IAEA-TECDOC-1697

Improvement of Computer CodesUsed for Fuel Behaviour

Simulation(FUMEX–III)

INTERNATIONAL ATOMIC ENERGY AGENCYVIENNA

ISBN 978–92–0–138610–6ISSN 1011–4289

IAEA

-TECD

OC

-1697 n

IMPR

OVEM

ENT O

F CO

MPU

TER C

OD

ES USED

FOR

FUEL B

EHAVIO

UR

SIMU

LATION

(FUM

EX–III)

IMPROVEMENT OF COMPUTER CODES USED FOR FUEL BEHAVIOUR

SIMULATION (FUMEX-III)

AFGHANISTANALBANIAALGERIAANGoLAARGENTINAARMENIAAUSTRALIAAUSTRIAAZERBAIJANBAHRAINBANGLADESHBELARUSBELGIUMBELIZEBENINBoLIVIABoSNIA AND HERZEGoVINABoTSWANABRAZILBULGARIABURKINA FASoBURUNDICAMBoDIACAMERooNCANADACENTRAL AFRICAN

REPUBLICCHADCHILECHINACoLoMBIACoNGoCoSTA RICACÔTE D’IVoIRECRoATIACUBACYPRUSCZECH REPUBLICDEMoCRATIC REPUBLIC

oF THE CoNGoDENMARKDoMINICADoMINICAN REPUBLICECUADoREGYPTEL SALVADoRERITREAESToNIAETHIoPIAFIJIFINLANDFRANCEGABoNGEoRGIAGERMANYGHANAGREECE

GUATEMALAHAITIHoLY SEEHoNDURASHUNGARYICELANDINDIAINDoNESIAIRAN, ISLAMIC REPUBLIC oF IRAQIRELANDISRAELITALYJAMAICAJAPANJoRDANKAZAKHSTANKENYAKoREA, REPUBLIC oFKUWAITKYRGYZSTANLAo PEoPLE’S DEMoCRATIC

REPUBLICLATVIALEBANoNLESoTHoLIBERIALIBYALIECHTENSTEINLITHUANIALUXEMBoURGMADAGASCARMALAWIMALAYSIAMALIMALTAMARSHALL ISLANDSMAURITANIAMAURITIUSMEXICoMoNACoMoNGoLIAMoNTENEGRoMoRoCCoMoZAMBIQUEMYANMARNAMIBIANEPALNETHERLANDSNEW ZEALANDNICARAGUANIGERNIGERIANoRWAYoMANPAKISTANPALAU

PANAMAPAPUA NEW GUINEAPARAGUAYPERUPHILIPPINESPoLANDPoRTUGALQATARREPUBLIC oF MoLDoVARoMANIARUSSIAN FEDERATIoNRWANDASAUDI ARABIASENEGALSERBIASEYCHELLESSIERRA LEoNESINGAPoRESLoVAKIASLoVENIASoUTH AFRICASPAINSRI LANKASUDANSWAZILANDSWEDENSWITZERLANDSYRIAN ARAB REPUBLICTAJIKISTANTHAILANDTHE FoRMER YUGoSLAV

REPUBLIC oF MACEDoNIAToGoTRINIDAD AND ToBAGoTUNISIATURKEYUGANDAUKRAINEUNITED ARAB EMIRATESUNITED KINGDoM oF

GREAT BRITAIN AND NoRTHERN IRELAND

UNITED REPUBLIC oF TANZANIA

UNITED STATES oF AMERICAURUGUAYUZBEKISTANVENEZUELAVIETNAMYEMENZAMBIAZIMBABWE

The following States are Members of the International Atomic Energy Agency:

The Agency’s Statute was approved on 23 october 1956 by the Conference on the Statute of the IAEA held at United Nations Headquarters, New York; it entered into force on 29 July 1957. The Headquarters of the Agency are situated in Vienna. Its principal objective is “to accelerate and enlarge the contribution of atomic energy to peace, health and prosperity throughout the world’’.

