Unclassified NEA/CSNI/R(2003)8/VOL2 - Nuclear Energy ...

Unclassified NEA/CSNI/R(2003)8/VOL2

Organisation de Coopération et de Développement EconomiquesOrganisation for Economic Co-operation and Development 30-Apr-2003___________________________________________________________________________________________

English - Or. EnglishNUCLEAR ENERGY AGENCYCOMMITTEE ON THE SAFETY OF NUCLEAR INSTALLATIONS

PROCEEDINGS OF THE TOPICAL MEETING ON RIAFUEL SAFETY CRITERIA

Aix-en-Provence, France13th-15th May, 2002

JT00143628

Document complet disponible sur OLIS dans son format d’origineComplete document available on OLIS in its original format

NE

A/C

SNI/R

(2003)8/VO

L2

Unclassified

English - O

r. English

NEA/CSNI/R(2003)8/VOL2

2

ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT

Pursuant to Article 1 of the Convention signed in Paris on 14th December 1960, and which came into force on 30thSeptember 1961, the Organisation for Economic Co-operation and Development (OECD) shall promote policies designed:

− to achieve the highest sustainable economic growth and employment and a rising standard of living in Membercountries, while maintaining financial stability, and thus to contribute to the development of the world economy;

− to contribute to sound economic expansion in Member as well as non-member countries in the process of economicdevelopment; and

− to contribute to the expansion of world trade on a multilateral, non-discriminatory basis in accordance withinternational obligations.

The original Member countries of the OECD are Austria, Belgium, Canada, Denmark, France, Germany, Greece,Iceland, Ireland, Italy, Luxembourg, the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland, Turkey, the United Kingdomand the United States. The following countries became Members subsequently through accession at the dates indicated hereafter:Japan (28th April 1964), Finland (28th January 1969), Australia (7th June 1971), New Zealand (29th May 1973), Mexico (18thMay 1994), the Czech Republic (21st December 1995), Hungary (7th May 1996), Poland (22nd November 1996), Korea (12thDecember 1996) and the Slovak Republic (14th December 2000). The Commission of the European Communities takes part in thework of the OECD (Article 13 of the OECD Convention).

NUCLEAR ENERGY AGENCY

The OECD Nuclear Energy Agency (NEA) was established on 1st February 1958 under the name of the OEECEuropean Nuclear Energy Agency. It received its present designation on 20th April 1972, when Japan became its firstnon-European full Member. NEA membership today consists of 27 OECD Member countries: Australia, Austria, Belgium,Canada, Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Japan, Luxembourg,Mexico, the Netherlands, Norway, Portugal, Republic of Korea, Spain, Sweden, Switzerland, Turkey, the United Kingdom and theUnited States. The Commission of the European Communities also takes part in the work of the Agency.

The mission of the NEA is:

− to assist its Member countries in maintaining and further developing, through international co-operation, thescientific, technological and legal bases required for a safe, environmentally friendly and economical use of nuclearenergy for peaceful purposes, as well as

− to provide authoritative assessments and to forge common understandings on key issues, as input to governmentdecisions on nuclear energy policy and to broader OECD policy analyses in areas such as energy and sustainabledevelopment.

Specific areas of competence of the NEA include safety and regulation of nuclear activities, radioactive wastemanagement, radiological protection, nuclear science, economic and technical analyses of the nuclear fuel cycle, nuclear law andliability, and public information. The NEA Data Bank provides nuclear data and computer program services for participatingcountries.

In these and related tasks, the NEA works in close collaboration with the International Atomic Energy Agency inVienna, with which it has a Co-operation Agreement, as well as with other international organisations in the nuclear field.

© OECD 2003Permission to reproduce a portion of this work for non-commercial purposes or classroom use should be obtained through theCentre français d’exploitation du droit de copie (CCF), 20, rue des Grands-Augustins, 75006 Paris, France, Tel. (33-1) 44 07 4770, Fax (33-1) 46 34 67 19, for every country except the United States. In the United States permission should be obtained throughthe Copyright Clearance Center, Customer Service, (508)750-8400, 222 Rosewood Drive, Danvers, MA 01923, USA, or CCCOnline: http://www.copyright.com/. All other applications for permission to reproduce or translate all or part of this book shouldbe made to OECD Publications, 2, rue André-Pascal, 75775 Paris Cedex 16, France.


3

COMMITTEE ON THE SAFETY OF NUCLEAR INSTALLATIONS

The NEA Committee on the Safety of Nuclear Installations (CSNI) is an international committee made up ofscientists and engineers. It was set up in 1973 to develop and co-ordinate the activities of the Nuclear Energy Agencyconcerning the technical aspects of the design, construction and operation of nuclear installations insofar as theyaffect the safety of such installations. The Committee’s purpose is to foster international co-operation in nuclearsafety amongst the OECD Member countries.

CSNI constitutes a forum for the exchange of technical information and for collaboration between organisationswhich can contribute, from their respective backgrounds in research, development, engineering or regulation, to theseactivities and to the definition of its programme of work. It also reviews the state of knowledge on selected topics ofnuclear safety technology and safety assessment, including operating experience. It initiates and conductsprogrammes identified by these reviews and assessments in order to overcome discrepancies, develop improvementsand reach international consensus in different projects and International Standard Problems, and assists in thefeedback of the results to participating organisations. Full use is also made of traditional methods of co-operation,such as information exchanges, establishment of working groups and organisation of conferences and specialistmeeting.

The greater part of CSNI’s current programme of work is concerned with safety technology of water reactors. Theprincipal areas covered are operating experience and the human factor, reactor coolant system behaviour, variousaspects of reactor component integrity, the phenomenology of radioactive releases in reactor accidents and theirconfinement, containment performance, risk assessment and severe accidents. The Committee also studies the safetyof the fuel cycle, conducts periodic surveys of reactor safety research programmes and operates an internationalmechanism for exchanging reports on nuclear power plant incidents.

In implementing its programme, CSNI establishes co-operative mechanisms with NEA’s Committee on NuclearRegulatory Activities (CNRA), responsible for the activities of the Agency concerning the regulation, licensing andinspection of nuclear installations with regard to safety. It also co-operates with NEA’s Committee on RadiationProtection and Public Health and NEA’s Radioactive Waste Management Committee on matters of common interest.


4


5

TABLE OF CONTENTS

VOLUME IPage

Highlights of the meeting 7

Conclusions and Recommendations From General Discussion 9

PART I"Best Estimate" Core Calculations for RIA Energy Deposition in High Burnup Fuel 13

Realistic Analysis of RIA in PWR and BWR 15T.Nakajima, NUPEC, Japan

Pulse Width In a Rod Ejection Accident 33D. Diamond, D. J. Bromley, A.L. Aronson, Brookhaven NL, USA

Pin-by-Pin Best-Estimate Core Calculation for LWR RIAs 45V.Malofeev, Kurchatov Institute, Russian federation

Uncertainties Analysis for Best-Estimate PWR RIA Modelling 57J.C Le Pallec, CEA/Serma,, N. Tricot IPSN/DES, France

Coupled Modelling of Fuel Behaviour, Neutronics, and ThermalHydraulics in Safety Assessments at High Burnups 73S. Kelppe, R. Kyrki-Rajamäki, H.Räty, VTT Processes,K. Valtonen, STUK, Finland

Likely Plant Response to RIA for Sizewell B 87R. Page, NNC Ltd., United KingdomJ.R. Jones, British Energy Ltd., United Kingdom

Representative Core Calculations of RIA for BWR 101L. Heins, Framatome - ANP, Germany

VOLUME II

Part IICurrent and New RIA Safety Criteria, the Technical Background 7

RIA Criteria in Japan 9T.Nakajima, NUPEC, Japan

Review of RIA Safety Criteria for VVER Fuel 21Z. Hózer, L. Maroti, KFKI, Hungary

Burnup Dependent RIA criteria in Switzerland 35W. van Doesburg, HSK, Switzerland


6

An Analysis of the CABRI REP Na Tests 39C. Vitanza, OECD/NEA

Part IIIOngoing RIA Experimental Programmes 59

Main Outcomes from the Cabri Tests Results. 61J. Papin, F. Lemoine, E. Fédérici,IRSN, France

NSRR RIA Tests Results and Experimental Programmes 83T. Nakamura, H. Sasajima, H.Uetsuka, JAERI, Japan

High Burnup Fuel and Cladding Characteristics as RIA Test Initial Condition 97K.Kamimura, NUPEC, Japan

Study of High Burnup VVER Fuel Rods Behaviourat the BIGR Reactor Under RIA Conditions: Experimental Results 115L. Egorova, KIAE, O. Netchaeva, Bochvar Inst., Russian Federation

Impact of Corrosion on Rapid Deformation Capabilities of ZIRLO Cladding. 131V. Grigoriev, R. Jakobsson, D. Schrire - Studsvik AB, Sweden,R. Kesterson, D. Mitchell -Westinghouse, USA,H. Pettersson - Vattenfall Fuel AB, Sweden.

List of Participants 141


7

Part II

Current and New RIA Safety Criteria, the Technical Background


8


9

RIA Criteria in Japan

T. Nakajima

Nuclear Power Engineering Corporation (NUPEC), 17-1, 3-Chome, Toranomon, Minato-ku, Tokyo 105-0001, JapanTel: +81(3)4512-2777, Fax: +81(3)4512-2799, e-mail:[email protected]

Abstract

To assess the Reactivity Insertion Events, Japanese Nuclear Safety Commission (NSC) formulated“Evaluation Guide for Reactivity Insertion Events of Light Water Nuclear Power Reactor Facilities” in1984 based on the data from NSRR experiments. For the reactivity insertion accident (RIA), the followingcriteria are applied.

(1) Peak fuel enthalpy must not exceed the prescribed limit of 230 cal/g.(2) Maximum pressure to the pressure boundary must be lower than 1.2 times of the design pressure.(3) Undue radiation exposure to the neighboring public must be avoided.

Besides, the following additional criteria is applied.(4) The reactor pressure vessel must not be damaged due to shock wave and water hammer generated by

waterlogged fuel rupture.To calculate the number of fuel failures related to criteria (3), the fuel failure thresholds due to high-

temperature rupture and PCMI failure are defined. Related to the criteria (4), mechanical energy generatedby waterlogged fuel rupture is evaluated.

The results of recent reactivity insertion accident(RIA) experiments with high burnup fuel performed inFrance(CABRI) and in Japan(NSRR) indicated that some fuel failed at the lower deposited energy in thefuel than was previously assumed. Taking it seriously, Japanese NSC had assessed the following highburnup issues based on the new data from CABRI and NSRR tests.

(a) Decrease of fuel melting point due to burnup increase, addition of gadolinium, plutonium etc.(b) PCMI failure threshold(c) Mechanical energy generated by fuel particle dispersal at low energy PCMI failures(d) Coolability for debris due to fuel dispersal(e) Fission gas releases

The Japanese NSC revised the licensing criteria for RIA in 1998 based on the assessment of high burnupissues. The enthalpy limit to avoid fuel melting and PCMI failure threshold were revised for high burnupfuels. Also, the evaluation of mechanical energies generated by PCMI failure was added in the safetyassessment of RIA to confirm the integrity of reactor pressure vessel.


10


11


12


13


14


15


16


17


18


19


20


21

Review of RIA Safety Criteria for VVER FuelZ. Hózer, L. Maróti

KFKI Atomic Energy Research Institute, Hungary

ABSTRACT

The recently published Vitanza and KAERI RIA correlations for failure enthalpy have been applied to 20VVER tests. Experimental data from Russian IGR and BIGR reactors have been used. The calculationshave shown that both burnup and cladding oxidation effects must be considered, however the pulse widthdependence of failure enthalpy has not been confirmed.

INTRODUCTION

Experimental data on the fuel failure behaviour under reactivity-initiated-accident (RIA) conditionsproduced in the last decade in French and Japanese test reactors indicated low failure enthalpy for highburnup fuel compared to fresh fuel [1]. However the high burnup was not the only phenomenoninfluencing the fuel failure. The oxide scale on the external surface of the fuel rod, hydrogen content of theZr cladding and the local hydriding seemed also be responsible for the failure at low enthalpy [2].Furthermore, differences have been found between Western design fuel and Russian type VVER fuel. Theburnup dependence of fuel failure for VVER fuel was found to be much less, probably due to the lowoxidation during normal operational conditions compared to other PWRs [3].

RIA CORRELATIONS FOR FAILURE ENTHALPY

Two similar approaches proposing more general RIA correlations have been published recently. C.Vitanza, OECD Halden Project, derived a correlation on the basis of CABRI experimental data andpresented at the IAEA Technical Committee Meeting on Fuel Behaviour Under Transient and LOCAConditions in Halden, September 2001 [5]. Nam, Jeong and Jung, KAERI applied a statistical approach tovarious RIA test data from open literature and published in Nuclear Technology, November 2001 [6]. Bothapproaches intended to produce a simple correlation using the available experimental data and without theneed for additional information on the tested fuel. The produced correlations can be used for thecalculation of traditionally applied fuel enthalpy.

Vitanza correlation

The failure treshold proposed by C.Vitanza is based on cladding deformation. CABRI REP Nadata have been used and fuel failure has been considered as the strain level, which cannot be tolerated bythe cladding. 1% permanent strain was accepted for cladding with ductile mechanical characeristics. Thefailure treshold of embrittled cladding is the onset of permanent strain (0%). The criterion predicts well theCABRI data and Japanese NSRR tests as well. The proposed correlation has the following form [5]:


22

285.0

13.01025

200

−

∆++=

W

OX

Bu

DH F τ (1)

where FH - fuel enthalpy failure limit, cal/gBu - burnup, MWd/kgUD - hoop strain limit, %

τ∆ - pulse width, msOX - oxide layer thickness, µmW - as fabricated cladding thickness, µm

• the calculated failure enthalpy is limited: if FH >200, FH =200

• hoop strain is 1% for ductile and 0% for brittle cladding. Two transition functions are proposed, onewith spalling oxide and one for cladding without spalling oxide layer. The failure strain drops from 1%to 0% as function of oxide layer thickness. For cladding with oxide scale less then 50 µm in both cases1% is applied.

• pulse width is also limited: if τ∆ > 75 ms, τ∆ =75 ms

For generic applications to cases other than the CABRI REP Na tests the effect of initial temperatureshould be considered according to the following form:

)280(85.0

13.025

200 ipF TcW

OX

BuH −−

−

∆+=∆ τ (2)

Correlation (2) predicts the failure treshold in terms of enthalpy increase above the initial insteadof total enthalpy. Ti is the initial temperature in oC.

KAERI correlation

A statistical regression model has been employed by Nam, Jeong and Jung to predict the failureenthalpy of irradiated PWR type fuel rods based on US, Japanese, French and Russian research reactorresults. The failure enthalpy in their correlation is expressed in terms of fuel burnup, oxide thickness andpulse width [6].

)log(41.29076.1774.06.156 τ∆+−−= BuOXH f (3)

where FH - fuel enthalpy failure limit, cal/gBu - burnup, MWd/kgU

τ∆ - pulse width, msOX - oxide layer thickness, µm

In the development of (3) correlation only the failed fuel data have been used and the peak fuelenthalpy has been considered as failure enthalpy. It must be mentioned, that this approach may cause someproblems, as very limited information is known on the precise failure enthalpy in most of the tests. TheCABRI facility is capable of measuring the failure enthalpy, but in other reactors only the peak enthalpy isdetermined, which can be much higher than the failure enthalpy.


23

VVER RIA TEST RESULTS

A large number of RIA experiments has been performed in Russia on the IGR, GIDRA and BIGRreactors in order to study the behaviour of VVER fuel rods [3,7,8,9]. Capsule type experiments werecarried out with fresh and irradiated fuel. Furthermore, some refabricated fuel samples were applied withfresh pellets and irradiated cladding. The effects of energy deposition, pulse width, pressurization of fuelrods were tested. In most of the cases water fill was used, but some experiments were conducted in air aswell.

According to the test results for highly pressurized fuel rods, ballooning was the basic mechanismof cladding failure for both fresh and irradiated fuel. Peak fuel enthalpies, that correspond to the lowerfailure boundary were found to be the same (~160 cal/g) for both fresh and irradiated fuel [3]. Theconducted tests covered a wide range of pulse width, but showed no effect of this parameter on the failuretreshold.

Part of the experiments was collected into a well described database and published in NUREGreports [7]. Table 1. contains the main parameters of tests conducted in water. The experiments with airatmosphere were used to create uniform conditions for later analysis and are not considered prototypicalfor reactor transients. For this reason, the air tests are not listed in the table. The last rows of Table 1.include recently published experimental data from the ongoing research programme on the BIGR reactor[8,9]. The failure enthalpy was not measured during the tests, these values were calculated with transientfuel behaviour codes. In the table only the peak fuel enthalpies are shown.


24

Table 1.

