microstructure of the Zircaloys used in BWRs is continually evolving due toirradiation damage and second phase precipitate (PP) dissolution. In the range 3050 Md/kgU(6x0

' n/cm

, E> MeV), changes in the microstructure induce an acceleration of uniform

corrosion, as described in Section 4.1. First, patches of white oxide appear in the otherwise blackor grey uniform background. These patches remain very thin, about the same as the black uniformcorrosion film. At some point the patches cover % of the surface and oxide thickening occursat an accelerated rate, labelled "late increased corrosion.

Importantly, at the same or somewhat higher fluences, in BWRs a marked increase inhydrogen pickup fraction occurs. This hydriding is potentially a more serious issue than thecorrOOn crease.

In BWRs (and RBKs) nodular corrosion, characterised by locallythick oxide appearing as white spots, occurs. SmallPP Zircaloys (average PP size less than about . Jm) in general donot form nodules. In largePP Zircaloys, nodules initiate early in life and grow at a decreasingrate with fluence. In some cases, for aggressive water chemistries and susceptible material, nodulescan coalesce to cover the entire surface. Nodular oxide thickness does not generally causeperformance problems; however, in severe cases spalled oxide can be a source of "grit in controldrive mechanisms. Serious fuel failure problems can be induced by a combination of heavynodular corrosion and copperzincladen crud, resulting in the CILC (crud induced localizedcorrosion) phenomena discussed in Section 44

The fourth type of corrosion is the socalled shadow corrosion, described in more detail inSection 4.2. Shadow corrosion is induced on all zirconium alloys when they are in close proximityto many nonzirconium alloys such as stainless steel or Inconel. The oxide thickness is unusually

large and often appears to be particularly dense and uncracked. A common occurrence in BWRsis shadow corrosion oxide induced by a stainless steel control blade handle. Shadow corrosion has"always been present in BWRs, but not in PWRs, which is primarily related to the high PWRhydrogen concentration, which in turn reduces or eliminates galvanic potentials betweendissimilar alloy components. In BWRs, shadow corrosion has caused no performance issues untilrecently when, at one reactor, fuel performance problems were induced by unusually severe"enhanced spacer shadow corrosion. Also, more recently, shadow corrosion has been alleged tobe involved in BWR channel bow problems. oth issues are addressed in Section 4.

Various sections of this R address in more detail the mechanisms which are believed to be most important in describing corrosion in zirconium alloys.



62(614)



ZIRATr 2 SPECL TOPC REPO 2007 - COOON MECHNM N ZCONUM ALLOY

6.2

6.2.1

Basics of corrosion and hydrogen pick-up(Section 2)

Mechanisms

Corrosion of zirconium alloys at elevated temperatures is basically an electrochemical processinvolving a galvanic cell. The growth of the oxide film by oxygen vacancy diffusion is the anodicprocess and reduction of hydrogen ions from the water by electrons diffusing through the oxidefilm is the cathodic process. Since Zr is a very good insulator, measurements of the electricpotential that develops across the oxide show that electron transport across the oxide film is moredifficult than the oxygen vacancy diffusion process. Thus, electron transport through the oxide isra tedetermining.

After the oxide grows to about 2 m by forming columnar oxide crystallites with preferredorientations, breakdown of the largely impervious oxide occurs with an attendant acceleration inthe corrosion process (posttransition corrosion). This breakdown is probably a result of accumulation of impurities (or alloying elements) at the columnar crystallite boundaries combined

with a decreasing compressive stress in the oxide that is unable to maintain the stability of the

tetragonal (t)Zr phase. ther factors, such as ingress of water may also be involved in thetetragonal - monoclinic (m)Zr phase transformation that initiates posttransition corrosion.

The electronic conductivities of Zr films can be degraded by the incorporation of alloyingelements or impurities (especially Fe) in the pretransition oxide film. The oxide electronicconductivity is significantly reduced by irradiation with any species having an energy greater thanthe Zr bandgap energy (�5eV). Thus, UV, Cerenkov, gamma and beta radiation are all capableof enhancing the corrosion of zirconium alloys under the right experimental conditions. Fastneutron bombardment will only directly influence corrosion by enhancement of electronicconduction. However, the electronic conductivity of the oxide film can be indirectly enhancedthrough the fastneutroninduced displacement of Fe atoms from PPs into the zirconium matrixand into the oxide structure.