IAEA-TECDOC-1697

IMPROVEMENT OF COMPUTER CODES USED FOR FUEL BEHAVIOUR

SIMULATION (FUMEX-III)

REPORT OF A COORDINATED RESEARCH PROJECT 2008–2012

INTERNATIONAL ATOMIC ENERGY AGENCY VIENNA, 2013

COPYRIGHT NOTICE

All IAEA scientific and technical publications are protected by the terms of the Universal Copyright Convention as adopted in 1952 (Berne) and as revised in 1972 (Paris). The copyright has since been extended by the World Intellectual Property Organization (Geneva) to include electronic and virtual intellectual property. Permission to use whole or parts of texts contained in IAEA publications in printed or electronic form must be obtained and is usually subject to royalty agreements. Proposals for non-commercial reproductions and translations are welcomed and considered on a case-by-case basis. Enquiries should be addressed to the IAEA Publishing Section at: Marketing and Sales Unit, Publishing Section International Atomic Energy Agency Vienna International Centre PO Box 100 1400 Vienna, Austria fax: +43 1 2600 29302 tel.: +43 1 2600 22417 email: sales.publications@iaea.org http://www.iaea.org/books

For further information on this publication, please contact:

Nuclear Fuel Cycle and Materials Section International Atomic Energy Agency

Vienna International Centre PO Box 100

1400 Vienna, Austria Email: Official.Mail@iaea.org

© IAEA, 2013 Printed by the IAEA in Austria

March 2013

IAEA Library Cataloguing in Publication Data Improvement of computer codes used for fuel behaviour simulation (FUMEX-III) : report of a coordinated research project 2008 – 2012. – Vienna : International Atomic Energy Agency, 2013. p. ; 30 cm. – (IAEA-TECDOC series, ISSN 1011-4289 ; no. 1697) ISBN 978-92-0-138610-6 Includes bibliographical references. 1. Nuclear reactors – Safety measures. 2. Fuel burnup (Nuclear engineering) – Mathematical models – Data processing. I. International Atomic Energy Agency. II. Series. IAEAL 13–00790

FOREWORD

It is fundamental to the future of nuclear power that reactors can be run safely and economically to compete with other forms of power generation. As a consequence, it is essential to develop the understanding of fuel performance and to embody that knowledge in codes to provide best estimate predictions of fuel behaviour. This in turn leads to a better understanding of fuel performance, a reduction in operating margins, flexibility in fuel management and improved operating economics.

The IAEA has therefore embarked on a series of programmes addressing different aspects of fuel behaviour modelling with the following objectives:

To assess the maturity and prediction capabilities of fuel performance codes, and to support interaction and information exchange between countries with code development and application needs (FUMEX series);

To build a database of well defined experiments suitable for code validation in association with the OECD Nuclear Energy Agency (OECD/NEA);

To transfer a mature fuel modelling code to developing countries, to support teams in these countries in their efforts to adapt the code to the requirements of particular reactors, and to provide guidance on applying the code to reactor operation and safety assessments;

To provide guidelines for code quality assurance, code licensing and code application to fuel licensing.

This report describes the results of the coordinated research project on the “Improvement of computer codes used for fuel behaviour simulation (FUMEX-III)”. This programme was initiated in 2008 and completed in 2012. It followed previous programmes on fuel modelling: D-COM 1982–1984, FUMEX 1993–1996 and FUMEX-II 2002–2006.

The participants used a mixture of data derived from commercial and experimental irradiation histories, in particular data designed to investigate the mechanical interactions occurring in fuel during normal, transient and severe transient operation. All participants carried out calculations on priority cases selected from a matrix of cases identified at the first research coordination meeting and designed to support their individual priorities. These priority cases were chosen as the best available to help determine which of the many models used in the codes best reflect reality. The participants also used these cases for verification and validation purposes, as well as for inter-code comparisons.

FUMEX-III was made possible as a result of the support and dedication of many organizations and individuals. The IAEA would like to thank the Technical Working Group on Fuel Performance and Technology (TWGFPT) for suggesting and supporting the programme, the OECD/NEA Halden Reactor Project for providing experimental data, the OECD/NEA for supporting the International Fuel Performance Experiments Database and the participants for performing the calculations and submitting summaries and meeting contributions. The IAEA would also like to thank J.A. Turnbull (United Kingdom) and all those who prepared the intermediate working material and the final report. The IAEA officer responsible for this publication was J. Killeen of the Division of Nuclear Energy and Waste Technology.