Russian RIA tests with VVER fuel

Test Burnup[MWd/kgU]

Oxidethickness

[µm]

Pulse width[ms]

Peak enthalpy[cal/g]

Failure

IGR experimentsH1T 51 5 800 151 NoH2T 50 8 760 213 YesH3T 50 10 820 212 YesH4T 50 5 760 110 NoH5T 50 8 840 176 YesH6T 50 5 800 87 NoH7T 47 5 630 187 YesH8T 48 5 850 61 NoH14T 0 5 900 61 NoH15T 0 5 900 195 YesH16T 0 5 850 121 NoH17T 0 5 950 91 NoH18T 0 5 850 85 NoH6C 0 5 800 219 Yes

BIGR experiments833 49 5 2.6 142 No834 48 5 3.2 116 No832 48 5 2.6 139 No846 61 5 2.6 120 No847 49 5 2.6 145 No845 48 5 2.6 153 No

Application of RIA correlations to VVER tests

Both the Vitanza and KAERI correlations are based on three parameters: fuel burnup, oxide layer thicknessand pulse width. These parameters were available for the above VVER tests and so correlation (1) and (3)were applied for the calculation of failure enthalpy of the listed experiments. The Vitanza correlation wascalculated without correction for the temperature, therefore, the second correlation (2) was not used. Thesummary of the calculated results is given in Table 2. In the experiments the exact failure enthalpy was notdetermined, only the peak value is known. Therefore, it was not possible to compare directly the calculatedand measured failure limits. However, in checking each test separately the calculated failure enthalpy wascompared with the peak fuel enthalpy. In case of no failure the calculated value was expected to be higherthan the measured peak enthalpy. In the case of failure, the calculated failure enthalpy should be below theexperimental data.

The Vitanza correlation (1) was calculated using 1% hoop strain, 685 µm cladding thickness and75 ms pulse width for the very long IGR tests. For fresh fuel 1 MWd/kgU burnup was applied. Theformula gave very high values for all fresh fuel IGR tests with long pulse width (630-950 ms), the failure


25

enthalpy was limited by the maximum 200 cal/g value. The effect of burnup and oxidation decreased thefailure limit to ~150 cal/g for a 50 MWd/kgU burnup fuel. In the case of BIGR tests the calculated valuesindicated the failure of four rods, while none of the fuel rods were failed in this series of experiments.

The KAERI correlation (2) was used with the experimental data listed in Table 1. It gave veryhigh failure enthalpy for fresh IGR fuel, which was much higher than the peak fuel enthalpy of failed freshrods H6C and H15T. The failure enthalpy was overestimated for the high burnup IGR test H5T as well. Inthe case of BIGR tests the formula indicated failure for all fuel rods, while all of them remained intactduring the test.

Table 2.

Calculated failure enthalpy using correlations (1) and (3)

Test Burnup [MWd/kgU] Failure Peak enthalpy[cal/g]

experiment

Failure enthalpy[cal/g]

correlation (1)


correlation (3)H1T 51 No 151 155.2 183.2H2T 50 Yes 213 156.6 181.3H3T 50 Yes 212 155.8 180.8H4T 50 No 110 157.8 183.7H5T 50 Yes 176 156.6 182.6H6T 50 No 87 157.8 184.3H7T 47 Yes 187 166.3 184.5H8T 48 No 61 163.3 187.2H14T 0 No 61 200.0 239.6H15T 0 Yes 195 200.0 239.6H16T 0 No 121 200.0 238.9H17T 0 No 91 200.0 240.3H18T 0 No 85 200.0 238.9H6C 0 Yes 219 200.0 238.1833 49 No 142 139.0 112.2834 48 No 116 142.0 115.9832 48 No 139 141.9 113.3846 61 No 120 112.3 99.3847 49 No 145 139.0 112.2845 48 No 153 141.9 113.3

The analysis of the results showed that the calculations were too sensitive to the pulse width value.For this reason a constant 75 ms pulse width was selected for all cases and the calculations were repeated.

There are several reasons to remove the pulse width from the proposed correlations. First of all,the RIA tests (Russian and others as well) showed no significant dependence of the failure enthalpy on thepulse width. Furthermore, the peak fuel enthalpy is calculated using the power history over the RIA time,so that the peak fuel enthalpy value already includes the information on characteristic pulse width.


26

The final form of the correlations applied for the VVER tests was the following:

{ }200,85.0

15.221

7000min

2

−

+

+=

W

OX

BuH F (4)

BuOXH f 076.1774.07.211 −−= (5)

Table 3.

Calculated failure enthalpy using correlations (1) and (3) with constant pulse width (75 ms)

Test Burnup [MWd/kgU] Failure Peak enthalpy [cal/g]experiment


correlation (1)


correlation (3)H1T 51 No 151 155.2 153.0H2T 50 Yes 213 156.6 151.8H3T 50 Yes 212 155.8 150.2H4T 50 No 110 157.8 154.1H5T 50 Yes 176 156.6 151.8H6T 50 No 87 157.8 154.1H7T 47 Yes 187 166.3 157.3H8T 48 No 61 163.3 156.2H14T 0 1.1.1.1.1. 61 200.0 207.9

H15T 0 Yes 195 200.0 207.9H16T 0 No 121 200.0 207.9H17T 0 No 91 200.0 207.9H18T 0 No 85 200.0 207.9H6C 0 Yes 219 200.0 207.9833 49 No 142 160.5 155.2834 48 No 116 163.3 156.2832 48 No 139 163.3 156.2846 61 No 120 133.7 142.2847 49 No 145 160.5 155.2845 48 No 153 163.3 156.2

Using a constant pulse width, the agreement between calculated and experimental values wasmuch better than in the first series of calculations. With this simplification both formulae gave higherfailure values than the measured peak enthalpy in tests with no fuel failure. In case of fuel failure, bothcorrelations indicated lower failure enthalpy than the measured value, except for one point H15T, wherethe difference between measured (195 cal/g) and calculated (200 cal/g and 207.9 cal/g) was small, but thecalculated value was higher. The results are listed in Table 3. and shown in Fig. 2-5. The better agreementusing the modified correlations is clearly seen in the figures. The failed fuel figures (Fig. 2-3) represent thecalculated failure enthalpies as a function of measured peak fuel enthalpy. The points are expected to bebelow the y=x curve, if the prediction is correct. In the figures with intact fuel (Fig. 4-5.) the correct valuesshould be above the y=x line (the peak fuel enthalpy for the intact fuel is below the failure limit). It can be


27

seen that the original correlations provided several values below and above the y=x line in both cases,while the correlations with constant pulse width indicated correct failure limit in all cases except for onetest. The two correlations with constant pulse width produced values very close to each other in most of theanalysed cases.

100 150 200100

150

200

250Failed VVER fuel rods

Vitanza correlation (1) KAERI correlation (3)

Pre

dict

ed F

ailu

re E

ntha

lpy

[cal

/g]

Experimental Peak Fuel Enthalpy [cal/g]

Fig. 2.Predicted failure enthalpy by Vitanza and

KAERI correlations vs VVER experimental peakfuel enthalpy for failed rods

100 150 200100

150

200

250Failed VVER fuel rods

Correlation (4) Correlation (5)

Pre

dict

ed F

ailu

re E

ntha

lpy

[ca

l/g]



KAERI correlations with constant pulse width (75ms) vs VVER experimental peak fuel enthalpy for

failed rods

0 20 40 60 80 100 120 140 160 180 200 2200

20

40

60

80

100

120

140

160

180

200

220

240

Intact VVER fuel rods

Vitanza correlation KAERI correlation (P

redi

cted

Fai

lure

Ent

halp

y [

cal/g

]



KAERI correlations vs VVER experimental peakfuel enthalpy for intact rods

0 20 40 60 80 100 120 140 160 180 200 2200

50

100

150

200

250

Intact VVER fuel rods

Correlation Correlation

Pre

dict

ed F

ailu

re E

ntha

lpy

[ca

l/g]



KAERI correlations with constant pulse width (75ms) vs VVER experimental peak fuel enthalpy for

intact rods


28

0 10 20 30 40 50 60 7050

100

150

200

250

300

350

Burnup [MWd/kgU]

Vitanza correlation, 75 ms pulse width, 10 µm oxide KAERI correlation, 75 ms pulse width, 10 µm oxide 140 cal/g limit Intact VVER fuel rods Failed VVER fuel rods

Fue

l ent

halp

y [

cal/g

]

Fig 6.Burnup dependence of VVER fuel failure enthalpy

The burnup dependence of VVER fuel failure enthalpy was calculated using correlations (4) and(5) with a constant oxide layers thickness of 10 µm, which is a typical value for VVER-440 fuel. In Fig. 6.beside the two curves the available experimental data are presented and the 140 cal/g limit is also shown.The 140 cal/g value is used in Hungary as fuel failure criteria in RIA accident, and this value does notdepend on burnup. Fig. 6. shows that the correlations, which proved to be capable to describe the boundarybetween failed and intact fuels, lie above this line and reach the 140 cal/g value at ~60MWd/kgU.

Analysis of non-VVER tests

A less detailed analysis was carried out for US, French and Japanese RIA experiments. Thenecessary data for the correlations (burnup, oxide layer thickness and pulse width) has been collected fromopen literature. Both Vitanza and KAERI correlations were applied in their original form (1) and (3). Incase of Vitanza correlation 1% ductility was applied for specimens with oxide layer thickness below 30µm and 0% was considered for samples with thicker oxide scale. The simplified correlations without pulsewidth (4) and (5) have been calculated as well (with the assumption of constant pulse width of 75 ms). Themain parameters of the tests and the calculated failure enthalpies are summarised in Tables 4. and 5. forintact and failed rods respectively.


29

In case of failed fuel rods, the failure enthalpy calculated according to (1) and (3) showedscattering character around the y=x curve indicating non-failure for several rods, which failed in the tests(Fig. 7.). The use of simplified correlation has not changed the scattered picture (Fig. 8.). The mostsignificant differences between experimental and calculated values were observed for low enthalpy failurecases, like RepNa-1 and HBO-1 tests. In those tests the hydriding of Zr cladding played important role andthis effect was not considered directly in the correlations.

0 50 100 150 200 2500

20

40

60

80

100

120

140

160

180

200

220

240

260

280

300Failed fuel rods

Vitanza correlatio KAERI correlation

Pre

dict

ed F

ailu

re E

ntha

lpy

[cal

/g]



KAERI correlations vs non-VVER experimentalpeak fuel enthalpy for failed rods

0 50 100 150 200 2500

20

40

60

80

100

120

140

160

180

200

220

240

260

280

300Failed fuel rods


Pre

dict

ed F

ailu

re E

ntha

lpy

[cal

/g]



KAERI correlations with constant pulse width (75ms) vs non-VVER experimental peak fuel

enthalpy for failed rods

0 50 100 150 200 2500

20

40

60

80

100

120

140

160

180

200

220

240

260

280

300Intact fuel rods

Vitanza correlatio KAERI correlation

Pre

dict

ed F

ailu

re E

ntha

lpy

[cal

/g]



KAERI correlations vs non-VVER experimentalpeak fuel enthalpy for intact rods

0 50 100 150 200 2500

20

40

60

80

100

120

140

160

180

200

220

240

260

280

300Intact fuel rods


Pre

dict

ed F

ailu

re E

ntha

lpy

[cal

/g]



KAERI correlations with constant pulse width (75ms) vs non-VVER experimental peak fuel

enthalpy for intact rods


30

The application of original correlations to tests with intact fuel rods indicated failure for severalfuel rods above 100 cal/g, which remained intact after the RIA experiment (Fig. 9.). The use of simplifiedcorrelations (4) and (5) improved the agreement between experimental and theoretical values: only a fewpoints remained below the y=x curve (Fig. 10.).

Table 4.

Calculated failure enthalpy by correlations (1),(3),(4) and (5) for failed non-VVER tests

Test Burnup[MWd/k

gU]

Oxidethickness[micron]

Pulsewidth[ms]

Peakenthalpy[cal/g]

Calculated failure enthalpy[MWd/kgU]

(1) (3) (4) (5)PWR PBF tests, US

RIA-ST-2 0 0 17 250 200.0 192.8 200.0 211.7RIA-ST-2 0 0 17 260 200.0 192.8 200.0 211.7801-1 4.6 5 13 285 200.0 180.5 200.0 202.9801-2 4.7 5 13 285 200.0 180.4 200.0 202.8801-3 0 0 13 285 200.0 189.4 200.0 211.7801-5 0 0 13 285 200.0 189.4 200.0 211.7802-3 4.4 5 16 185 200.0 183.4 200.0 203.1804-1 6.1 5 11 277 200.0 176.8 200.0 201.3804-3 5.5 5 11 277 200.0 177.4 200.0 202.0804-7 5.9 5 11 277 200.0 177.0 200.0 201.5804-9 5.7 5 11 277 200.0 177.2 200.0 201.7804-10 4.4 5 11 255 200.0 178.6 200.0 203.1804-4 5 5 11 255 200.0 178.0 200.0 202.5804-6 5.1 5 11 255 200.0 177.9 200.0 202.4804-8 4.7 5 11 255 200.0 178.3 200.0 202.8804-5 5.5 5 11 234 200.0 177.4 200.0 202.0

BWR SPERT tests, USCDC-568 3.5 0 24 165 200.0 193.4 200.0 208.0CDC-567 3.1 0 18 219 200.0 190.2 200.0 208.4CDC-569 4.1 0 14 289 200.0 185.9 200.0 207.3CDC-709 1 0 13 198 200.0 188.3 200.0 210.7CDC-756 32.7 65 17 146 176.6 107.3 141.8 126.3CDC-859 31.8 65 16 158 181.1 107.5 145.1 127.2

CABRI tests, FranceREP-Na1 63.8 80 9.5 30 86.2 54.8 77.5 81.2REP-Na7 55 50 40 120 117.5 105.8 95.9 113.9REP-Na8 60 200 75 82 68.2 -7.6 51.9 -7.6REP-Na10 60 80 31 79 96.5 74.0 81.3 85.3

NSRR tests, JapanHBO-1 50.4 43 4.4 73 121.8 88.0 106.1 124.2HBO-5 44 4.4 79 156.9 128.2 178.1 164.4


31

Table 5.

Calculated failure enthalpy by correlations (1),(3),(4) and (5) for intact non-VVER tests

Test Burnup[MWd/k

gU]

Oxidethickness[micron]

Pulsewidth[ms]

Peakenthalpy[cal/g]

Calculated failure enthalpy[MWd/kgU]

(1) (3) (4) (5)PWR PBF tests, US

RIA-ST-1 0 0 22 185 200.0 196.1 200.0 211.7RIA-ST-3 0 0 20 225 200.0 194.9 200.0 211.7802-1 5.2 5 16 185 200.0 182.5 200.0 202.3802-2 5.1 5 16 185 200.0 182.7 200.0 202.4

BWR, SPERT tests, USCDC-571 4.6 0 31 134 200.0 195.5 200.0 206.8CDC-703 1.1 0 15 159 200.0 190.0 200.0 210.6CDC-685 13.1 0 23 154 200.0 182.6 200.0 197.6CDC-684 12.9 0 20 166 200.0 181.0 200.0 197.9

CABRI tests, FranceREP-Na2 33 4 9.5 209 200.0 146.8 200.0 173.1REP-Na3 52.8 40 9.5 125 84.8 97.6 135.1 124.0REP-Na4 62.3 80 64 96 76.4 80.8 103.5 82.8REP-Na5 64.3 20 9.5 115 103.6 100.7 122.2 127.1REP-Na6 47 40 35 148 101.5 120.5 149.0 130.2REP-Na9 28 10 40 210 200.0 165.8 200.0 173.9

Japanese testsMH-1 38.9 4 6.8 47 175.5 136.1 195.8 166.8MH-2 38.9 4 5.5 54 175.2 133.4 195.8 166.8MH-3 38.9 4 4.5 67 174.9 130.9 195.8 166.8GK-1 42.1 10 4.6 93 159.3 123.1 179.9 158.7GK-2 42.1 10 4.6 90 159.3 123.1 179.9 158.7OI-1 39.2 15 4.4 106 168.3 121.7 188.6 158.0OI-2 39.2 15 4.4 108 168.3 121.7 188.6 158.0HBO-2 50.4 35 6.9 37 90.0 100.0 143.8 130.4HBO-3 50.4 23 4.4 74 129.0 103.5 148.8 139.7HBO-4 50.4 19 5.3 50 130.7 109.0 150.5 142.8HBO-6 49 4.4 79 141.3 122.8 162.5 159.0HBO-7 49 5 88 141.5 124.4 162.5 159.0TK-1 37.8 7 5 126 178.4 131.1 199.0 165.7TK-3 50 5 99 138.8 123.4 159.8 157.9TK-4 50 20 5 98 131.3 107.9 151.1 142.5TK-5 48 25 5 101 134.6 106.2 154.2 140.7TK-6 38 15 5 125 173.6 124.7 193.8 159.2TS-1 26 6 5.8 55 200.0 146.4 200.0 179.1TS-2 26 6 5.2 66 200.0 145.0 200.0 179.1TS-3 26 6 4.6 88 200.0 143.5 200.0 179.1TS-4 26 6 4.3 89 200.0 142.6 200.0 179.1TS-5 26 6 4.6 98 200.0 143.5 200.0 179.1FK-1 45.4 20 5 130 144.1 112.8 164.0 147.4


32

FK-2 45.5 20 5 70 143.8 112.7 163.7 147.3FK-3 41.3 20 5 145 158.0 117.2 177.8 151.8FK-4 56 20 5 140 117.6 101.4 137.4 136.0FK-5 56 20 5 70 117.6 101.4 137.4 136.0ATR-1 20 15 5 80 200.0 144.0 200.0 178.6ATR-2 20 15 5 110 200.0 144.0 200.0 178.6ATR-3 20 15 5 120 200.0 144.0 200.0 178.6ATR-4 20 15 5 140 200.0 144.0 200.0 178.6


33

CONCLUSIONS

The Vitanza correlation (based on French and Japanese tests) and the KAERI correlation (based onUS, French, Japanese and Russian data) with slight modifications proved to be applicable to VVER testsand provided results very similar to each other.

The analysis of 20 VVER fuel tests data using the failure enthalpy correlations, indicated that thiskind of correlation could be successfully used for the calculation of failure threshold during a RIAaccident. The formula reflects the degradation of fuel properties due to burnup and oxidation. The effect ofpulse width was not confirmed. Furthermore, avoiding this parameter from the correlation gave a betteragreement with measured data.