Galvanic corrosion processes between zirconium alloys and other metallic components of the

system (e.g. spacer grids, fuel channels, PPs) are possible in BWRs because of the effects of irradiation on electron conduction for all the components of the reactor. This is because thesystem is not reducing enough to prevent galvanic potentials being present between the zirconiumalloy and these other metals. These effects give rise to shadow corrosion and nodular corrosioneffects. Such effects are not seen in PWRs provided the dissolved hydrogen content of the reactor water is high enough to make all the various metallic species in the system are at the same,reversible hydrogen potential, such that no galvanic couples can develop between dissimilar metalsand alloys.

The uptake of hydrogen by the zirconium alloy matrxi during its corrosion, arising from thereduction of protons by the electrons from the Zr cathodichalfcell reaction is now considered tobe a function of the number of throughwall flaws present in the Zr film. Diffusion of H in theZr lattice, or through PP particles, can be dismissed on present evidence.

6.2.2 Testing

The various techniques used for measuring outofreactor corrosion of zirconium alloys (e.g.autoclaves and outofreactor loops) are discussed. Also discussed are the sources of variability inexperimental data, such as specimen shape; environmental impurities such as F, andimpurities/alloying content of the alloys tested. Examples of each of these effects are given and thevariation that these can induce in the oxidation kinetics that are subsequently measured.






7 References

Adamson R. B., Luz D. R., Davies J. H., "Hot Cell Observations of Shadow CorrosionPhenomena Proceedings Facagung der KTG-Facgruppe, Brennelemene und Kernbauelle,29 Februar Marz 2000, Forscungszenrum Karlsrue, 2000.

Adamson R. B., Cox B., Garzarolli F., Srasser A., Rudling P. and Wikmark G., "Corrosion of Zirconium Alloys ZIRAT- Special Topical Repor, Dec., 2002.

Adamson R., Cox B., Garzarolli F., Massi A., Riess R., Srasser A. and Rudling P., "ZIRAT Annual Report 2005.

Adamson R and Cox B., "ZIRAT- Special Topic on Impact of Irradiation on MaterialPerformance 2005.

Adamson R., Cox B., Garzarolli F., Riess R., Sabol G. Srasser A. and Rudling P., ZiratAnnual Repor, 2006

Anada H. and Tekada K., "Microstructure of oxides on ZircaloY-4, .Nb ZrY-4, and Zircaloy-z formed in .3-MPa steam at 673 °C ASTM STP 1295, pp. 34-54, 1996.

Anada H. e a., "Effect of annealing temperature on corrosion behaviour and ZrO2microstructure of ZrY-4 cladding tube ASTM STP 1295, pp. 74-93, 1996.

Anada H. e a., "Chemical state analysis of Sn and Fe in ZrO2 by Mssbauer spectroscopy

ASTM STP 1423, pp. 154-168, 2002.Andersson B., "The Enhanced Spacer Shadow Corrosion Phenomenon Facagung der KTG

Facgruppe Brennelemene und Kernbaueile, 29 February Marc 2000, ForscungszenrumKarlsrue, 2000.

Andersson B., Limback M., Wikmark G., Hauso E., Jonsen T., Ballinger R. G. and Nysrand A-C., "Test Reactor Studies of the Shadow Corrosion Phenomenon Zirconium in e NuclearIndustry: 13th In'l Symposium, ASTM STP 1423, Moan G. D. and Rudling P., Eds., ASTM 2002.

Arasi H. and Isigame J.M., Pys Saus Solidi Vol. A7, p. 313, 1982.Aser R. C. and Trowse F. W., "The Distribution of Hydrogen in Zirconium Alloy Fuel

Cladding: The Effects of Heat Flux J. Nucl. Ma. 35, p.5-121, 1970.Baur K, e a., "Electrochemical examinations in 350°C water with respect to the mechanism

of corrosion-hydrogen pickup ASTM STP 1354, pp. 836-852, 2000.Beie H.J. e a., "Examination of the corrosion mechanism of Zr alloys ASTM STP 1245, pp.