EDITORIAL NOTE

The papers in these proceedings are reproduced as submitted by the authors and have not undergone rigorous editorial review by the IAEA.

The views expressed do not necessarily reflect those of the IAEA, the governments of the nominating Member States or the nominating organizations.

The use of particular designations of countries or territories does not imply any judgement by the publisher, the IAEA, as to the legal status of such countries or territories, of their authorities and institutions or of the delimitation of their boundaries.

The mention of names of specific companies or products (whether or not indicated as registered) does not imply any intention to infringe proprietary rights, nor should it be construed as an endorsement or recommendation on the part of the IAEA.

The authors are responsible for having obtained the necessary permission for the IAEA to reproduce, translate or use material from sources already protected by copyrights.

CONTENTS

1. INTRODUCTION ............................................................................................................ 1 1.1. The D-COM blind exercise ..................................................................................... 1 1.2. The FUMEX blind exercise .................................................................................... 3

1.2.1. Description of codes used in the FUMEX exercise .................................. 4 1.2.2. Experimental data used for the comparison exercise ................................ 4

1.3. FUMEX-II coordinated research project ................................................................. 5 1.4. The fuel performance experiments database ........................................................... 9

1.3.1. Description of FUMEX-II ......................................................................... 6 1.3.2. Conclusions from the FUMEX-II CRP ..................................................... 8

2. FUMEX-III ..................................................................................................................... 10

2.1. Objective ............................................................................................................... 10 2.2. Requirements for FUMEX-III ............................................................................... 10 2.3. Participants ............................................................................................................ 11 2.4. Cases ...................................................................................................................... 12 2.5. Additional information .......................................................................................... 12

3. DESCRIPTION OF CASES ........................................................................................... 16

3.1. Normal operation cases ......................................................................................... 16 3.1.1. US PWR 16×16 ....................................................................................... 16 3.1.2. AREVA idealized case ............................................................................ 17 3.1.3. Risø-3 AN2, AN3 .................................................................................... 19 3.1.4. R. E. Ginna rodlets .................................................................................. 19 3.1.5. Experimental fuel elements (EFE) 51 and 89 ......................................... 20 3.1.6. AECL BDL-406 ...................................................................................... 21

3.2. Gadolinia doped fuel ............................................................................................. 21 3.2.1. GAIN project ........................................................................................... 21

3.3. Transients .............................................................................................................. 22 3.3.1. IFA 535.5 and IFA 535.6: Effect of pre-pressurization on

fission gas release .................................................................................... 22 3.3.2. Risø-3 bump test II5 ................................................................................ 23 3.3.3. IRDMR experiment ................................................................................. 24 3.3.4. KOLA-3 MIR ramp experiment .............................................................. 25

3.4. Load follow ........................................................................................................... 27 3.4.1. Bundles JC and NR ................................................................................. 27 3.4.2. IFA-519.9 ................................................................................................ 27

3.5. PCI criterion .......................................................................................................... 28 3.5.1. SUPER-RAMP and INTER-RAMP ....................................................... 28 3.6. Pellet–cladding mechanical interaction ................................................... 29 3.6.1. Risø-2 GE-m and Risø-3 rod II-3 ............................................................ 29 3.6.2 Risø-3 rod GE-7 ...................................................................................... 29 3.6.3. OSIRIS .................................................................................................... 29 3.6.4. CNEA-MOX-RAMP experiments .......................................................... 30

3.7. RIA and LOCA ..................................................................................................... 30 3.7.1. BWR segments FK-1, FK-2 and FK-3 under RIA test

conditions in NSRR ................................................................................. 30 3.7.2. CANDU LOCA tests FIO-130 and FIO-131 .......................................... 31 3.7.3. IFA-650 ................................................................................................... 32

3.8. MOX ...................................................................................................................... 32 3.8.1. IFA 629 rod 1 .......................................................................................... 32 3.8.2. PRIMO .................................................................................................... 33

4. DESCRIPTION OF THE CODES ................................................................................. 34

4.1. Code information ................................................................................................... 34 4.1.1. Model overview ...................................................................................... 34 4.1.2. Transient modelling ................................................................................ 34 4.1.3. Fuel swelling ........................................................................................... 34 4.1.4. Fuel-clad friction ..................................................................................... 35 4.1.5. Summary ................................................................................................. 35