The correlations indicated the decrease of failure enthalpy with fuel burnup and oxide scalegrowth. The application of typical VVER conditions showed that current RIA fuel safety criteria applied inHungary (140 cal/g) is satisfied up to ~60 MWd/kgU.

REFERENCES

[1] R.O. Meyer, R.K McCardell, H.M. Chung, D.J. Diamond, H.H. Scott: A Regulatory Assessment ofTest Data for Reactivity-Initiated Accidents, Nuclear Safety, vol 37, No.4, 1996, pp. 271-288

[2] F. Nagase, K. Ishiyima, T. Furuta: Influence of Locally Concentrated Hydrides on Ductiltity ofZircaloy-4, NEA/CSNI/R(95)22, 1995, pp.433-443

[3] V. Asmolov, L. Yegorova: The Russian RIA Research Program: Motivation, Definition, Execution andResults, Nuclear Safety, vol 37, No.4, 1996, pp. 343-371

[4] An overview of Fuel Safety Criteria Used in NEA Member States, Draft version April 14 2002

[5] C Vitanza: A RIA Failure Criterion Based on Cladding Strain, IAEA TCM on Fuel Behaviour UnderTransient and LOCA Conditions, Halden, 10-14 September 2001

[6] C Nam, YH Jeong, YH Jung: A Statistical Approach to Predict the Failure Enthalpy and Reliability ofIrradiated PWR Fuel Rods During Reactivity-Initiated Accidents, Nuclear Technology, Vol. 136. (Nov.2001)

[7] L Yegorova: Data Base on the Behaviour of High Burnup Fuel Rods with Zr1%Nb Cladding and UO2

Fuel (VVER Type) under Reactivity Accident Conditions, NUREG/IA-0156, NSI RRC KI 2179, 1998

[8] L Yegorova, F Schmitz, J Papin: Mechanical Behaviour of Fuel Element During RIA Transients, Proc.of EUROSAFE, 18-19 November 1999, Paris

[9] Yu Bibilashvili, N Sokolov, O Nechaeva, A Salatov, F Sokolov, V Asmolov, L Yegorova, E Kaplar,Yu Trutnev, I Smirnov, V Ustinenko, V Sazhnov, V Smirnov, A Goryachev: Experimental Study of VVERHigh Burnup Fuel Rods at the BIGR Reactor under Narrow Pulse Conditions, Proc. of Int. TopicalMeeting on LWR Fuel Performance on CD, 10-13 April 2000


34


35

Burnup Dependent RIA Criteria in Switzerland

Willem van Doesburg, Swiss Federal Nuclear Safety Inspectorate (HSK)

ABSTRACT

Since November 1994, the following provisional RIA limits are in place in Switzerland: (a) no fuelmelting, as core coolability criterion, and (b) a burnup dependent max. value for the radially ave. enthalpyincrease (from 125 cal/g at 0 MWd/t to 45 cal/g at 50 MWd/t, the so-called "Swiss curve") as fuel failurecriterion. The industry proposed RIA limits as documented in the1996 EPRI report (235 cal/g coolabilitycriterion and burnup dependent failure limit based on the so-called "Region of Success") were found to beplausible. However, the licensing basis was not changed pending more experimental verification. Today,with a much improved experimental basis and analytical appreciation of the underlying phenomena, HSKconsiders that a less conservative approach is appropriate. The fuel failure and core coolability criteriaproposed in the EPRI/ANATECH topical report 1002865, which was recently submitted to the USNRC,appear to provide a basis for such an approach. HSK is currently reviewing this, as well as supplementaryMOX studies that ANATECH performed for HSK. Additional analysis is being done by PSI &ANATECH for selected Swiss NPPs, as plant specific verification of the proposed RIA limits. The HSKaims at updating the provisional limits by early 2003. In order to ensure the availability of a properexperimental basis for RIA and other safety criteria, especially in view of high burnup, a High BurnupStrategy has been agreed upon between the HSK and the Swiss utilities.


36


37


38


39

An Analysis of the CABRI REP Na Tests

by C. Vitanza, OECD Nuclear Energy Agency, Paris, France

This note intends to contribute to the interpretation of the RIA tests performed during the last decade onBWR and PWR rods and reported in the open literature. A failure criterion based on cladding deformationis proposed. This criterion accurately predicts all CABRI REP Na data with the exception of one (REP Na-1). This exception is discussed in the paper. NSRR fuel failures are also reasonably well predicted by theproposed criterion. Further MOX tests are needed in order to clarify the differences between MOX andUO2. With regard to mechanisms, fuel swelling is important and FGR less important for the failuremechanism at high burnup (~60 MWd/kg).

Content

1. Introduction and summary

2. Cladding strain, tests with UO2 fuel

3. Cladding strain as basis of a failure criterion

4. Inference of a RIA failure threshold

4.1 Correlation for the RIA failure threshold

4.2 Predictability of the REP Na tests

4.3 Generic failure threshold, application to the NSRR tests

5. Fission gas release, tests with UO2 fuel

6. MOX fuel

7. Discussion on Rep Na-1

8. Concluding remarks

Appendix


40

1. Introduction and summary

The CABRI REP Na data available in the literature (Table 1) provide a very consistent basis for theassessment of fuel behaviour mechanisms relevant for RIA transients. For the REP Na rods that did notfail, valuable information on fission gas release and especially on cladding deformation have beengenerated in the test programme.

REP Na cladding strain data have been used as a start point for deriving an admittedly simple RIA failurecriterion. This is based on a maximum cladding strain which can be tolerated by the cladding, i.e. failurecan occur only when this level of strain is exceeded.

� For cladding that still retains ductility, failure is predicted beyond a 1% (permanent) diameter strain.

� For cladding that has been embrittled due to large corrosion and hydriding, a zero ductility is assumed,i.e. the failure threshold is at onset (0%) of permanent strain.

� According to the above, one derives that the lowest failure limit is at 60 MWd/kg is ~65-70 cal/g,which applies to heavily corroded/hydrided fuel, i.e. for oxide thickness of ~80 µm and in the presenceof oxide spalling. For corrosion resistant fuel, i.e. for oxide thickness well below 80 µm and in theabsence of spalling, the failure threshold at the same burnup is 100-120 cal/g.

This failure criterion accurately predicts three of the four failed REP Na tests, but not the REP Na-1 test.For this, the predicted failure threshold is 63 cal/g, whereas the reported experimental value is 30 cal/g.The criterion also applies well to the failed NSRR PWR/BWR rods.

The PIE data of unfailed REP Na rods show a remarkable similarity between the trend exhibited by thecladding strain and by the fission gas release. This is briefly discussed in this paper with regard to apossible FGR mechanism (i.e. it is implied that the gas may mostly be released upon cooling, whencladding constraints are reduced). In any case, it appears that FGR as such has little to do with themechanism of failure when this occurs at low enthalpy.

A comparison MOX versus UO2 fuel based on the REP Na tests is also attempted. The data indicatepossible differences between the two fuels at low burnup (~30 MWd/kg). However, at higher burnup thedifference between MOX and UO2 fuel is not so clear (possibly except for fuel ejection), at least on thebasis of the limited MOX data available so far.

In conclusion, the REP Na test series have proven to constitute a very valuable and consistent data set. Theinformation on cladding strain fits very well together and has enabled to derive a failure criterion, whichpredicts reasonably well three out of four REP Na failures and the NSRR failures. However, the reportedfailure level of the REP Na-1 test cannot be explained based on the present analysis, a point which isdiscussed in the paper.


41

Table 1. The CABRI REP Na tests

Test Rod Pulse(ms)

Energy endof peak(cal/g)

Corrosion(µ)

RIM(µ)

Results and observations

Na-1(11/93)

GRA 54.5% U64 GWd/t

9.5 110(at 0.4 s)

80initial

spalling

200 - Failure, brittle type for HF = 30cal/g.

- Hydride accumulation- Fuel dispersion 6 g., including

fuel fragments outside RIM (>40µ)

- Pressure peaks in Na of 9-10bars

Na-2(6/94)

BR36.85% U33 GWd/t

9.1 211(at 0.4 s)

4 No failure HMAX = 210 cal/gMax. strain: 3,5% average, 3.1%mid-pellet, FGR: 5.5%

Na-3(10/94)

GRA 54.5%53 GWd/t

9.5 120(at 0.4 s)

40 100 No failure HMAX = 125 cal/gMax. strain: 2% FGR: 13.7%

Na-4(7/95)

GRA 54.5% U62 GWd/t

# 75 95(at 1.2 s)

80no initialspalling

200 No failure HMAX = 99 cal/gCladding spalling under transientMax. strain: 0.4% FGR: 8.3%

Na-5(5/95)

GRA 54.5% U64 GWd/t

9.5 105(at 0.4 s)

20 200 No failure HMAX = 115 cal/gMax. strain: 1% FGR: 15.1%

Na-8(07/97)

GRA 54.5%60 GWd/t

75 106(at 0.4 s)

130lim. initial

spalling

200 Failure HF≤ 82 cal/g,HMAX = 110 cal/gno fuel dispersion

Na-10(07/98)

GRA 54.5%62 GWd/t

31 107(st 1.2 s)

80important

initialspalling

200 Failure at HF = 79 cal/g,HMAX = 110 cal/gno fuel dispersalExaminations to be performed

Na-9(04/97)

MOX2 cycles28 GWd/t

34 197 at 0.5s

241 at 1.2s

< 20 No failure HMAX = 210 cal/gMax. strain: 7.4% averageFGR: ~34%

Na 6(03/96)

MOX3 cycles47 GWd/t

35 125 at0.66s

165 at 1.2s

35 No failure HMAX = 148 cal/gMax. strain: 3.2% (2.5% average)FGR: 21.6%

Na 7(1/97)

MOX4 cycles55 GWd/t

40 125 at0.48s175 at1.20s

50 Failure HF= 120 cal/g(t=0.452 s)

Strong flow ejection, pressurepeaks of 200-110b, fuel motion inthe lower half zone


42

2. Cladding strain, CABRI REP Na tests with UO2 fuel

Unfailed UO2 fuel rods tested in the REP Na series have been examined in hot cells after the tests in theCABRI reactor. Through the post-test diameter profilometry one can determine how the diameterpermanent strain at a given axial position depends on the fuel enthalpy deposited at that position during thetransient. This has been done for the UO2 fuel rods and the outcome is shown in Fig. 1.

The burnup effect on cladding strain can be seen in Fig. 1 by comparing curve 2, which refers to a fuel rodof 33 MWd/kg, with curves 3 and 5, which are for 53 and 64 MWd/kg respectively. One can also discern aslight difference between 3 and 5, which is attributed to the burnup difference between these two fuel rods.

The cladding strain as derived in Fig. 1 is substantially greater than what it would be expected from fuelthermal expansion only, typically 2 to 4 times greater. In fact, fuel swelling is believed to be the mostimportant contributor to cladding strain.

The plots in Fig. 1 show that for the high burnup fuel the slope of the strain vs. energy curve tends toincrease gradually, indicating that the fuel swelling might become progressively more pronounced (seecurve 3 and 5). The difference between curve 4 and 5 is ascribed to the difference of pulse width, whichwas larger in test 4 (75 ms versus 9.5 ms).

The Appendix provides examples of PIE diameter profilometry and axial power profile.

3. Cladding strain as basis of a failure criterion

Regardless of the details on the mechanisms involved, the ability of the fuel to withstand a RIA transientdepends on its capability to accommodate cladding strain. The criterion suggested here reflects this, as it isbased on a maximum tolerable cladding strain, i.e. failure can occur for cladding strain exceeding a givenlimit.

� For cladding that has not been embrittled by hydrogen, the (permanent) strain limit is set at 1%. This iscertainly a conservative value, since REP Na tests show that fuel rods with burnup from 53 to 64MWd/kg could tolerate cladding strains in the range 1 to 2% without failing (See REP Na-3 and -5,Fig. 1).

� The REP Na tests provide evidence that large cladding oxidation and presence of oxide spalling mayreduce the cladding ductility due to hydrogen embrittlement, making the cladding more prone tofailure. In this case the threshold of cladding failure is set at onset of permanent strain (i.e. permanentstrain = 0).

The enthalpy failure limits corresponding to the 1% and 0% ductility threshold are derived from Fig. 1 in astraightforward manner. The resulting enthalpy failure limit is shown in Fig. 2. It can be seen that theenthalpy limit decreases with burnup. At 60 MWd/kg the threshold is as low as ~70 cal/g for embrittledcladding and ~110 cal/g for cladding which has no hydrogen embrittlement. As Fig. 2 shows, an upperlimit of 200 cal/g is set on the curves in the low burnup range, since failure mechanisms other than PCMIprevail for high energy depositions.


43

4. Inference of a RIA failure threshold

4.1 Correlation for the RIA failure threshold

A correlation for the fuel failure limit has been derived based on the consideration made in the previoussections and on a closer analysis of the data. The correlation is as follows

285.01 3.0

1025200

−∆++⋅=

WOX

BuD

FH τ /1/

where HF is the fuel enthalpy failure limit, cal/gIf HF from (1) is > 200,. set HF = 200Bu is burnup in MWd/kgD is the (ductility or) hoop strain limit (1% for cladding with residual ductility and 0% for

embrittled cladding), in percent∆τ is the pulse width in ms ( ≤ 75 ms, use 75 ms beyond that)OX is the oxide thickness in µmW is the as-fabricated wall thickness as-fabricated in µm (576 µm for the REP Na rods)

The strain to failure D varies from 1% for ductile cladding to 0% for embrittled cladding. Embrittlementoccurs for large oxide and in presence of spalling. The REP Na tests indicate that for cladding having anoxide thickness of 80 µm, embrittlement (D=0) occurs for spalled cladding oxide, whereas ductility ismaintained (D=1) for un-spalled cladding. Recently published ANL data indicate that cladding ductilitycan gradually decrease to zero also for uniform oxides [1]. These data show that ductility decreases withincreasing thickness of the peripheral hydride layer, approaching zero when this thickness goes beyond~100 µm. For the purpose of this analysis, these findings have been provisionally converted in decreasingductility vs. oxide thickness. The resulting picture is given in Fig.3, which shows how ductility drops tozero for spalled oxide (inferred from the failed REP Na rods) and for un-spalled oxide (inferred from [1]).One should notice that in most practical cases one does not know a-priori if the cladding is spalled or not,except that spalling may occur for thick oxide. For such cases, the use of the left curve of Fig. 3 isconservatively recommended.

Fig. 3. Suggested dependency of the term D on cladding oxide

0

0,5

D,%

1% D=1

D=0

oxide thickness, � m

50 100

D=1

150

with spalling

without spalling

D=0


44

The oxide thickness in equation /1/ is important not only because it affects the ductility term D. Oxidationalso causes a reduction of cladding metal wall thickness and an increase of hydride formation at thecladding periphery, both of which lower the constraint posed by the cladding on the swelling fuel. This isaccounted for by the last term of equation /1/. In practice, the above means that low or moderately oxidisedfuel will fail at appreciably higher enthalpy than heavily oxidised fuel.

In addition to oxide thickness, burnup is very important. Swelling during both normal operation and duringa RIA transient becomes more pronounced with burnup, causing larger cladding strains. This isacknowledged in equation /1/ by the inverse relation between failure enthalpy and burnup.

Since the oxide thickness is also a function of burnup, at the end of the day the failure threshold can beexpressed only in terms of burnup, once the relation between oxide thickness and burnup can beanticipated. Two practical examples are shown in Fig. 4, which gives the failure limits for two types of fuelcladding, one having significant corrosion and spalling, the other one having moderate corrosion. Thepredictions give a threshold that is ~40 cal/g lower for the more oxidised/spalled cladding. (The cases ofFig. 4 are only meant as examples, as oxide dependency on burnup may be different for differentreactors/materials). For cladding having oxide thickness larger than 130 µm, the predicted failure thresholdat 60 MWd/kg is lower than ~55 cal/g.

In this evaluation, burnup and oxide thickness are kept as independent, separate variables that affect failurelevels in the way expressed by Eq. /1/. They are considered as separate variables because the corrosion-burnup relation may vary substantially from case to case, depending on cladding alloy, water chemistryand fuel duty during service in a commercial reactor. It is possible that some sort of duty index can be usedas better parameter to predict fuel failure enthalpy, but such refinements are beyond the purpose of thisnote. Burnup and oxide thickness are used here simply because they are parameters of common use whichhave a direct physical meaning– in other words, everyone knows what these parameters represent.

Equation /1/ acknowledges only a moderate effect of pulse width (∆τ). In fact, the predicted failurethreshold increases by only ~5 cal/g when the pulse width increases from 9 to 30 ms. Since in the REP Naseries the ∆τ ranged between 9 and 75 ms, a maximum value ∆τ = 75 ms should be used when the pulsewidth exceeds 75 ms.

4.2 Predictability of the REP Na tests

The calculated enthalpy to failure based on equation /1/ for the various REP Na tests is given in Table 2 onthe next page.

For the rods that had failed, the comparison between predicted and actual failure enthalpy is shown in Fig.5. Three tests are accurately predicted by Eq. /1/, but REP Na-1 is not. In this case the failure is predictedat 63 cal/g, whereas the quoted enthalpy to failure is a factor of 2 lower.

A prediction of the failure limit has also been done for the REP Na tests that did not fail. The results areshown in the diagram of Fig. 6. One can observe that all the data points lie rather close to the 1:1 diagonal,typically ~5-10 cal/g above it. This confirms that, with reference to the REP Na-1 database, the proposedcorrelation is conservative by ~5-10 cal/g, i.e. not an "unreasonably" conservative one.