615-643, 1994.Benini G.G. e a., "The thermal oxidation of Ion-implanted Zirconium Corrosion Science

Vol. 20, pp. 17-39, 1980.Billo P. e a., Proc. 4. BNES-Conference on Waer Cemisry for Nuclear Reacor Sysems,

pp. 97, 1986.Blank H., "On the Origin of Different Rates of Inpile Post-transition Oxidation for Zr-based

Cladding Alloys Z. Meallkunde 85, pp. 645-657, 1994.Bla M. and Noel D., "Detrimental role of hydrogen on the corrosion rate of Zr alloys

A ST M STP 1295, pp. 319-337, 1996.Bla M., Legras L., Noel D. and Amanric H., "Contribution to a Better Understanding of the

Detrimental Role of Hydrogen on the Corrosion Rate of ZrY-4 Cladding Materials Zirconiumin e Nuclear Indusry: 12 In'l Symposium, ASTM STP 1354, G. Sabol and G. Moan, Eds.

ASTM Wes Consoocken, PA, 563-591, 2000.Brion C.F. and Wanklin J.N., "Inhibition by boric acid of the oxidation of Zr in high pressuresteam J.Nucl. Ma. N.3, pp. 326-337, 1962.

Bossis P. Lelievre G., Barberis P., is X. and LeFebvre F., "Multi-Scale Characterisation of theMetal-Oxide Interface of Zirconium Alloys Proc. 12th In. Symp. on Zr in e Nucl. Ind., ASTMSTP-q54, pp. 918-940, 2000.

Bossis P. e a., "Study of the Mechanism controlling the oxide growth under irradiation ASTM STP 1423, 2002, .90-22

Bossis P. e a., "Comparison of the high burnup corrosion of M5 and Low-Tin-ZrY-4; ASTMSTP 1467, pp. 494-524, 2006.






Bossis P., Veraege B. Dorio S. Gilbon D., Cabreou V., Dalmais A., Mardon J.P., Blad M. and Miue A., "In PWR comprehensive study of high burn-up corrosion and growth behaviour of M5@and recrystallied low-tin ZircaloY-4 Zr in e Nuclear Indusry: 15th InernaionalSymposium, Sunriver OR, June25-28, 2007.

Bouineau V., e al. "In situ measurements of the metal-oxide strain tensor in case of irconiumalloy oxidation Poser a 15. ASTM Zr Conference, Sanriver, Oregon, 2007.

Bradurs D. H., Draley J. E. and Van Drunen C. J., An Electrochemical Mode for theOxidation of irconium J. Elecrocem. Soc., 112, 1171-1177, 1965.

Broy Y., Garzarolli F., Seibold A. and Van Swam L. F., "Influence of Transition Elements Fe,Cr, and Von Long-Time Corrosion in PWRs Zirconium in e Nuclear Indusry: 12th In'lSymposium, ASTM STP 1354, G. P. Sabol and G. D. Moan, Eds., ASTM Wes Consoocken,PA, 609-622, 2000.

Bryner J. A., J., "The Cyclic Nature of Corrosion of ZircaloY-4 in 633K Water NuclearMaer., 82, 84-101, 1979.

Celain A., Anderson B., Ballinger R. G., and Wikmark G., "Enhanced Corrosion of Zirconium Base Alloys in Proximity to Other Metals: the Shadow Effect .48 5-498, In. TopicalMeeing on L WR Fuel Performance, Park Ciy, UT, April, 2000.

Cen J. S. F. and Adamson R. B., "Observations of Shadow Phenomena on Zirconium Alloys Proceedings of e Inernaional Topical Meeing on L WR Fuel Performance, ANS, Wes Palm Beac, Florida, 309-317, 1994.