5. CODE COMPARISONS ................................................................................................ 40

5.1. Overview ............................................................................................................... 40 5.2. Normal operation cases ......................................................................................... 40

5.2.1. US 16×16 PWR: Comparison between solid and hollow pellets at high burnup ............................................................................................. 40

5.2.2. AREVA idealized case ............................................................................ 47 5.2.3. RE Ginna rodlet 2 and rodlet 4 ............................................................... 49 5.2.4. Normal operation: Elements RO-51 and RO-89 ..................................... 50

5.3. Gadolinia doped fuel ............................................................................................. 52 5.3.1. GAIN rodlets 301 and 701 ...................................................................... 52

5.4. Transients .............................................................................................................. 55 5.4.1. IFA 535.5 and IFA 535.6: Effect of pre-pressurazation on fission gas

release ...................................................................................................... 55 5.4.2. Risø-3, bump test II5 ............................................................................... 58 5.4.3. CANDU bundles measured in IRDMR FIO 118/119 ............................. 64 5.4.4. Kola–MIR test ramps .............................................................................. 64

5.5. Load follow ........................................................................................................... 69 5.5.1. Tests JC and NR ...................................................................................... 69 5.5.2. The IFA-519 study .................................................................................. 71

5.6. PCI criterion .......................................................................................................... 71 5.6.1. The INTER-RAMP and SUPER-RAMP failure criterion exercise ........ 71

5.7. PCMI ..................................................................................................................... 75 5.7.1. Ramp test Risø GE7 ................................................................................ 75 5.7.2. OSIRIS rods J12 and H09 ....................................................................... 79

5.8. RIA and LOCA ..................................................................................................... 81 5.8.1. BWR segments FK-1, FK-2 under RIA test conditions in NSRR .......... 81 5.8.2. CANDU LOCA test FIO-131 ................................................................. 85 5.8.3. CANDU MOX ramp test CNEA BU15 .................................................. 86

5.9. MOX ...................................................................................................................... 86 5.9.1. IFA 629 rod 1 .......................................................................................... 86 5.9.2. PRIMO rod BD8 ..................................................................................... 88

6. DISCUSSION ................................................................................................................. 89

6.1. Issues raised ........................................................................................................... 89 6.2. Data needs for fuel performance modelling .......................................................... 89 6.3. Thermal and fission power .................................................................................... 90 6.4. Burnup calculations ............................................................................................... 90 6.5. High burnup fission gas release ............................................................................. 91

6.6. Transient fission gas release .................................................................................. 92 6.7. Swelling ................................................................................................................. 91 6.8. Pellet–cladding mechanical interaction (PCMI) ................................................... 91 6.9. CANDU ................................................................................................................. 93 6.10. RIA/LOCA ............................................................................................................ 93 6.11. MOX analysis ........................................................................................................ 93

6.11.1. Temperature and thermal conductivity degradation ............................... 93 6.11.2 FGR ......................................................................................................... 94 6.11.3. He generation and release ....................................................................... 94 6.11.4. Mechanical behaviour ............................................................................. 95

7. CONCLUSIONS ............................................................................................................ 95

REFERENCES ......................................................................................................................... 97

LIST OF ABBREVIATIONS .................................................................................................. 98

ANNEX I: REPORT OF THE FIRST RESEARCH COORDINATION MEETING ........ 101

ANNEX II: REPORT OF THE SECOND RESEARCH COORDINATION MEETING ... 108

ANNEX III: REPORT OF THE THIRD RESEARCH COORDINATION MEETING ....... 118

CONTRIBUTORS TO DRAFTING AND REVIEW ........................................................... 125

1

1. INTRODUCTION

Reliable prediction of fuel behaviour is a basic requirement for safe operation of nuclear fuel, for design purposes and for fuel performance assessments. As a consequence, it is essential to develop the understanding of fuel performance and to embody this knowledge in computer codes to provide the best estimate predictions of fuel behaviour. This, in turn, leads to a reduction in operating margins, flexibility in fuel management and improved operating economics. The ultimate goal is a description of fuel behaviour in both normal and off-normal conditions. From this knowledge, operation rules can be derived to prevent fuel failures and the release of fission products to the environment, also, in extreme cases, to prevent escalation of fuel and core damage and the consequential hazards.