4.3 Generic failure threshold, application to the NSRR tests

While Eq. /1/ already contains a reasonable degree of conservatism, the correct setting of the parameter Dmust be considered further for cases different than the REP Na tests. As said earlier, in REP Na tests it isapparent that an inverse relation does exist between oxide thickness/spalling and cladding ductility.


45

However, other factors than oxide morphology may also contribute to ductility degradation for fuel andconditions different than the ones of the REP Na tests. The power reactor coolant temperature, for instance,may affect the residual ductility in that it affects the balance between irradiation induced embrittlement andtemperature induced annealing. Changes of cladding type may also result in different residual ductility athigh burnup, as affected by irradiation as well as by hydrogen content and distribution in the cladding.Further, the test temperature conditions may affect ductility.

For these reasons, a failure criterion of more generic applicability must necessarily have a greater degree ofconservatism than the one given by Eq. /1/. In practice, this means that a more generic (and conservative)failure criterion can be obtained by using the most conservative value for the cladding ductility parameterD, i.e. D=0.

For applications to other cases than the REP Na tests, another point to consider is the temperature orenthalpy at the start of the test. In Eq. /1/, HF is the total enthalpy to failure, with a start enthalpy level of[Cp ⋅ 280oC] = 20 cal/g (1). For generic applications to a lower (or different) initial temperature, such as inthe NSRR tests, the failure threshold should be expressed in terms of enthalpy increase above the initialvalue. Thus for those generic applications to cases other than the REP Na ones, the failure threshold shouldbe expressed by

gcalW

OX

BuH F /20

85.01 3.0

25200

2

−

−

∆+⋅=∆ τ /2/

which is equivalent to Eq. /1/ except that D is set D=0 and the enthalpy increase ∆HF above the initialinstead of total enthalpy.

The performance of Eq. /2/ is predicting the NSRR PWR and BWR tests where failure occurred is shownin Table 3.

____________________

(1) Cp = 0.30 J/goC or Cp = 0.072 cal/goC


46

Table 2. Failure predictions of REP Na tests based on Eq. /1/

Test ∆τ Bu OX D Experimentfuel

enthalpy

HF (Eq. 1)

REP Na -1 9.5 64 80(spalled)

0 *HFail = 30 63

REP Na-2 9.1 33 4 1 HMAX = 210 200

REP Na-3 9.5 53 40 1 HMAX = 125 119

REP Na-4 75 62 80 1 HMAX = 99 105

REP Na-5 9.5 64 20 1 HMAX = 115 107

REP Na-6 35 47 35 1 HMAX = 148 142 MOX

REP Na-7 40 55 50 1 *HFail = 120 120 MOX

REP Na-8 75 60 130(spalled)

0 *HFail � 82 70

REP Na-9 34 28 20 1 HMAX = 210 200 MOX

REP Na-10 31 62 80(spalled)

0 *HFail = 79 71

Table 3. Comparison of the NSRR PWR/BWR Experimental Enthalpy

at Failure with the Values Predicted by Eq. /2/

Test ∆τ Bu OX Exp. ∆H atfailure

∆HF

(Eq. 2)Fuel type

NSRR

HBO-1 5 50 48 60 67 PWR

HBO-5 5 44 60(1) 77 75 PWR

TK-2 5 48 35 60 73 PWR

TK-7 5 50 30 86 73 PWR

NSRR/BWR

FK-6 5 61 ~20 70 59 BWR

FK-7 5 61 ~20 62 59 BWR

FK-9 5 61 ~20 86 59 BWR

(1) Maximum value


47

5. Fission gas release, tests with UO2 fuel

An evaluation has been made in order to derive the dependency of FGR on enthalpy, on the reasonableassumption that the release depends mainly on fuel enthalpy and burnup. The resulting FGR versusenthalpy curves for REP Na rods having different burnup are shown in Fig. 7. Not surprisingly, one canobserve that the release increases with fuel burnup and enthalpy. By comparing this figure with Fig. 1, onecan also note that the onset of FGR is very close to the onset of permanent deformations.

As an additional exercise, the enthalpy at 5% FGR has been derived for the four curves given in Fig. 7, andthen plotted as a function of the corresponding fuel burnup, as shown in Fig. 8. In the same figure, theonset of a cladding permanent strain curve has also been plotted. One can again note that the 5% FGRcurve is remarkably close to the onset of cladding strain curve, which was derived in Fig. 2.

Fig. 8 basically says that appreciable fission gas release is observed beyond the onset of a permanentcladding strain. Since the latter is also the failure threshold for brittle cladding, it follows that FGR"occurs" beyond the low-enthalpy failures (i.e. those related to brittle cladding). Said in other terms, whilefission gas induced fuel swelling is important, FGR has little to do with the failure mechanism for theselow enthalpy failures. Fission gas release - more than the cause of clad diameter deformation - seems to bethe consequence of it. This can be rationalised, for instance, if the gas release takes place mainly in thecooling phase, i.e. after the cladding has been deformed and, upon cooling, does not provide any moreconstraint to the fuel.

6. MOX fuel

The diameter strain for the two unfailed MOX tests are compared with the diameter strain of the unfailedUO2 tests in the diagram of Fig. 9. Although this is only an indicative comparison, one can observe that thediameter strain observed in REP Na-6 are comparable with the ones observed for UO2 fuel at high burnup.Instead, REP Na-9 gave larger deformations than the UO2 test at corresponding burnup (REP Na-2). Itshould be noted, however, that the REP Na-2 fuel was of a quite different source than the rest of the REPNa series. Fig. 10, where the FGR of MOX and UO2 fuel are compared, gives approximately the samepicture, i.e. the MOX REP Na-6 FGR is compatible with UO2 fuel, whereas the REP Na-9 gives a higherrelease than REP Na-2.

In conclusion, the high burnup MOX data (REP Na-6) on cladding strain and FGR are comparable withUO2 high burnup data. Whether the differences observed at lower burnup between REP Na-9 and 2 aredue to the MOX vs. UO2 fuel differences or to other factors (e.g., the REP Na-2 had BR-3 fuel), cannot beconcluded at this time.

Based on the microstructural differences between MOX and UO2 fuel, one would expect that appreciabledifferences exist in the failure behaviour of the two types of fuel. However, the data available on MOX,shown in Fig. 5 and Fig. 6, are too few to ascertain of any clear differences between MOX and UO2 failurepropensity, at least in terms of failure predictability based on Eq. /1/, and more data are needed in order todraw firm conclusions.


48

7. Discussion on REP Na-1

The REP Na-1 test started at a steady temperatureof 280oC, corresponding to ~20 cal/g. The tests,which were to achieve 110 cal/g, resulted in a fuelfailure that reportedly occurred when the transienthad just started. At that point-in-time, only 8.9cal/g had been injected into the fuel, bringing thetotal enthalpy to (~20 + 8.9) = ~30 cal/g. Fuelfailure at such conditions is difficult to explain forthe reasons outlined below.

A ∆H = 8.9 cal/g means an average fueltemperature of ~400oC. Fuels are not expected tofail in such conditions, also considering that in thiscase it is likely that the fuel was exposed to morechallenging conditions for a long time duringservice in the power reactor. The fact that thetemperature was higher than 400oC in the pelletrim does not change the substance of thisobservation.

The CABRI-PROMETRA tests show that an UTS of about 600 MPa was reached in cladding specimensrepresentative of REP Na-1 fuel. This means that REP Na-1 cladding was able to withstand appreciablestress. A failure at ∆H = 8.9 cal/g, instead, would imply a cladding not able to withstand any stress. In fact,the expected stress at ∆H = 8.9 cal/g is less than:

YSgcal

gcalYS 2.0

/)2065(

/9.8 =−

=σ (65 cal/g is the onset of permanent strain, 20 cal/g is the start of the transient)

It is difficult to reconcile such low "tolerable" stress (less than 20% of yield stress) with the results of thePROMETRA tests.

At a total enthalpy of ~30 cal/g the fuel is far below the onset of permanent deformations and of fission gasrelease. In fact, the other REP Na tests show that the onset of permanent deformation and FGR is beyond65-70 cal/g (Figs. 1 and 7).

It is reported that the failure in the test in question resulted in almost immediate fuel ejection due to fissiongas pressure (and in flow disturbance due to fission gas emission). Yet, the other REP Na data show thatthe quoted failure level for REP Na-1 was still far away from FGR onset.

It has been speculated that REP Na-1 failed at very low enthalpy because of the very fast transient in thatcase (combined with spalled cladding). However, the faster part of the transient hadn’t started yet at thereported failure time. The figure in the previous page shows that the derivative of the energy-time curve upto the quoted failure time is still low (approximately 1/3 of the maximum derivative).

0.04 0.06 0.08 0.1

Failure reported here

Co

re e

ner

gy

dep

osi

t

Tim e (s/TOP onset)

REP Na-1


49

A CABRI TAG Task Force is dealing with this, i.e. with the interpretation/explanation of the REP Na-1test, and two issues have been considered in this connection. One is that "something" might have occurredduring the test pre-conditioning, which considerably weakened the cladding. A second consideration wasthe experimental timing of the failure in REP Na-1. As to the latter, one should note that in REP Na-1 thetotal energy deposition was 10 times greater than the quoted level at failure (1) and that this happened inonly 9 ms. It follows that the outcome of the test is crucially dependent on an unquestionably exact timingof the failure event.

8. Concluding remarks

� The expected failure limit based on Eq. /1/ for BU = 60 MWd/kg and oxide thickness 30-40 µm is~100-110 cal/g and this should be conservative values. It follows that, for corrosion resistant cladding,one expects low failure probability unless the achievable fuel enthalpy is well above 100 cal/g. If thisis not the case, one may learn too little on what causes failures (in corrosion resistant fuel) and onrelated failure margins.

� UTS and TE (+ YS) are key parameters for RIA failure propensity. The YS and UTS are expected tobe relatively close to each other and basically provide the measure for the cladding ability to withstandstress. The tensile elongation provides an indication of cladding residual ductility. Large UTS (>600MPa) and TE (> some percent) indicate a good capability of the cladding to withstand RIA transients.Having said this, one should be aware that these basic mechanical properties have to be considered inthe context of a more complex picture. Fuel-cladding bonding, for instance, may create localised stressand strain well above uniform circumferential values.

� Testing at intermediate burnup may seem of academic interest, but the failure threshold dropssubstantially beyond 30 MWd/kg, while the fuel still retains a relatively large reactivity. The databaseat 30-40 MWd/kg seems too limited at present.

� There are too few MOX tests available to enable to point out the clear differences in MOX versus UO2

failure propensity. More data are needed on this subject.

� A study on FGR versus swelling (on which of the two comes first and determines the cladding stress)seems of little or no relevance, at least for explaining the worst failures, i.e. those occurring at lowenthalpy (~60-80 cal/g). As said in this report, these failures occur before FGR comes into the picture.

_______________________

(1) (110 - 20) = 90 cal/g versus 8.9 cal/g


50

� The on-line detection of failure should be based on reliable and acknowledged methods. The fact thatthe CABRI WL tests will be run with ∆τ = ~30 ms will enhance the accuracy of the time-to-failuredetermination (with respect to faster transients).

� There are indications that the CABRI axial power profile, especially at the top/bottom end, should bere-visited.

Reference

[1] R.S. Daum et al., "Embrittlement of Hydrided and Irradiated Zircaloy-4 Under RIA-typeConditions". 13th ASTM Int. Sump. on Zirconium in the Nuclear Industry", Annecy, France (2201).


51

Appendix A1.

Details on profilometry

Fig. A1 and A2 show the axial enthalpy profile (based on calculations) and the experimental diameterstrain versus height for two of the UO2 REP Na tests. These data have been used to derive the strain versusenthalpy curves of Fig. 1. One should note that the power profile seems to become less reliable towards therod lower/upper end. For this reason data were taken only in the region ÷200 to +150 mm around the peakzone.

As a confirmation that the plots shown in Fig. 1 are realistic, the shaded area in Fig. A3 shows a coarserendition of the maximum strain data as reported in Table 1. One can see that the data as plotted in Fig. 1are somewhat more conservative in terms of enthalpy at onset of cladding strain - besides being moreprecise.


52

Fig. 1. Permanent diameter strain versus enthalpy as derived from the post-test rodprofilometry of the four REP Na UO2 fuels that did not fail. One can see that thecurves tend to shift towards left at increasing burn-up. The pulse width also has an effect as shown by the comparison of curve 4 and 5.

Fig. 2. Failure threshold derived from the experimental cladding strainplots (in Fig.1). The upper curve is the failure limit for cladding thatstill retains ductility (1% ∆D/D permanent). The lower curve is forembrittled cladding (0% ∆D/D). For fuel at burnup of 60 MWd/kg, the

200

100

0 0 10 20 30 40 50 60 70

Based on 1% strain

Based on zero strain

RIA

fai

lure

th

resh

old

, cal

/gr

Fuel burn-up, MWd/kg

2

3

0 0 50 100 150 200

5 Burnup 64 MWd/kg

3 Burnup 53 MWd/kg

4 Burnup 62 MWd/kg

2 Burnup 33 MWd/kg

4 Pulse 75 ms, for all others 9 ms

= REP Na-X

' D,% D

Enthalpy at peak, cal/gr

2

3

5

4

1%

0%


53

50

100

150100

50

00 10 20 30 40 50 60 70

00 10 20 30 40 50 60 70

2

1

1

2

Burn-up, MWd/kg

Oxide thicknessvs burn-up

12

Cladding with significant oxidation

Corrosion resistant cladding

Burn-up, MWd/kg

ox, µ

m

Fai

lure

th

resh

old

, cal

/gr

200

Fig. 4. RIA failure threshold as predicted by Eq./l/in the case of a cladding exhibitingsignificant oxidation and spalling (curve 1) and in case of a corrosion resistantcladding (curve 2). At 60 MWd/kg, the predicted enthalpy-to-failure thresholdis~40cal/gr higher for the corrosion resistant cladding (70 vs. 110 cal/gr).

Fig. 5. Comparison of calculated vs experimental enthalpy to failure for the four REPNa tests that failed. Three tests, including a MOX test, are well predicted. The

0

50

100

0 50

R E P N a-8

(M O X )

Exp

erim

enta

l H F

, cal

/gr

C alcu la ted H F , ca l/g r

R E P N a-1

R E P N a-10

R E P N a-7

100


54

FG

R, %

R E P N a - 5

R E P N a - 4

R E P N a - 3

R E P N a - 2

F G R = 5 %

R e p N o 5 6 4 M W d /k gR e p N o 4 6 2 M W d /k gR e p N o 3 5 3 M W d /k gR e p N o 2 3 3 M W d /k g

E x p e r im e n t a l d a taI n fe r r e d

0 5 0 1 0 0 2 0 0

1 0

2 0

F u e l e n t h a lp y , c a l /g r

Fig. 6. Comparison between the calculated enthalpy to failure and the actuallyachieved fuel enthalpy for the REP Na tests that did not result in fuelfailure.

Fig. 7. Fission gas release (FGR) data, plotted as function of the fuelenthalpy, for the UO2 REP Na tests.

0

50

100

150

200

250

0 50 100 150 200 250

REP Na-4

REP Na-5

REP Na-3

REP Na-6

REP Na-2,9

MOX

MOX

Ach

ieve

d f

uel

en

thal

py,

cal

/gr

Calculated H F , cal/gr


55

200

Burnup, MWd/kg

10 30 40 50 60 700

50

100

150

200

Onset of claddingpermanent strain(0% strain)

5% FGRthreshold

En

thal

py,

cal

/gr

Pla

stic

dia

met

er s

trai

n,%

Enthalpy, cal/gr

At (0.5 s)

At HMAX

At (1.2 s)

REP Na-64 7 MWd/kg

REP Na-3,4,553-64 MWd/kg

0

1

2

3

4

5

REP Na-928 MWd/kg

REP Na-233 MWd/kg

0 100 200

6

7

Fig. 8. Plot of the 5% FGR threshold derived from the previous figure(Fig. 7). Together with it, the curve for onset of cladding plasticdeformation is also plotted. The latter is identical to (the lowercurve of ) Fig. 2 and is derived from the cladding strain data shownin Fig. 1. The trend of the two curves is remarkably similar.

Fig. 9. Comparison of the percent diameter strain in UO2 and MOX fuel. TheREP Na-6 test gave strains comparable with the UO2 tests at highburnup. The REP Na-9 MOX test strains appreciably greater than theREP Na-2 UO2 (test at comparable burnup). Note however that the REPNa-2 was a "special" fuel.


56

Fig. 10. Comparison of the fission gas release UO2 and MOX fuel. The REP Na-6 MOXtest (47 MWd/kg) gave FGR compatible with the REP Na-3 UO2 test (53MWd/kg). The REP Na-9 MOX test (28 MWd/kg) gave much higher gas releasethan the REP Na-2 UO2 test (33 MWd/kg)-Note however that the latter has a“special” type of fuel.

At (0.5 s)

At H MAX

At (1.2 s)

0

10

0 200 100

UO 2 33 MWd/kg

MOX 28 MWd/kg

Fis

sio

n G

as R

elea

se, %

Enthalpy, cal/gr

20

30

40

UO 2 53 MWd/kg

MOX 47 MWd/kg

UO 2 62-64 MWd/kg


57

Strain profile

Energyprofile

REP Na-3

Hei

gh

t, m

m

FU

EL

CO

LUM

N

usef

ul a

xial

reg

ion

(

50 m

m)

Energy, cal/gr

0,4 0,6 0,8 1,0 1,2 1,4 1,6 1,8

strain,%

70 80 90 100 110 120 1300

100

200

300

400

Fig. A2. Diameter strain versus height and energy axialprofile in the REP Na-5 test.