Ceng B. and Adamson R. B., "Mechanistic Studies of Zircaloy Nodular CorrosionZirconium in e Nuclear Indusry: h In'l Symposium, ASTM STP 393, Adamson R. B. and Van Swam L. F. P., Eds., ASTM Piladelpia, 387-416, 1987.

Ceng B., Levin H. A., Adamson R. B., Marlowe M. O. and Monroe, V. L., "Development of a Sensitive and Reproducible Steam Test for Zircaloy Nodular Corrosion Zirconium in eNuclear Indusry, h In'l Symposium, ASTM STP 939, Adamson R. B. and Van Swam L. F. P.,Eds., ASTM Piladelpia, 257-283, 1987.

Ceng B-C., Krger R. M. and Adamson R. B., "Corrosion Behaviour of Irradiated ZircaloyProc. th In. Symp. in e Nucl. Ind., -STP-245, pp. 400-418, 1994.

Crisensen H., e a., "Experimental studies of radiolysis in an in-core loop in the Studsvik Rzreactor Proc. BNES conference on Waer Cemisry of Nuclear Reacors Sysems, pp. 483-485,

1996.Crisensen H. e a., BNES 3,d Worksop on L WR coolan waer radiolysis and elecrocemisry, Dorse, UK, Oc. 2000.

Cox B, "Degradation of Zirconia Ceramics in 3000 C Sulphate Solutions UKAEA, AEREHarwell. Repor HARD(c), p. 33, 1957.

Cox B., "Some factors which affect the rate of oxidation and hydrogen absorption of Zircaloy-z in steam J. Elecrocem. Soc., v. 109, pp 6-12, 1962.

Cox B., and Jonson T. J., "The Oxidation and Corrosion of Niobium (Columbium),Trans. Me. Soc, AIME v.�, pp. 36-46, 1963.

Cox B. and Roy c., "The Use of Tritium as a Tracer in Studies of Hydrogen Uptake byZirconium Alloys Aomic Energy of Canada, Repor AECL 2519, 1965.

Cox B. and Roy c., "Transport of oxygen in oxide films on irconium determined by thenuclear reaction '70 He, a) '60", Elecrocem. Tec., , 121-127, 1966.

Cox B. and Pemsler J. P., "Diffusion of oxygen in growing irconia films Nuclear Maer.,28, 73-78, 1968.Cox B. and Donner A., The Morphology of Thick Oxide Films on Zircaloy-z J. Nuclear

Maer., , 72-80, 1973.Cox B., "Oxidation of Zirconium and It's Alloys Advances in Corrosion Science and

Tecnology, Vol. 5, Edied by Mars G. Fonana and Roger W. Saele, Plenum, N. Y. pp. 173-391, 1976.

Cox B., "Assessment of in-reactor corrosion models and data for Zircaloys in water Proc. 2

In. Symp.on Environmenal Degradaion of Maerials in Nuclear Power Sysems-Waer Reacors,pp. 219-226, Monery, Ca., USA, 1985.





ZIRATr 2 SPECL TOPC REPO 2007 - COOON MECHNM N ZCONUM ALLOY

Cox B. and Fidleris V., "Enhanced Low-Temperature Oxidation of Zirconium Alloys underIrradiation Proc. 8th In. Symp. on Zr in e Nuclear Ind., San Diego, CA, ASTM - STP 1023,245-265, 1989.

Cox B. and Wu c., "Dissolution of Zirconium oxide film in 300°C LiOH J. Nucl. Ma. 199,pp. 272-284, 1993·

Cox B., Ungurelu M., Wong Y-M., and Wu c., "Mechanisms of LiOH Degradation and

H B Repair of Zr2 Films Proc. 11th In. Symp. on Zr in e Nuclear Ind., GarmiscParenkircen, Ger., ASTM - STP 1295, 114-136, 1996.Cox B., Hydrogen Uptake During Oxidation of Zirconium Alloys J. Alloys and Comp.,

Vol. 256, pp 244-246, 1997.Cox B. and Seik H., "Redistribution of the Alloying Elements During Zircaloy-z

Oxidation J. Nucl. Maer. V. 249, pp. 17-32, 1997.Cox B., "A mechanism for the hydrogen uptake process in irconium alloys J. Nucl. Maer,

Vol. 264, pp. 283-294, 1999.Cox B. and Wong Y -M., J., "A Hydrogen Uptake Micro-mechanism for Zr Alloys Nuclear

Mat er., 270, 134-146, 1999. Cox B., "Mechanisms of Zirconium Alloy Corrosion in Nuclear Reactors Proc. In. Symp.

on Corrosion Science in e 21s, Cenury, Manceser, UK, July 6-10t\ J. Corr. Sci. and Eng.,vol. 6, pp. 14, 2003.