Fig. A1. Diameter strain and energy axial profile for the REP Na-3 test.

Strain profile

Energy profile

REP Na-5

Hei

gh

t, m

m

FU

EL

CO

LUM

N

usef

ul a

xial

reg

ion

(350

mm

)

Energy, cal/gr

0,2 0,4 0,6 0,8 1,0

strain,%

0 1,2

50 60 70 80 90 100 110 0

100

200

300

400

500

600


58

REP Na-3

REP Na-5

REP Na-4

Dia

met

er s

trai

n, %

Slope of

REP Na-2

(ax, profile)

Enthalpy, cal/gr

.

Quoted data (Table 1)Authors checkBased on axialprofilometry (Fig.1)Coarse rendition of dataat peak position

UO2 rods only

All non-failedREP-Na rods above47MWd/kg

0 50 100 1500

1

2

3

Fig. A3. Comparison of the strain versus enthalpy trend derived from profilometryand from a coarse rendition of the maximum strain data reported foreach of the unfailed UO2 rods.


59

Part III

Ongoing RIA Experimental Programmes


60


61

Main Outcomes from the Cabri Tests Results

J. Papin, F. Lemoine, E. Fédérici

Département de recherches en SécuritéInstitut de Radioprotection et de Sûreté Nucléaire (IRSN)

CE Cadarache, BP n°3, 13115, St Paul lez Durance - France

1. THE CABRI REP NA PROGRAMME

The evolution of the fuel management strategy in the French PWRs with the increase of the UO2 fueldischarge burnup and the introduction of the MOX fuel, created the need for new investigation of fuelbehaviour under reactivity initiated accidents (RIAs) as resulting from control rod ejection.

In this framework, the “Institut de Protection et de Sûreté Nucléaire” (now IRSN) initiated in 1992 theCABRI REP Na research programme [1] [2] [3] which was conducted in collaboration with Electricité deFrance (EDF) and with the participation of the US NRC.

The main objectives were to study the behaviour of high burnup UO2 fuel (including advanced fuel) andMOX fuel, verify the adequacy or modify the present safety criteria previously defined for lower burnupfuel and to evaluate the safety margins.

As a first step of investigation in France, the CABRI-REP Na experimental programme was launched inthe sodium loop of the CABRI reactor [4] in which the consequences of a fast power transient applied to asingle rod could be studied in a sodium coolant environment. Due to this last point, the investigation wasfocused on the first phase of the power transient when strong pellet-clad mechanical interaction (PCMI)occurs with limited clad heat-up and was devoted to the study of rod failure mechanism and onset of fueldispersal, if any.

In parallel to the REP Na experiments, the SCANAIR code [5] was developed in order to interpret the testresults, perform sensitivity studies and translate the results to reactor conditions.

Separate-effects tests are also being conducted for the study of the cladding mechanical properties(PROMETRA ), the clad-fluid heat transfer under fast transients (PATRICIA) and the transient behaviourof fission gases (SILENE-RIA).

From 1993 to 1998, eight tests with UO2 fuel and four tests with MOX fuel were performed using mostlyrefabricated rods from PWR fuel, with low internal pressure (0.3 Mpa He pressure, consistent with sodiumchannel pressure). The coolant conditions were : inlet temperature of 280°C, fluid velocity of 4 m/s andchannel pressure of 3b.

The following parameters were studied :

- rod burn-up from 33 to 64 GWd/t,

- cladding material : standard Zircaloy-4 (Zr-4), M5

- cladding corrosion thickness of Zr-4 from 4 to 130 µm of ZrO2,


62

- corrosion conditions from uniform to spalled with hydride blisters resulting from reactor operation,

- power transients starting from almost zero power to different energy levels and with various pulsewidths leading to different energy injection rates.

As a first outcome, the REP Na tests which led to rod failure at fuel enthalpy levels (radial average) fromaround 30 cal/g (REP Na 1) to 120 cal/g (REP Na 7) underlined the lack of adequacy of the present safetycriteria for high burnup UO2 and MOX fuel.

A similar conclusion was derived from tests performed in the NSRR facility in Japan and showing a lowenthalpy failure level of high burnup UO2 fuel.

On the other hand, the detailed analysis and interpretation of the first REP Na tests allowed for theidentification of the physical mechanisms and the key parameters influencing the fuel behaviour.

A deleterious effect of a large clad corrosion level with spalling and hydride concentration, reducing theclad ductility (also confirmed by the PROMETRA mechanical testing) has been revealed. On the otherhand , moderate or large corrosion without spalling (up to 80 µm) may withstand significant clad straining,which is obtained as a result of fuel thermal expansion and fission gas swelling and depends on the level ofenergy injection.

Transient spalling of the oxide layer has been evidenced as a result of clad straining and is enhanced byazimuthally non-uniform initial oxide thickness and associated ovalizing : such transient spalling may leadto earlier boiling crisis occurrence in PWR conditions.

The contribution of fission gases on clad loading, in addition to the classical fuel thermal expansion effect,has been underlined [6]: such gas contribution and fission gas release are increased with burnup and in caseof MOX fuel due to its non-homogeneous, structure with UPuO2 agglomerates.

In the case of high energy injection (leading to mean fuel enthalpy above 130-150 cal/g), the contributionof the intra-granular gas swelling on clad loading has been evidenced and confirmed by hydrostatic densitymeasurements.

On the other hand, extensive fuel fragmentation (grains separation) has been observed in most of the REPNa tests. This phenomenon is attributed to the high overpressure that is developed in the small inter-granular and porosity bubbles during fast heating rates and which induces high stress fields between thegrains, leading to the grain boundaries cracking. Subsequent grains separation depends on the respectiveinfluences of gas pressure and external fuel constraint. Largely observed in UO2 fuel, it appears also clearlyin the fuel matrix with MOX tests, in spite of the relatively low burnup level. The main consequences ofthis phenomenon are a degradation of fuel mechanical properties, the fast availability of all the grainboundary gases with associated driving pressure leading to solid fuel pressurization and swelling, cladloading with risk of failure and finally to gas release.

Due to the high gas content in inter-granular and porosity bubbles associated to agglomerates behaviourunder irradiation, solid fuel pressurization and clad loading may be increased in MOX fuel and mayexplain the failure of REP Na 7 suggesting a high burnup effect with MOX fuel [7].


63

2. The SCANAIR code

The objective of the SCANAIR code is to simulate the thermo-mechanical behaviour of a PWR rod (UO2,MOX), especially at high burnup level, under RIA conditions.

It is developed in close link with the interpretation of the CABRI-REP Na and future CABRI –WATERLoop programmes and has to be validated on global tests (CABRI, some NSRRS tests) as well as onseparate effect tests (PATRICIA thermo-hydraulic experiments, SILENE-RIA future tests, …) for correcttranslation to reactor case.

The main characteristic of the SCANAIR code is to be able to deal with intimately coupled phenomenaoccurring during rapid power transients such as thermal, mechanical and fission gas aspects. It includes a1.5D modelling, starting with an initial rod state given by an irradiation code and a power pulse providedas data.

At the present time, a satisfying status of validation against the REP Na tests has been obtained.

3. The future CABRI WATER LOOP Programme

Although the CABRI-REP Na program has been the basis for the EDF authorization to increase the fuelburnup up to 52 GWd/t (mean assembly), questions still remain about the impact of the transient fissiongas behaviour on clad loading for high burnup UO2 and MOX fuel, the rod behaviour after boiling crisis,the influence of internal rod pressure, and the post-failure phenomena with the possibility of -coolantinteraction with. fragmented solid fuel.

The complexity of these phenomena and their important coupling make it difficult to have confidence inthe current results without experimental confirmation with integral tests under representative PWRconditions. Moreover, the lack of representativity was also identified in separate effect tests such as thePROMETRA mechanical program under RIA conditions, preventing any reliable prediction of the rodbehaviour on the sole basis of such experiments at the present time. Prototypical PWR conditions areparticularly important for the qualification of any further increases in fuel burnup in power reactors.

These are the reasons why IRSN has decided to replace the present sodium loop in Cabri with apressurized water loop (PWL) and to propose an international programme called Cabri InternationalProgramme (CIP).

Twelve tests have been proposed for the CIP program and they include high-burnup fuel tests combinedwith mechanical tests to provide the understanding necessary to extrapolate to a broad spectrum of reactorconditions. Several test series have been identified :

CIP0 two tests in the sodium loop using advanced fuels (Zirlo and M5 cladding)

CIPQ qualification test of the water loop (rod with Zr-4 cladding)

CIP1 reference tests in the water loop with the same advanced fuels as in CIP0 toprovide a link to REP-Na tests,

CIP2 tests with ultra high burnup UO2 fuel (80-100 GWd/t)

CIP3 tests specifically designed to improve the understanding of RIA phenomena

CIP4 tests with high burnup MOX fuel

CIP5 complementary tests (open)


64

These integral tests will be coupled with separate-effect tests (mechanical testing, fission gas behaviourexperiments) and code development to facilitate translation to power reactor conditions, development ofsafety criteria or limits and evaluation of the safety margins.

The CIP program is now in its active phase, the CIP0 test series being scheduled before end 2002.

References :

[1] “French Studies on High-Burnup Fuel Transient Behavior under RIA Conditions,” J. Papin et al,Nuclear Safety, Vol. 37, 1996, pp. 289-327.

[2] “High burn-up effects on fuel behavior under accident conditions : the tests CABRI REP Na” F.Schmitz and J. Papin, J. Nucl. Mater., 270 (1999) 55-64

[3] “Further results and analysis of MOX fuel behavior under reactivity accident conditions inCABRI” J. Papin, F. Schmitz, B. Cazalis, 27th WRSM, October 1999, Bethesda, USA

[4] “SURA : a Test Facility to Investigate the Safety of LMFBR and PWR Fuels,”C. Marquie et al, IAEA International Symposium on Research Reactor Utilization,

Lisbon, 1999.

[5] “Status of development of the SCANAIR code for description of fuel behaviour under reactivityinitiated accident”, E. Federici, F. Lamare, V. Bessiron, J. Papin, International Topical Meeting onLight Water Reactor Fuel Performance, Park city, Utah, USA, du 10 au 13/04/2000

[6] “The role of fission gases on the high burnup fuel behaviour in reactivity initiated accidentconditions”, F. Lemoine – B. Cazalis – H. Rigat, 10th international Symposium onThermodynamics of Nuclear Materials, Halifax, 6 – 11/08/2000

[7] “The MOX fuel in the CABRI REP Na programme : analysis and main outcomes”, B. Cazalis, F.Lemoine, J. Papin, International Topical Meeting on Light Water Reactor Fuel Performance, Park city, Utah, USA, du 10 au 13/04/2000

NE

A/C

SNI/

R(2

003)

8/V

OL

2

65

MA

IN O

UT

CO

ME

S FR

OM

TH

E C

AB

RI

TE

STS

RE

SUL

TS

J.P

apin

, F. L

emoi

ne, E

. Féd

éric

i

IPSN

/DR

S/SE

MA

R

NE

A/C

SNI/

R(2

003)

8/V

OL

2

66

TH

E C

AB

RI

RE

P N

a PR

OG

RA

MM

E

CO

NT

EX

T A

ND

OB

JEC

TIV

E

�

Stud

y of

the

cont

rol r

od e

ject

ion

acci

dent

�

Bur

nup

incr

ease

in th

e Fr

ench

PW

Rs

(→ 5

2 G

Wd/

t mea

n as

sem

bly,

≈60

GW

d/t l

ocal

)�

IP

SN-E

DF

colla

bora

tion,

sup

port

ed b

y N

RC

�

Inve

stig

atio

n ot

the

tran

sien

t rod

res

pons

e w

ith m

oder

ate

clad

hea

t-up

(be

fore

boi

ling

cris

is):

rod

fai

lure

mec

hani

sm, o

nset

of

fuel

dis

pers

alfo

r ad

apta

tion

of s

afet

y cr

iter

ia

�

Eva

luat

ion

of s

afet

y m

argi

ns

�

Stud

y of

the

beha

viou

r of

hig

h bu

rnup

UO

2 fue

l and

MO

X f

uel,

incl

udin

g ad

vanc

ed f

uels

NE

A/C

SNI/

R(2

003)

8/V

OL

2

67

TH

E C

AB

RI

RE

P N

a PR

OG

RA

MM

E

12 C

AB

RI

RE

P N

a te

sts

(8

UO

2, 4

MO

X)

reco

nditi

oned

rod

s : L

=56

cm, 3

b H

eco

olan

t sod

ium

: T

inle

t= 2

80°C

, v=

4m/s

, P=

3bze

ro in

itial

pow

er

SCA

NA

IR c

ode

deve

lopm

ent f

or d

escr

iptio

n of

rod

them

o-m

echa

nica

l beh

avio

ur

Sepa

rate

eff

ect t

ests

:M

echa

nica

l pro

pert

ies

→ P

RO

ME

TR

AC

lad

to w

ater

hea

t tra

nsfe

r un

der

fast

tran

sien

ts →

PAT

RIC

IAT

rans

ient

fis

sion

gas

beh

avio

ur →

SIL

EN

E –

RIA

NE

A/C

SNI/

R(2

003)

8/V

OL

2

68

The

CA

BR

I R

EPN

a T

ests

- U

O2

Fuel

Tes

tR

odPu

lse

(ms)

Ene

rgy

end

ofpe

ak (

cal/

g)C

orro

sion

(µ)

Cla

ddin

gm

ater

ial

Res

ults

and

obs

erva

tion

s

Na-

1(1

1/93

)G

RA

54.

5 %

U64

GW

d/t

9.5

110

(at 0

.4 s

)80

initi

al s

palli

ngZ

r-4

stan

dard

- Fa

ilure

, bri

ttle

type

for

Hf ≈

30

cal/

g-

Hyd

ride

acc

umul

atio

n-

Fuel

dis

pers

ion

6 g,

incl

udin

g fu

el f

ragm

ents

out

side

RIM

(>

40 µ

) P

ress

ure

peak

s in

Na

of 9

-10

bars

Na-

2(6

/94)

BR

36.

85 %

U33

GW

d/t

9.1

211

(at 0

.4 s

)4

Zr-

4N

o fa

ilure

H

max

= 2

10 c

al/g

Max

. str

ain

: 3.5

% a

vera

ge,3

.1%

mid

pel

let

FG

R :

5.5

%N

a-3

(10/

94)

GR

A 5

4.5

%53

GW

d/t

9.5

120

(at 0

.4 s

)40

Zr-

4 lo

w ti

nN

o fa

ilure

H

max

= 1

25 c

al/g

Max

. str

ain

: 2 %

FG

R :

13.7

%

Na-

4(7

/95)

GR

A 5

4.5

% U

62 G

Wd/

t

# 75

95(a

t 1.2

s)

80no

initi

alsp

allin

g

Zr-

4st

anda

rdN

o fa

ilure

H

max

= 9

9 ca

l/g

Cla

ddin

g sp

alli

ng u

nder

tran

sien

tM

ax. s

trai

n : 0

.4 %

FG

R :

8.3

%N

a-5

(5/9

5)G

RA

54.

5 %

U64

GW

d/t

9.5

105

(at 0

.4 s

)20

Zr-

4st

anda

rdN

o fa

ilure

H

max

= 1

15 c

al/g

Max

. str

ain

: 1.1

%

FGR

: 15

.1 %

Na-

8(0

7/97

)G

RA

54.

5 %

60 G

Wd/

t

7510

6(a

t 1.2

s)

130

lim. i

niti

alsp

allin

g

Zr-

4st

anda

rdFa

ilur

e

Hf ≤

82

cal/

g ,

Hm

ax=

110

cal/

gno

fue

l dis

pers

ion

Na-

10(0

7/98

)G

RA

-54.

5%62

GW

d/t

3110

7(a

t 1.2

s)80

initi

al s

palli

ngZ

r-4

stan

dard

Fail

ure

H

f= 7

9 ca

l/g

,

Hm

ax=1

10 c

al/g

no f

uel d

ispe

rsal

Na-

11G

RA

-54.

5%63

,4 G

Wd/

t

3110

4(a

t 1.2

s)15

M5

No

ro

d f

ailu

re

Hm

ax =

110

cal

/gP

ost

exa

min

atio

n u

nd

erw

ay

NE

A/C

SNI/

R(2

003)

8/V

OL

2

69

TH

E R

EP

Na

MO

X T

ES

TS

Tes

t R

od

P

uls

e (m

s)

En

erg

y en

d

of

pea

k (c

al/g

)

Co

rro

sio

n

(�)

Res

ult

s an

d o

bse

rvat

ion

s

Na-

9 (0

4/97

) M

OX

2

cycl

es

28 G

Wd

/t

34

197

at 0

.5 s

24

1 at

1.2

s

< 2

0 N

o f

ailu

re

H

max

= 2

10 c

al/g

M

ax. S

trai

n :

7.4

% a

vera

ge

FG

R :

~ 3

4%

Na-

6 (0

3/96

) M

OX

3

cycl

es

47 G

Wd

/t

35

126

at 0

.66

s

165

at 1

.2 s

35

N

o f

ailu

re

H

max

= 1

48 c

al/g

M

ax. S

trai

n :

3.2

%

(2.

5% a

vera

ge)

F

GR

: 2

1.6

%

Na-

7 (1

/97)

MO

X

4 cy

cles

55

GW

d/t

40

125

at 0

.48

s 17

5 at

1.2

0 s

50

Fai

lure

,

Hf =

120

cal

/g (

t=0.

452

s)

Str

on

g f

low

eje

ctio

n, p

ress

ure

pea

ks o

f 20

0-11

0b, f

uel

m

oti

on

in t

he

low

er h

alf

zon

e

Na

12

(12/

00)

MO

X

5 cy

cles

65

GW

d/t

62.5

10

5 at

1.2

s 80

N

o f

ailu

re,

H

max

= 1

09 c

al/g

P

ost

tes

t ex

amin

atio

ns

un

der

way

Fue

l man

ufac

turin

g p

roce

ss :

MIM

AS

/AU

C ,

Zirc

aloy

4 c

ladd

ing

NE

A/C

SNI/

R(2

003)

8/V

OL

2

70

CL

AD

CO

RR

OSI

ON

IM

PAC

T O

N R

OD

BE

HA

VIO

UR

(1)

larg

e co

rros

ion

thic

knes

s +

oxi

de s

palli

ng →

pres

ence

of

blis

ters

→ d

rast

ic lo

ss o

fduc

tility

of

Zr

clad

ding

(co

nfir

med

by

PRO

ME

TR

A

resu

lts)

cf R

EP

Na

1-8-

10

RE

P N

a1 C

R1

8mm

bfc

RE

P N

a 8

CR

1 31

5mm

bfc

NE

A/C

SNI/

R(2

003)

8/V

OL

2

71

CL

AD

CO

RR

OSI

ON

IM

PAC

T O

N R

OD

BE

HA

VIO

UR

(2

)

• M

oder

ate

or la

rge

corr

osio

n (u

p to

80

µm)

with

out i

nitia

l spa

lling

wit

hsta

nds

sign

ific

ant c

lad

stra

inin

g

• T

rans

ient

spa

lling

link

ed to

cla

d st

rain

ing,

enh

ance

d by

azi

mut

hall

y no

n un

ifor

m in

itial

oxi

de th

ickn

ess

with

ova

lisin

g

Fuel ZrO

2

Cla

d0°

90°

0°

90°

Max

str

ain

Min

str

ain

Spal

ling

zone

45°

270°

225°

→ p

ossi

ble

effe

ct o

n ea

rlie

r bo

iling

cri

sis

occu

rren

ce in

pre

ssur

ised

wat

er c

ondi

tion

s

NE

A/C

SNI/

R(2

003)

8/V

OL

2

72

TH

E F

ISSI

ON

GA

SES

CO

NT

RIB

UT

ION

ON

RO

D B

EH

AV

IOU

R (

1)

• In

case

of

high

ent

halp

y le

vel

(≥,1

50 c

al/g

) co

ntri

butio

n of

int

ra-g

ranu

lar

fiss

ion

gas

swel

ling

on c

lad

load

ing

and

defo

rmat

ion

inad

diti

on to

ther

mal

exp

ansi

on (

RE

P N

a2 ,

RE

P N

a9)

; con

firm

ed b

y hy

dros

tati

c de

nsity

red

ucti

on m

easu

red

afte

r te

st

• Ext

ensi

ve f

uel f

ragm

enta

tion

in th

e R

EP

Na

test

s

20

µm20

µmR

EP

Na5

le

ft :r

im ,

right

0.9

5 R

RE

P N

a6

left

:0.9

8R, r

ight

0.9

R

• FG

R in

crea

sing

with

bur

n-up

, cor

rela

ted

to c

lad

defo

rmat

ion

(RE

P N

a4-5

)

→ R

ole

of g

rain

bou

ndar

y ga

ses

NE

A/C

SNI/

R(2

003)

8/V

OL

2

73

TH

E F

ISSI

ON

GA

SES

CO

NT

RIB

UT

ION

ON

RO

D B

EH

AV

IOU

R (

2)

Pres

ent u

nder

stan

ding

:

The

gra

in b

ound

ary

gas

cont

ent i

s in

crea

sing

ver

sus

burn

up (

rim

zon

e, U

PuO

2 par

ticle

s in

MO

X f

uel)

Rap

id f

uel h

eat-

up c

ause

s ga

s bu

bble

pre

ssur

isat

ion

(int

er-g

ranu

lar,

por

osity

)

→ g

rain

bou

ndar

y cr

acki

ng +

grai

n se

para

tion

depe

ndin

g on

gas

pre

ssur

e an

d fu

el c

onst

rain

t

→ p

rom

pt g

as a

vaila

bilit

y fr

om h

igh

burn

up r

egio

nco

ntri

butio

n to

cla

d lo

adin

g +f

ailu

re r

isk

→ in

crea

sed

effe

ct w

ith M

OX

fue

l at h

igh

burn

up a

nd n

on h

omog

eneo

us m

icro

stru

ctur

em

ay e

xpla

in R

EP

Na

7 fa

ilure

with

gas

and

fue

l eje

ctio

n

NE

A/C

SNI/

R(2

003)

8/V

OL

2

74

T

HE

FIS

SIO

N G

ASE

S C

ON

TR

IBU

TIO

N O

N R

OD

BE

HA

VIO

UR

(3)

Gra

in b

ound

ary

sepa

rati

on c

reat

es p

athe

s fo

r FG

R :

• C

orre

late

d to

cla

d de

form

atio

n

• C

onfi

rmed

by

GB

G in

vent

ory

afte

r te

st (

AD

AG

IO o

xida

tion

met

hod

on R

EP

Na

4-5

sam

ples

)

GB

G r

elea

se in

inne

r an

d ou

ter

fuel

zon

es in

RE

P N

a5G

BG

rel

ease

in o

uter

zon

e on

ly in

RE

P N

a4

→ c

ontr

ibut

ion

of th

e ri

m z

one

to F

GR

und

er s

low

pow

er p

ulse

s(R

EP

Na

4)

NE

A/C

SNI/

R(2

003)

8/V

OL

2

75

TH

E S

CA

NA

IR C

OD

E

Des

crip

tion

of th

e th

erm

o-m

echa

nica

l beh

avio

ur o

f PW

R h

igh

burn

up f

uel r

ods

(UO

2,M

OX

) un

der

RIA

• 1.

5 D

mod

elin

g , i

ntim

atel

y co

uple

d ph

enom

ena

: the

rmal

dyn

amic

s, m

echa

nics

, fis

sion

gas

es, t

herm

al-h

ydra

ulic

s•

deve

lope

d in

clo

se li

nk w

ith in

terp

reta

tion

of g

loba

l CA

BR

I te

sts

+ s

uppo

rt te

sts

(PR

OM

ET

RA

, PA

TR

ICIA

)

Dev

elop

men

ts u

nder

way

:-

appr

oach

for

cla

ddin

g fa

ilur

e cr

iteri

on u

nder

PC

MI

or g

as p

ress

ure

load

ing

: du

ctile

(cu

mul

ated

str

ain,

CSE

D)

and

britt

le (

frac

tura

l m

echa

nics

) ca

ses

- m

echa

nica

l pro

pert

ies

of a

dvan

ced

allo

ys (

Zir

lo,M

5)-

ther

mal

-hyd

raul

ics

base

d on

Pat

rici

a →

boili

ng c

risi

s w

ithou

t nuc

leat

e bo

iling

reg

ime

- cl

ad d

efor

mat

ion

unde

r pr

essu

re (

cree

p m

echa

nism

,hig

h T

emp.

)

NE

A/C

SNI/

R(2

003)

8/V

OL

2

76

TH

E S

CA

NA

IR M

OD

EL

ING

Neu

tron

ic c

alcu

latio

nPo

wer

tran

sien

t

Com

puta

tion

of

the

fuel

be

havi

or

duri

ngir

radi

atio

n

Fuel

rod

initi

al s

tate

bef

ore

the

tran

sien

t

(“t o s

tate

”)

ther

mic

s +

ther

mal

-hyd

rau

lics

(Na

or

wat

er)

NE

A/C

SNI/

R(2

003)

8/V

OL

2

77

TH

E S

CA

NA

IR C

OD

E Q

UA

LIF

ICA

TIO

N

fis

sion

gas

rel

ease

aft

er th

e tr

ansi

ent.

010

20

30

40

exp.results(%)

0

10

20

30

40

SCANAIRcalculations(%)

Na3

Na4

Na5

Na9

Na6

Leg

end

UO

2 tes

ts

MO

X te

sts

FG r

elea

se o

f th

e un

faile

d C

AB

RI

RE

P N

a U

O2 a

nd M

OX

test

s.

NE

A/C

SNI/

R(2

003)

8/V

OL

2

78

TH

E S

CA

NA

IR C

OD

E Q

UA

LIF

ICA

TIO

N

Cla

ddin

g pl

astic

hoop

str

ain

of th

eun

faile

d C

AB

RI

RE

P N

a U

O2 a

ndM

OX

test

s

01

23

45

67

8exp.results(%)

012345678

SCANAIRcalculations(%)

Na3

Na4

Na5

Na9

Na6

Na7(*)

Leg

end

UO

2 tes

ts

MO

X te

sts

(*)

Axi

ally

av

erag

ed

tota

l ho

opst

rain

at t

ime

of f

ailu

re(e

xp.

resu

lts

+

SCA

NA

IRre

sults

).

NE

A/C

SNI/

R(2

003)

8/V

OL

2

79

NE

ED

S FO

R F

UR

TH

ER

IN

VE

STIG

AT

ION

S

1/ A

nsw

er to

pen

ding

que

stio

ns

• T

rans

ient

fis

sion

gas

beh

avio

ur w

ith i

mpa

ct o

n cl

ad l

oadi

ng (

shor

t an

d lo

ng t

erm

s) f

or h

igh

burn

up U

O2

fuel

and

for

MO

X f

uel

(rim

zon

e,U

PuO

2 ag

glom

erat

es),

• Fu

el b

ehav

iour

aft

er b

oilin

g cr

isis

• In

tern

al p

ress

ure

effe

ct•

Pos

t-fa

ilur

e ph

enom

ena

: fu

el e

ject

ion,

fue

l coo

lant

inte

ract

ion

with

fin

ely

frag

men

ted

soli

d fu

el

2/ P

rovi

de u

nder

typi

cal P

WR

con

diti

ons,

bas

es f

or a

sses

smen

t of

new

RIA

rel

ated

saf

ety

crite

ria

for

adva

nced

hig

h bu

rnup

UO

2 an

d M

OX

fue

ls

3/ E

valu

ate

safe

ty m

argi

ns

NE

A/C

SNI/

R(2

003)

8/V

OL

2

80

TH

E C

AB

RI

INT

ER

NA

TIO

NA

L P

RO

GR

AM

ME

( C

IP )

TE

ST M

AT

RIX

( 1

2 te

sts

in to

tal )

• C

IP0

: 2 r

efer

ence

test

s in

the

sodi

um lo

op, a

dvan

ced

fuel

s (Z

irlo

, M5

clad

ding

)

• C

IPQ

: qu

alif

icat

ion

test

of

the

wat

er lo

op

• C

IP1

: 2 r

efer

ence

test

s in

the

wat

er lo

op, f

or c

ompa

riso

n to

CIP

0 te

sts

• C

IP3

: ultr

a hi

gh b

urnu

p (8

0-10

0 G

Wd/

t, on

e D

uple

x ro

d)

• C

IP4

: im

prov

emen

t of

phys

ical

und

erst

andi

ng o

f R

IA

• C

IP4

: MO

X f

uel,

high

bur

nup

• C

IP5

: com

plem

enta

ry te

sts

( op

en )

NE

A/C

SNI/

R(2

003)

8/V

OL

2

81

CO

NC

LU

SIO

N

The

CA

BR

I R

EPN

a pr

ogra

mm

e an

d as

soci

ated

stu

dies

hav

e gi

ven

maj

or r

esul

ts f

or th

e kn

owle

dge

of h

igh

burn

up f

uel b

ehav

iour

und

er R

IA :

• no

n ad

equa

cy o

f pr

esen

t saf

ety

crite

ria

• ba

sis

for

safe

ty s

tudi

es r

elat

ed to

bur

nup

incr

ease

aut

hori

sati

on u

p to

52

GW

d/t (

mea

n as

sem

bly)

The

key

par

amet

ers

have

bee

n id

entif

ied

for

the

firs

t pha

se o

f th

e tr

ansi

ent

- de

lete

riou

s ef

fect

of

high

cor

rosi

on le

vel w

ith

hydr

ides

con

cent

ratio

n an

d sp

allin

g on

cla

d fa

ilure

- in

flue

nce

of g

rain

bou

ndar

y ga

ses

on c

lad

load

ing

and

FGR

The

CA

BR

I R

EPN

a pr

ogra

mm

e ha

s pr

ovid

ed in

form

atio

n an

d ba

sis

for

SCA

NA

IR c

ode

deve

lopm

ent a

nd v

alid

atio

n

The

CA

BR

I In

tern

atio

nal

Prog

ram

me

now

in

its a

ctiv

e ph

ase,

will

pro

vide

und

er t

ypic

al P

WR

con

ditio

ns,

base

s fo

r as

sess

men

t of

new

saf

ety

crite

ria

for

RIA

and

allo

w e

valu

atio

n of

the

safe

ty


82


83

NSRR RIA Tests Results and Experimental Programmes

T. Nakamura, H. Sasajima and H. Uetsuka

Japan Atomic Energy Research Institute (JAERI), Tokai-mura, Ibaraki-ken, 319-1195 JapanTel: +81(29)282-6386, Fax: +81(29)282-5429, e-mail:[email protected]

ABSTRACT

Pulse irradiation tests of PWR, BWR and ATR/MOX fuels have been conducted in the Nuclear SafetyResearch Reactor (NSRR) under reactivity initiated accident (RIA) conditions. Short test fuel rods werefabricated from the commercial reactor fuel rods at burnups from 20 to 61GWd/t. Thermal energy from196 to 607 J/g (47 to 145 cal/g) was promptly subjected within about 20ms to the test rods, which werecontained in capsules filled with water at room temperature and at ambient pressure. Transient behavior ofthe rods was monitored through on-line measurements of rod internal pressure, pellet stack/cladding axialdeformation, cladding hoop deformation, cladding surface temperature and capsule water temperature.Change of the fuel micro structure, fuel plastic deformation and fission gas release were evaluated from preand post-test examinations of the rods.

In the recent BWR fuel tests at a burnup of 61GWd/t, brittle cladding fracture occurred early during thetransient irradiation, while the cladding remained cool. The fracture surface exhibited slightly differentnature of the cracks from those observed in the earlier PWR fuel tests at a burnup of about 50GWd/t.Hydride distribution in the BWR cladding was different from those observed in the PWR fuels, whichlikely contributed to the BWR fuel failure at lower hydrogen contents of about 150-200ppm than that ofPWR cladding of about 400ppm. Failure enthalpies were in a range of 250 to 359 J/g (60 to 86 cal/g) andcomparable in the BWR and PWR fuel tests. Transient hoop strain measurements of the cladding in theearly phase of the transients indicated small deformation below 0.4%, suggesting that the deformation wascaused mainly by thermal expansion of the pellets.

At the time of the fuel failure, pressure pulse generation and movement of capsule water were observed.Thermal interaction between fine fuel fragments of about 0.05mm and the capsule water is believed toconvert the thermal energy of the pellets to the mechanical energy in forms of pressure pulses and thewater hammer. Transient fission gas release by the pulse irradiation was in a range from 4 to 23%depending mainly on the fuel burnups and the peak fuel enthalpies, which could be additional radioactivesources in case of the fuel failure. In the MOX fuel tests, relatively large fission gas release was observed,which could be resulted from the locally higher burnups and fuel enthalpies at Pu rich spots in the MOXfuel.


84


85


86


87


88


89


90


91


92


93


94


95


96


97

High Burnup Fuel and Cladding Characteristics as RIA Test Initial Condition

K. Kamimura, NUPEC, Japan

ABSTRACT

The Nuclear Power Engineering Corporation (NUPEC) carried out the “NUPEC-HB Project”, whichconsisted of the BWR and PWR verification tests on high burnup fuel, under the sponsorship of theMinistry of Economy, Trade and Industry (METI).

The fuel rods out of 8x8 type BWR fuel assemblies, which were irradiated up to 5 cycles (burnup assemblyav.48GWd/t, pellet peak 61GWd/t) in Fukushima Daini-2 NPP, were supplied for PIE and for power ramptests in the Japanese Material Test Reactor (JMTR).

The fuel rods out of 17x17 type PWR fuel assemblies, which were irradiated up to 4 cycles (burnupassembly av.53GWd/t, pellet peak 61GWd/t) in Vandellos-2 NPP, were supplied for PIE and for powerramp tests in R-2 Reactor in Studsvik.

The data driven from those PIE and ramp tests were evaluated, and the observations of high burnup fuelperformance during steady state irradiation and power ramp were obtained. The reviewed characteristics ofhigh burnup fuel and cladding to be considered for RIA test initial condition are as follows.

BWR FUEL

The corrosion was not nodular but uniform, and oxide thickness was less than 20 micrometer even after a5-cycle irradiation.

The hydrogen content in the cladding tube was less than 200 ppm after a 5-cycle irradiation, although thehydrogen pickup ratio remarkably increased during the fourth and fifth irradiation cycles.

Radial hydrides were observed at the outer rim of the cladding tubes irradiated for 4 and 5 cycles andramp-tested rods.

The result of a cladding tube burst test at 300 degrees C and RT showed axial sprit and small strain.

The fission gas release ratio was at the low level at less than 6 %.

Pellet-cladding bonding layers were observed on the almost whole inside cladding surface after 5 cycles.

The ramp test results showed that the failure threshold power for higher burnup rods decreased, and it wascaused by PCMI connected with hydrogen precipitation behaviour.