Cox B., "Some thoughts on the mechanism of in-reactor corrosion of irconium alloysJ. Nucl. Maerials 336, pp. 331-368, 2005.

Davidson S. e a., J. Solid Sae Cemisry, Vol. 73, .47, 1988.Dalgaard S.B., Proc. IAEA Conference on "Corrosion of Reactor Materials Vol 2, p. 133,

1962.Deville S., Cevalier J. and El Aaoui H., "Atomic Force Microscopy Study and Qualitative

Analysis of Martensite Relief in Zirconia J. Amer. Ceram. Soc., B (5), 1261-1267, 2005.Edwards W. H., "Corrosion Investigations in U. K. Nuclear Plants Proc. 1" In. Congress on

Meallic Corrosion, London, U. K., May 8-14, 1961.Elmoseli M. e a., "A study of the hydrogen uptake mechanism in irconium alloys Proc.

th In. Symp. On Zr in e Nucl. Ind., Balimore, MD., -STP-1245, pp 62-79, 1994.Elmoseli M.B. and Donner A., "Inhibitors for reducing hydrogen ingress during corrosion of

Zr alloys ASTM STP 1467, pp 563-581, 2005.Eo Y., Simada S., Yasuda T., Adamson R. B., Cen J-S. E., Isii Y and Takey K.,"Development of new Zr alloys for a BWR Proc. 11th In. Symp. On Zr in e Nucl. Ind., ASTMSTP 1295, pp 825-849, Garmisc-Parkenkircen, Germany, 1996.

Fukuya K., Ecigoya H., Haori Y., Kobayasi K., Kobayasi K., and Sasaki T., "BWR FuelChannel Performance and Localied Corrosion at High Burnups Proceedings of eInernaional Topical Meeting on L WR Fuel Performance, ANS, Wes Palm Beac, Florida, 580-586, 1994·

Garbe K., Hensaw J., Polley M. V. and Sims H. E., "Oxygen and Hydrogen Behaviour inPWR Primary Circuits Proc. JAIF In'l Conf. Waer. Ce. Nucl. Power Plans, Oc. 13-16,1998, Kasiwasaki Ciy, Japan, 902-909, 1998.

Garde A. M., Pai S. R., Krammen M. A., Smi G. P. and Ender R. K., "Corrosion Behaviorof ZircaloY-4 Cladding with Varying Tin Content in High-Temperature Pressuried Water

Reactors Zirconium in e Nuclear Indusry: th In'l Symposium, ASTM STP 1245, A. M.Garde and E. R. Bradley, Eds., ASTM Piladelpia, 760-778, 1994.Garzarolli F., Manzel R., Rescke S. and Tenckoff E., "Review of corrosion and dimensional

behavior of Zircaloy under water reactor conditions ASTM STP 681, pp. 91-106, 1979.Garzarolli F. e a., "External corrosion in water cooled power reactors IAEA Tecn.

Commiee Meeing, Caderac, France, 1985(a).Garzarolli F., Bodmer R., Sele H., Trapp-Priscing S., Progress in understanding PWR fuel

rod waterside corrosion Proc. ANS Topical Mg. On Lig Waer Reacor Fuel Performance,Orlando, ANS, La Grange Park, IL, pp. 3-55 o 3-72, 1985(b).

Garzarolli F. e a., "Inuence of various additions to water on Zircaloy corrosion in autoclavetests at 350°C Enl. Halden Program Group Meeing, Loen, Norway, 1988.

first chapter - zirat12 str corrosion mechanisms

Documents