98

PWR FUEL

The oxide thickness of MDA and ZIRLOTM was smaller than that of low Sn Zry-4. The maximum oxidethickness of low Sn Zry-4 was 100 micrometer after a 4-cycle irradiation.

The hydrogen content in the cladding tube was as large as 300~900 ppm after a 4-cycle irradiation,although the hydrogen pickup ratio was almost the same as 15% during the third and fourth irradiationcycle for any cladding materials.

Dense circumferential hydrides were observed at the outer rim of the cladding tubes.The fission gas release ratio was at the low level at less than 3 %.

The ramp test results showed that the failure threshold power had not decreased even in the range of higherburnup as 4-cycle.


99


100


101


102


103


104


105


106


10710


108


109


110


111


112


113


114


115

Study of High Burnup VVER Fuel Rods Behaviour at the BIGRReactor Under RIA Conditions: Experimental Results.

Yu.Bibilashvili, A.Goryachev,O.Nechaeva, A.Salatov, V.Sazhnov, I.Smirnov, N.Sokolov,

Yu.Trutnev, V.Ustinenko, L.Yegorova.

A.A.Bochvar All-Russian Research Institute of Inorganic MaterialsNuclear Safety Institute of RRC “Kurchatov Institute”

Russian Federation Nuclear Center “All-Russian Research Institute of Experimental Physics”State Research Center “Research Institute of Atomic Reactors”

ABSTRACT

The experimental results received during tests of VVER-type high burnup refabricated fuel rods at theBIGR (the name of a research reactor) reactor are submitted. The pulse reactor BIGR providesreproduction of fast energy deposition increase in experimental rod fuel, simulating conditions for RIAaccident. The scheme of the lateral experimental channel arrangement near to BIGR active zone, scheme ofthe most experimental channel device, including tested fuel rod, placed in an ampoule filled water undernormal conditions, and also refabricated fuel rod scheme are given. The spatial - temporary distributions ofenergy deposition and enthalpy in refabricated rods fuel are briefly stated. The test results of twelverefabricated fuel rods with burnup 48 and 60 •W d/kg U and initial pressure 0,1 and 2,0 •P• under pulsepower conditions (peak enthalpy - 115-190 cal/g, pulse width - 3 •s) are described. Fuel rods fragmentationwas not observed. Four fuel rods had cladding failure. For all fuel rods the type of cladding failure issimilar – formation of local non-axis-symmetrical ballooning with rupture in the maximal deformationfield. Fuel rods with burnup 48 MW d/kg U had on one rupture site. Fuel rods with burnup 60 MW d/kg Uhad two and four rupture sites. The comparison of experimental data of BIGR-tests with the data of IGR-tests, allows to generalize results of researches in a common database used to support safety substantiationat forecasting of VVER high burnup fuel rod behaviour in design RIA type accidents.


116

Study of High Burnup VVER Fuel Rods Behaviour at the BIGRReactor Under RIA Conditions: Experimental Results.

Yu.Bibilashvili, A.Goryachev,O.Nechaeva, A.Salatov, V.Sazhnov, I.Smirnov, N.Sokolov,

Yu.Trutnev, V.Ustinenko, L.Yegorova.

A.A.Bochvar All-Russian Research Institute of Inorganic MaterialsNuclear Safety Institute of RRC “Kurchatov Institute”

Russian Federation Nuclear Center “All-Russian Research Institute of Experimental Physics”State Research Center “Research Institute of Atomic Reactors”

The fuel criteria phenomena requiring an experimental support for the RIA safety analysis /1-3/

• Fuel rod fragmentation – criteria is intended to ensure coolability of the core and preclude theenergetic dispersal of fuel particles into the coolant

(for VVER fuel - 230 cal/g fuel on peak radial averages fuel rod enthalpy )

• Cladding failure - criteria is used to indicate cladding failure and to estimate amount of non-hermetic fuel rods in calculating radiological releases

(for VVER fuel is developed)

NE

A/C

SNI/

R(2

003)

8/V

OL

2

117

Exp

erim

enta

l stu

dy o

f V

VE

R f

uel b

ehav

iour

und

er R

IA c

ondi

tion

s

Rea

ctor

s IG

R, G

IDR

A /3

/R

eact

or B

IGR

/4/

1983

- 1

990-

uni

rrad

iate

d fu

el r

ods

1994

-199

7 -

unir

radi

ated

fuel

rod

sD

efin

ition

of

failu

re th

resh

olds

for

unir

radi

ated

fue

l rod

s w

ith E

110

cla

ddin

gR

eact

or B

IGR

upd

atin

g an

d m

etho

dica

l res

earc

hes.

Com

para

tive

test

s of

uni

rrad

iate

d fu

el r

ods

with

E11

0 an

d E

635

clad

ding

Rea

ctor

IG

R /

3,6/

Rea

ctor

BIG

R /4

,5/

- 19

90-1

992

– re

fabr

icat

ed h

igh

burn

up f

uel r

ods

- 19

97-2

000

– re

fabr

icat

ed h

igh

burn

up f

uel

rods

Tes

t c

ondi

tion

s

Sing

le f

uel r

od is

pla

ced

in a

cap

sule

wit

h th

e w

ater

und

er n

orm

al c

ondi

tions

(20

• •, 0

,1 •

P•)

- bu

rnup

48-

51•W

d/•g

U-

burn

up 4

7-60

•W

d/•

g

- pu

lse

wid

th 6

00-9

00 m

s-

puls

e w

idth

4-5

ms

- in

itia

l pre

ssur

e in

a f

uel r

od 1

,7 •

P•

(He)

- in

itia

l pre

ssur

e in

a fu

el r

od 0

,1;

2,0

•P•

(He)

- m

axim

al p

eak

fuel

ent

halp

y 25

0 c•

l/g-

max

imal

pea

k fu

el e

ntha

lpy

190

c•l/g

Pos

ttes

t ex

amin

atio

n

The

cal

cula

tion

ana

lysi

s, in

terp

reta

tion

of

resu

lts,

dev

elop

men

t of

a d

atab

ase

1995

-199

920

00-2

001

NE

A/C

SNI/

R(2

003)

8/V

OL

2

118

VV

ER

fue

l rod

test

s in

the

late

ral e

xper

imen

tal c

hann

el (

LE

C)

of r

eact

or B

IGR

und

er R

IA c

ondi

tion

s

The

acc

omm

odat

ion

sche

me

of t

he L

EC

The

sch

eme

of t

he L

EC

dev

ice


119

THE DETERMINATION OF PULSE POWER IN EXPERIMENTAL FUEL AS TIME ANDCOORDINATES FUNCTION IS BASED ON:

• methodology used at realization of IGR pulse experiments /6 /,• the experimental data are given from BIGR tests with special samples and unirradiated VVER type

fuel rods,• results of neutronic calculations used two independent codes WIMS-6 and •-95 (Monte-Carlo type)

/7,8 /,• results of radiochemical measurements.

0 20 40 60Time (ms)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

Rea

ctor

pow

er (

per-

unit

)

BIGR power versus time

• The axial and azimuth non-uniformity of fissions in the VVER fuel - no more than 5 %.

• The fission distribution on fuel radius was determined for everyone experimental fuel rod on the basisof neutron-physical accounts in view of the NPP irradiation history and radiochemical measurementresults.

NE

A/C

SNI/

R(2

003)

8/V

OL

2

120

The

def

init

ion

spat

ial

- te

mpo

rary

fue

l en

thal

py d

istr

ibut

ions

and

oth

ers

ther

mal

mec

hani

cal

para

met

ers

of t

he t

este

d V

VE

R h

igh

burn

up f

uel i

n B

IGR

-tes

ts w

as c

arri

ed o

ut o

n tw

o in

depe

nden

tco

des:

AP

TA

-5 i

s de

velo

ped

in R

ussi

a w

ith t

he p

urpo

se t

o an

alys

e V

VE

R f

uel

rod

ehav

iour

in a

ccid

ents

with

loss

of c

oola

nt a

nd w

ith re

activ

ity in

crea

se /5

, 9-1

1 /.

odif

ied

vers

ion

of F

RA

P-T

6 /1

2 /,

deve

lope

dor

con

ditio

ns o

f IG

R/R

IA t

ests

with

VV

ER

igh

burn

up fu

el ro

ds /1

3/.

00

.04

0.0

80

.12

0.1

60

.2T

ime

(s)

0E

+0

00

1E

+0

04

2E

+0

04

3E

+0

04

4E

+0

04

5E

+0

04

6E

+0

04

Linear power of fuel ros (kW/m)

040

80

12

0

16

0

20

0

24

0

Fuel enthalpy (cal/g)

FR

AP

-T6

RA

PT

A-5

Po

wer

Fu

el e

nth

alp

y

00.

040.

080.

120.

160.

Tim

e (s

)

0

500

1000

1500

2000

2500

3000

3500

Temperature (K)

Fuel

tem

pera

ture

(FR

AP-

T6)

Fuel

tem

pera

ture

(RA

PTA

-5)

Cla

ddin

g te

mpe

ratu

re (F

RA

P-T6

)

Cla

ddin

g te

mpe

ratu

re (R

APT

A-5

)

Red

ial a

vera

ges

enth

alpy

vs.

tim

e ca

lcul

ated

with

FR

AP-

T6

and

RA

PTA

-5co

des

(RT

6).

Fuel

and

cla

ddin

g te

mpe

ratu

re v

s. ti

me

calc

ulat

ed w

ith F

RA

P-T

6 an

d R

APT

A-

5 co

des

(RT

6).

NE

A/C

SNI/

R(2

003)

8/V

OL

2

121

Hig

h bu

rnup

fue

l rod

test

s at

the

BIG

R r

eact

or

Sche

me

of h

igh

burn

up f

uel r

od r

efab

rica

ted

from

a c

omm

erci

al f

uel e

lem

ent

of t

he V

VE

R t

ype

• E

ight

ref

abri

cate

d fu

el r

ods

wit

h bu

rnup

48

•Wt

d/kg

U•

Fou

r r

efab

rica

ted

fuel

rod

s w

ith

burn

up 6

0 •W

t d/

kg U

• P

ulse

wid

th 4

-5 m

s•

Init

ial g

as p

ress

ure

(He)

in fu

el r

ods

0,1;

2,0

•P

••

Pea

k fu

el e

ntha

lpy

- 1

15-1

90 c

al/g

UO

2.


122

Parameters of VVER refabricated fuel rods tested at the BIGR reactor

Peak fuel enthalpy,cal/g

Burnup,•Wt d/kg U

Initial pressure (He) infuel rod,

•p•

Cladding hoopstrain,%(average)

Test Result forcladding

� – No failure� – Failure

� � � � � � � • • • • �

Refabricated fuel rodnumber (RT) 1 2 3 4 5 6 7 8 9 10 11 12

0

100

200

00

0

0

0

0

0

0

2

0

5

10

NE

A/C

SNI/

R(2

003)

8/V

OL

2

123

Tes

t re

sult

s of

VV

ER

ref

abri

cate

d fu

el r

ods

at t

he B

IGR

rea

ctor

• Fu

el r

ods

fra

gmen

tatio

n w

as n

ot o

bser

ved;

• Fo

ur f

uel r

ods

had

clad

ding

fai

lure

;

• Fo

r al

l fu

el r

ods

the

type

of

clad

ding

fai

lure

site

s is

sim

ilar

– fo

rmat

ion

of l

ocal

non

-axi

s-sy

mm

etri

cal

ball

ooni

ng w

ith

rupt

ure

in t

he m

axim

al d

efor

mat

ion

fiel

d;•

Fuel

rod

s w

ith b

urnu

p 48

MW

t d/k

g U

had

on

one

rupt

ure

site

;•

Fuel

rod

s w

ith

burn

up 6

0 M

Wt d

/kg

U h

ad tw

o an

d fo

ur r

uptu

re s

ites

�� ru

ptur

e si

te, t

he f

ourt

h ru

ptur

e on

RT

9 is

loca

ted

on th

e re

turn

par

ty

App

eara

nce

of r

efab

rica

ted

fuel

rod

s w

ith

burn

up 6

0 M

Wt d

/kg

U a

fter

BIG

R te

sts

NE

A/C

SNI/

R(2

003)

8/V

OL

2

124

Tes

t re

sult

s of

VV

ER

ref

abri

cate

d fu

el r

ods

at t

he B

IGR

rea

ctor

(co

ntin

ue)

Abo

ut in

crea

se o

f bu

rnup

to 6

0- M

Wt

d/kg

U:

• T

he ty

pe c

ladd

ing

failu

re s

ites

- no

n-ax

is-s

ymm

etri

cal b

allo

onin

g w

ith r

uptu

re -

is k

ept;

• Fu

el r

od c

ladd

ing

keep

s a

suff

icie

nt s

tock

of

plas

tic p

rope

rtie

s,

Uni

form

elo

ngat

ion

of th

e V

VE

R a

nd P

WR

irra

diat

ed c

ladd

ing

vs. t

empe

ratu

re /1

4/

NE

A/C

SNI/

R(2

003)

8/V

OL

2

125

Tes

t re

sult

s of

VV

ER

ref

abri

cate

d fu

el r

ods

at t

he B

IGR

rea

ctor

(co

ntin

ue)

• T

he a

bsen

ce o

r si

gnif

ican

t re

duct

ion

of a

n in

itial

gap

bet

wee

n fu

el a

nd c

ladd

ing,

and

als

o ga

s re

leas

e fr

om f

uel

rend

ers

the

esse

ntia

l in

flue

nce

on c

ladd

ing

defo

rmat

ion.

• In

all

cas

es o

f cl

addi

ng f

ailu

re th

e ho

op s

trai

n m

ade

not l

ess

than

5 %

, thu

s th

e re

lati

ve n

arro

win

g of

a c

ladd

ing

cros

s se

ctio

n re

ache

s 10

0%.

Def

orm

ing

of r

efab

rica

ted

fuel

rod

cla

ddin

g w

ith

burn

up 6

0 •W

t d/

kg U

in r

uptu

re c

ross

sec

tion


126

Test results of VVER refabricated fuel rods at the IGR reactor /15/

Peak fuelenthalpy,

cal/g

Burnup,��

Test Result forcladding

� – No failure� – Failure

� • • � • � • �

Refabricated

fuel rod

number

� � � � � � � �

Pulse width 600-900 msInitial gas pressure (He) in fuel rods 1,7 •P•

• the type for all cladding failure sites is similar – formation of local non-axis-symmetrical

ballooning with rupture in the maximal deformation field;

• •2•, •3• and •7• had till two cladding failure sites, the cladding residual hoop strain in rupture

cross sections made 12-25 %,

• •5• had one cladding failure site, the cladding residual hoop strain in rupture cross sectionsmade 6,5 %..

Refabricated fuel rod (H5T) cladding in rupture cross section

0

100

200

00

155 223 246 114 172 87 184 61

0

40

80

49.2 47.9 49.3 48.7 49.0 49.3 47.3 46.8

127

127

Tes

t re

sult

s of

VV

ER

ref

abri

cate

d fu

el r

ods

at t

he I

GR

and

BIG

R r

eact

ors

1.1.

1.1.

1.1.

2

1.1.

1.1.

1.1.

3

��n

o fa

ilur

e,

� fa

ilure

4050

6070

��

��

0

100

200

300

Peak fuel enthalpy, cal/g

128

128

Conclusion

• The behaviour of 12 VVER type refabricated fuel rods was researched at BIGR reactor in power pulse

conditions simulating reactivity initiated accident

Burnup 48 and 60 •Wt d/kg U,

Pulse width 4-5 ms,

Peak fuel enthalpy 115-190 cal/g

• Four refabricated fuel rods had cladding failure; type of failure was plastic deformation with rupture in

a place of local gas ballooning; hoop cladding strain not less than 5%.

• Fuel rod fragmentation was not observed.

• The comparison of experimental data of BIGR-tests with the data of IGR-tests, allows to generalize

results of researches in a uniform database used to support safety substantiation at forecasting of

VVER high burnup fuel rod behaviour in design RIA type accidents.

129

129

References

1. Guidelines for Accident Analysis for WWER Nuclear Power Plants. Draft Report., , IAEAConsultants Meetings, Vienna, Austria, 12-14.01.94, 29.08-02.09.94 and 24.03-2 04.95, under preparation.

2. Current Fuel Safety Criteria for PWR and WWER reactors. Appraising and comparing. Draft ReportIAEA, Vienna, Austria , 10.2000-11.2001, to be published

3. Nuclear safety, vol 37,no 4, october-december 19964. M.Kuvshinov, V.Kolesov, A.Voinov, I.Smirnov “A periodic Self-Quenching BIGR Reactor”,

Proceedings of the Topical Meeting ”Physics, Safety, and Applications of Pulse Reactors”,Washington DC, November 13–17, 1994.

5 Yu.K.Bibilashvili, O.A.Nechaeva, L.A.Yegorova, I.G. Smirnov, V.P.Smirnov e.a. Experimental Studyof VVER High Burnup Fuel Rods at the BIGR Reactor under Narrow Pulse Conditions. 2000International Topical Meeting on Light Water Reactor Fuel Performance, Park City, Utah, April 10-13, 2000, pp.306-314.

6. L.Yegorova, V.Asmolov, G.Abyshov, V.Malofeev, A.Avvakumov, E.Kaplar, K.Lioutov,A.Shestopalov, A.Bortash, L.Maiorov, K.Mikitiouk, V.Polvanov, V.Smirnov, A.Goryachev,V.Prokhorov, and A.Vurim “Data Base on the Behavior of High Burnup Fuel Rods with Zr-1%NbCladding and UO2 Fuel (VVER Type) under Reactivity Accident Conditions”, NUREG/IA-0156(IPSN99/08-02, NSI/RRC KI 2179), Vol.2, 1999.

7. WIMS-E User Guide, AEA Technology, Winfrith (ANSWERS/WIMS-E(95)).8. E.Donskoy, V.Yeltsov, A.Zhitnik et al “The Monte Carlo Method in VNIIEF”, VANT, issue 2

“Mathematical Simulation of Physical Processes”, 1999 (Russian).9. Yu.K.Bibilashvili, N.B.Sokolov, O.A.Nechaeva, A.V.Salatov Yu.A.Trutnev, I.G.Smirnov,

V.A.Ustinenko, V.V.Sashnov V.P.Smirnov, A.V.Goryachev V.G.Asmolov, L.A.Egorova .Experimental Researches and alculation Modelling of WWER High Burnup Fuel Rod Behaviourduring Pulse Tests on the Big Pulse Graphite Reactor (BIGR). Third International Seminar onWWER Fuel Performance, Modelling and Experimental Support, Pamporovo, Bulgaria, 4-8October 1999, pp.175-180.

10. Yu.Bibilashvili, N.Sokolov, A.Salatov, O.Nechaeva, L.Andreeva-Andrievskaya, F.Vlasov “Modelingof VVER Fuel Rod Behavior in Accident Conditions Using RAPTA-5 Code”. Second InternationalSeminar on VVER Fuel Performance, Modeling and Experimental Support, Sandanski, Bulgaria, 21-25 April 1997.

11. Yu.K.Bibilashvili, N.B.Sokolov, A.V.Salatov, L.N.Andreyeva-andriyevskaya, O.A.Nechayeva,F.YU.Vlasov. “RAPTA-5 Code: Modelling of Behaviour of Fuel Elements of VVER Type inDesign Accidents. Verification Calculations”. IAEA Technical Committee on Behaviour ofLWR Core Materials under Accident Conditions, held in Dimitrovgrad, Russia, on 9-13October 1995. IAEA-TECDOC-921, Vienna, 1996, pp. 139-152.

12. L.Siefken, Ch.Allison, M.Bohn, S.Peck “FRAP-T6: Computer Code for the Transient Analysis ofOxide Fuel Rods”, NUREG/CR-2148 EGG-2104, May 1981

13. A.Shestopalov, K.Lioutov, L.Yegorova, G.Abyshov, K.Mikitiouk “Modification of U.S. NRC’sFRAP-T6 Fuel Rod Transient Code for High Burnup Fuel” Report of RRC “KI”, NUREG/IA-0164(IPSM99/10, NSI/RRC KI 2180), 1999.

14. L.Yegorova, F.Schmitz, J.Papin “Mechanical Behavior of Fuel Element during RIA Transients”,Proceedings of the EUROSAFE-1999 Meeting, November 18-19, Paris, 1999.

15. V.Asmolov, L.Yegorova “The Russian RIA Research Program: Motivation, Definition, Execution,and Results”, Nuclear Safety, Vol.37, No.4, 1996.

130

130

131

131

RIA Topical Meeting (OECD/CSNI ), May 13-14, 2002, Cadarache, France

Impact of Corrosion on

Rapid Deformation Capabilities of ZIRLOTM Cladding

V. Grigoriev, R. Jakobsson and D. SchrireStudsvik Nuclear AB, SwedenR. Kesterson and D. Mitchell

Westinghouse, USAH. Pettersson

Vattenfall Fuel AB, Sweden

Abstract

Simulated RIA mechanical testing of the cladding specimens was performed by means of the"Expansion Due to Compression" (EDC) polymer mandrel technique recently developed atStudsvik. The test specimens were sections of ZIRLO cladding from a fuel rod operated in theRinghals-2 PWR reactor up to a burnup of about 50 MWd/kgU.

In total, four 20 mm long specimens have been tested: two of them were taken from the peakoxide region of the rod (~3 m from the bottom of the rod) and two specimens were taken from aposition ~1 m from the bottom of the rod. The wall-average hydrogen concentration in thecladding was measured by hot vacuum extraction. The measurements showed hydrogenconcentrations of 125 wtppm at an elevation of 1 m and about 550 wtppm at an elevation of 3 m.

The EDC testing of the specimens is performed for time intervals of 40-100 ms at roomtemperature (RT) or at 340°C. The only specimen fractured, 3 m/RT/100 ms, experienced 3,4%hoop strain before failure. Relatively high hoop strains (up to ~10%) were measured at 340°C onirradiated ZIRLO specimens with hydrogen concentrations on the order of 500 wtppm. Intensiveoxide scaling is observed at the outer surface of all specimens tested. The pattern of the scaledoxide reveals numerous localised shear bands, which appear to be more intensive at higher testtemperature and at higher hoop strains.

The results for ZIRLO cladding have been compared to the data from the EDC testing ofZircaloy-4 cladding from fuel rods of similar burnup. The ZIRLO specimens had a somewhathigher strain to failure than the Zircaloy-4 specimens at similar hydrogen levels and testingconditions.

The maximum SED (in unfailed specimens) and CSED (in failed specimens) were determinedfrom the EDC tests and compared to published values calculated for CABRI tests and other typesof mechanical properties tests.

132

132

133

133

134

134

135

135

136

136

137

137

138

138

139

139

140

140

141

141

LIST OF PARTICIPANTS

BELGIUMDr. Nadine HOLLASKY Tel.: +32 0 2 5280244AVN Association Vinçotte Nuclear Fax: 32 0 2 5280102Rue Walcourt, 148 Eml.: [email protected] Brussels

Dr. Marc VERWERFT Tel.: +32 14 33 30 48SCK/CEN Fax: +32 14 32 12 16Reactor Materials Research Eml.:[email protected] BoeretangB-2400 MOL

Mr. Albert CHARLIER Tel: +32 2 773 98 67Tractebel Energy Engineering Fax: +32 2 773 89 00Avenue Ariane 7 Eml: [email protected] BRUXELLES

CZECH REPUBLIC

Dr. Mojmir VALACH Tel.: +420 2 6617 2100Nuclear Research Institute Fax: +420 2 6617 2100Dept. of Reactor Technology Eml.: [email protected] - REZ U PRAHY

Mr. Ladislav BELOVSKY Tel. (+420) 2 272 921 678 ALIAS CZ s.r.o. Fax (+420) 2 57 21 28 66 Bachova 1586/16 E-mail: [email protected] 149 00 Praha-415

Dr. Jiri MACEK Tel: +420 2 20940 954 Head Fax: +420 2 20940 954 Thermal-Hydraulic Analyses Department Eml: [email protected] Nuclear Research Institute Rez plc 250 68 REZ

Dr. Alexander MIASNIKOV Tel.:+420 2 216 24 732State Office for Nuclear Safety Fax: +420 2 216 24 202Senovazne namesti 9 Eml.:[email protected] 00 Prague 1

142

142

Mr. Svatobor STECH Tel: +420 618 81 3690 Dukovany Nuclear power Plant Fax: +420618 866 495675 50 Dukovany Eml: [email protected]

FINLAND

Mr. Seppo KELPPE Tel.: +358 9 456 5026VTT Energy Fax: +358 9 456 5000P.O. Box 1604 Eml.: [email protected] VTT

Mr. Keijo VALTONEN Tel.: +358 (9) 759 331 Finnish Centre for Radiation Fax: +358 (9) 759 88382 and Nuclear Safety (STUK) Eml.: [email protected] P.O. Box 14 SF-00881 HELSINKI

FRANCE

Mr. Georges HACHE Tel.: +33 4 42 25 20 55IRSN/DRS Fax: +33 4 42 25 76 79

Bat 250 Eml.: [email protected] CEA/Cadarache 13108 ST PAUL LEZ DURANCE Cdx

Mme. Joelle PAPIN Tel.: +33 + 33 4 42 25 34 63IRSN/DRS/SEMAR Fax: +33 + 33 4 42 25 61 43

CEA/CEN/CADARACHE Eml.: [email protected] 13108 Saint Paul lez Durance Cedex

Mr. Nicolas TRICOT Tel.: +33 (0)1 46 54 90 34 IRSN/DES Fax: +33 (0)1 46 54 95 99 CEN FAR Eml.: [email protected] BP 17 92262 FONTENAY AUX ROSES CEDEX

Mr. Patrice DESCHAMPS Tel.:IRSN/DES Fax:

CEN FAR Eml.:[email protected] BP 17 92262 FONTENAY AUX ROSES CEDEX

143

143

Dr. Nicolas WAECKEL Tel : 04 72 82 72 44EDF/SEPTEN Fax : 04 72 82 77 1112-14 Avenue Dutrievoz Eml : [email protected] Villeurbanne Cedex F

Dr. Alain MAILLIAT Tel : +33 4 42 25 36 37Head of Project Fax: +33 4 42 25 29 29IRSN/DRS/DIR - Bat 250 Eml.: [email protected] - CEN CadaracheF-13108 St Paul-lez Durance Cedex

Mr. C. GETREY Tel.: +33 1 46 54 85 16IRSN/DES Fax: +33 1 46 54 95 99CEN/FAR Eml.: [email protected] Fontenay-aux-Roses

Mr. Jean-Charles LE PALLEC Tel. : +33 1 69 08 26 26CEA Saclay bat. 470 Fax : +33 1 69 08 23 81DM2S/SERMA/LCA Eml: [email protected] Gif-Sur-Yvette Cedex

Mr. A. FRICHET Tel. : +33 4 72 74 72 59Framatome ANP Fax : +33 4 72 82 77 1310 rue Juliette Récamier Eml.:alain.frichet@framatome-69456 LYON Cedex 06 anp.fr

GERMANY Dr. Heinz G. SONNENBURG Tel.: +49 89 320 04 578 GRS mbH Fax: +49 89 320 04 599 Forschungsgelaende Eml.: [email protected] D-85748 GARCHING

Dr. Lothar HEINS Tel: +49 9131 1892068Framatome ANP GmbH Fax: +49 9131 184799P.O. Box 3220 Eml: [email protected] Erlangen

Mr. Mathias BRETTNER Tel.: +49-421-323571Physikerbuero Bremen Fax: +49-421-328719Landweg 6 Eml: [email protected] Bremen

Mr. Siegfried LANGENBUCH Tel.: +49-89-32004-424GRS mbH Fax: +49-89-32004-599Forschungsgelaende Eml.: [email protected] GARCHING

144

144

HUNGARY Dr. Laszló MAROTI Tel.: +36 1 275 4090 Deputy Director Fax: +36 1 395 9293 Atomic Energy Research Institute Eml.: [email protected] Konkoly Thege 39 P.O. Box 49 H-1525 BUDAPEST 114

JAPANDr. Takehiko NAKAMURA Tel.: +81 (29) 282 6386Senior Engineer Fax: +81 (29) 282 5429Fuel Safety Research Laboratory Eml.: [email protected] of Reactor Safety ResearchJapan Atomic Energy Research InstituteTokai-mura, Naka-gun, Ibaraki-ken 319-1195

Mr. Katsuichiro KAMIMURA Tel: +81 3 4512 2671Senior Manager - Nuclear Fuel Department Fax: +81 3 4512 2699NUPEC Eml: [email protected] Floor, Fujita Kanko Toranomon Bldg.3-17-1 ToranomonMinato-ku, Tokyo 105-001

Mr. Tetsuo NAKAJIMA Tel: +81 (3) 4512 - 2742Senior Chief Engineer Fax: +81 (3) 4512 - 2799Institute of Nuclear Safety E-mail: [email protected] Power Engineering Corporation (NUPEC)7th Floor, Fujita-Kanko Toranomon Bldg. 3-17-1Toranomon, Minato-kuMinato-ku, Tokyo 105-0001

KOREA (REPUBLIC OF)Dr. C. B. LEE Tel: +82 42 868 2129Principal Research Scientist Fax: +82 42 868 8868Korea Atomic Energy Research Institute Eml: [email protected]. Box 105Yusung, Taejon 305-600

Dr. Sun-Ki Kim Tel. : +82-42-868-8661Senior researcher Fax. : +82-42-861-8819Korea Atomic Energy Research Institute Eml : [email protected]. Box 105Yusung, Taejon 305-600

145

145

RUSSIAN FEDERATIONProfessor Vladimir ASMOLOV Tel: (7-095) 196-9320Director RRC KI for R&D Fax: (7-095) 196-1702Kurchatov sq. 1 Eml: [email protected] 123182

Dr. Larisa YEGOROVA Tel: (7-095) 196-7283Head of Department NSI RRC KI Fax: (7-095) 196-1702Kurchatov sq. 1 Eml: [email protected] 123182

Dr. Valery MALOFEEV Tel: (7-095) 196-7466Chief of Laboratory NSI RRCKI Fax: (7-095) 196-1702Kurchatov sq. 1 Eml: [email protected] 123182

Mr. Iouri Vladimirovch PIMENOV Tel.: 007 095 239-41-98Senior Expert Fax: 007 095 239-44-04J - Stock Company”TVEL” Eml: [email protected] 24/26101000 Moscow

Ms. Olga Artiemovna NETCHAEVA Tel.: 007 095 190-62-36Senior Researcher Fax : 007 095 190-51-16,FSUE A.A. Bochvar VNIINM 095 196-65-91P.O.Box 369 VNIINM Eml: [email protected], Moscow

Mrs. L. N. ANDREEVA-ANDRIEVSKAYA Tel.: 007 095 190-62-36 Senior Researcher Fax: 007 (095) 190-51-16, VNIINM (095) 196-65-91 123060 Moscow, P.O. Box 369 Eml.: [email protected]

SPAINMr. M. José Manuel CONDE LOPEZ Tel.: +34 91 346 02 53Head, Nuclear Engineering D. Fax: +34 91 346 05 88Consejo de Seguridad Nuclear Eml.: [email protected] Dorado, 1128040 MADRID

Mr. José M. REY Tel.: 34-91-3460215Consejo de Seguridad Nuclear Fax.: 34-91-3460215c/Justo Dorado, 11 Eml.: [email protected] Madrid

Dr. Luis HERRANZ Tel: +34 91 3466219 CIEMAT Fax: +34 91 3466233 Dept. of Nuclear Fission Eml: [email protected] Avda. Complutense, 22 E-28040 Madrid

146

146

Dr. Manuel QUECEDO Tel.: 34-913474264Fuel Rod Technology Fax: 34-913474215ENUSA Eml: [email protected] Rusiñol 12,28040 Madrid

SWEDEN

Mr. Jan IN DE BETOU Tel.: +46 8 698 8459Swedish Nuclear Power Inspectorate Fax: +46 8 661 9086SKI Eml.: [email protected] Stockholm

Mr. David SCHRIRE Tel.:+46-155-22 16 30Studsvik Fax.:+46-155-26 31 56SE-611 82 Nyköping Eml: [email protected]

SWITZERLANDDr. Willem VAN DOESBURG Tel.: +41 56 310 3862Swiss Federal Nuclear Fax: +41 56 310 4979Safety Inspectorate (HSK) Eml.: [email protected] fuer die psi.chSicherheit der KernanlagenCH - 5232 VILLIGEN-HSK

Dr. Claude MAEDER Tel.: +41 56 310 3811Swiss Federal Nuclear Fax: +41 56 310 38 54Safety Inspectorate (HSK) Eml.: [email protected] fuer dieSicherheit der KernanlagenCH - 5232 VILLIGEN-HSK

Mr. Frank HOLZGREWE Tel.: ++41 (0)31 754 74 72Nuklearingenieur Fax: ++41 (0)31 754 71 25Ressort Physik Eml: [email protected] FMB Energie AGKernkraftwerk MühlebergCH-3203 Mühleberg

UNITED KINGDOMMr. J.R. JONES Tel.: + 44 1452 65 33 86British Energy Ltd. Fax: + 44 1452 65 22 06Barnet Way Eml.: [email protected] GL4 3RS.

147

147

UNITED STATES OF AMERICADr. Ralph O. MEYER Tel.: +1 301 415 6789Office of Nuclear Regulatory Research Fax: +1 301 415 5160USNRC Eml.: [email protected] DC 20555

Dr. Rosa YANG Tel.: +1 650 855 2481Nuclear Power Division Fax: +1 650-855 1026EPRI Eml.: [email protected] Hillview Ave P.O. Box 10412PALO ALTO, CA 94303

Mr. David J. DIAMOND Tel.: +1 631 344 2604Head Fax: +1 631 344.3957Nuclear Energy and Infrastructure Systems Division Eml: [email protected] National LaboratoryBNL-130 Upton, NY 11973-5000

International Organisations

OECD/Halden Reactor ProjectDr. Wolfgang WIESENACK Tel.: +47 (69) 21 2200Acting Project Manager Fax: +47 (69) 21 2201OECD Halden Reactor Project Eml.: [email protected] for EnergiteknikkOs Alle 13 - P.O. Box 173N-1751 HALDEN

OECD Nuclear Energy Agency, Issy-les-MoulineauxMr. Miroslav HREHOR Tel.: +33 1 45 24 10 58OECD Nuclear Energy Agency Fax: +33 1 45 24 11 10Le Seine St. Germain Eml.: [email protected], boulevard des IlesF-92130 ISSY-LES-MOULINEAUX

Dr. Carlo VITANZA Tel.: +33 1 45 24 10 62OECD Nuclear Energy Agency Fax: +33 1 45 24 11 10Le Seine St. Germain Eml.: [email protected], boulevard des IlesF-92130 ISSY-LES-MOULINEAUX

Unclassified NEA/CSNI/R(2003)8/VOL2 - Nuclear Energy ...

Documents

Transcript of Unclassified NEA/CSNI/R(2003)8/VOL2 - Nuclear Energy ...