Fuel modelling codes have been developed independently by many Member States and other States are

customising international codes (such as TRANSURANUS) for use in their countries. These codes have been developed with differing objectives and for various reactor types, some have been designed as research tools whilst others are used for licensing. The different codes are in differing stages of development and have varied validation and verification status. These codes need to be continuously updated and maintained to ensure that they are fit for the design purpose.

The IAEA has run a series of coordinated research projects (CRPs) in the area of fuel modelling, helping

Member States to develop state of the art codes, starting in 1982 with "The development of computer models for fuel element behaviour in water reactors" (D-COM), which terminated in 1984, the FUMEX CRP “Fuel modelling at extended burnup” which started in 1993 and concluded in 1996 and the FUMEX-II CRP, “Fuel modelling at extended burnup, FUMEX-II” which started in 2002 and concluded in 2006. This previous work has seen marked improvements in the capabilities of codes to predict temperature and fission gas release to high burnups in light water reactors (LWRs), and to a lesser extent in pressurised heavy water reactors (PHWRs).

However, fuel burnup is continuing to increase in many countries, and with the current programme of new

build of Advanced Reactors, new and more onerous fuel cycles are foreseen, which the fuel computer models will need to properly address. These factors are stretching the capability of fuel codes to model the behaviour of fuel under onerous operating conditions, at the end of life and particularly for transient and accident conditions. Further, the fuel itself is being developed to meet increasing demands of safety and endurance. This means that new materials, such as advanced cladding or doped fuels are being introduced and codes need to be developed to deal with these changes. Particular concerns include for LWRs, the pellet–cladding interaction at high burnup, where new phenomena are occurring, such as the rim effect and gap closure in Russian designed LWRs (WWERs). For PHWR systems there is special interest in slightly enriched fuel (SEU), with a consequent large increase in discharge burnup.

Fuel performance modelling requires a continuous improvement process, and the work of the IAEA has

been greatly welcomed by previous participants. The Technical Working Group on Fuel Performance and Technology (TWGFPT) requested that a third FUMEX exercise be undertaken at its meetings in April 2006 and April 2007, and this report describes the work undertaken for this exercise, the CRP on “Improvement of computer codes used for fuel behaviour simulation (FUMEX-III)”. It was conducted over the period 2008–2011 with twenty-two countries participating. 1.1. THE D-COM BLIND EXERCISE

The list of participants in the D-COM blind exercise is given in Table 1. The detailed consultants’ report presenting the state of the art in modelling the fuel rod behaviour and including a comprehensive review of fuel rod computer codes at that time is given in Refs [1, 2].

2

TABLE 1. PARTICIPANTS IN THE D-COM BLIND PROBLEM Country Organization Code

Denmark Risø Experiment

Argentina CNEA BACO

Belgium BN COMETHE III-L

Canada AECL ELESIM2.MOD10

Czechoslovakia Rež PIN/RELA

F.R. Germany/CEC TU-Darmstadt/ITU URANUS

Finland VTT FRAPCON-2

France CEA-Grenoble CREOLE

France EdF CYRANO-2

France CEN-Saclay RESTA

India BARC PROFESS

Japan CRIEPI FEMAXI-III

Sweden Studsvik GAPCON-SV

United Kingdom BNFL HOTROD

United Kingdom UKAEA MIN1PAD-E

United States of America Exxon RAMPX2

As part of this programme, a code exercise was conducted [3], where the objective was to investigate the

capability of fuel performance codes to predict fission gas release. The test cases to be calculated by the codes consisted of three mini pins irradiated together (test HP 096) in the Danish DR 3 test reactor to a burnup of 32 MW·d/kgU. Two pins were finally bumped together with average heat ratings of 33.7 and 36.2 kW/m respectively at the end of the bump. The blind code predictions were presented at the OECD/NEA/CSNI/IAEA Specialists' Meeting on Water Reactor Fuel Safety and Fission Product Release in Off-Normal and Accident Conditions, Risø National Laboratory, 1983 [4]. However, the results were not included in the proceedings of the meeting, but some are given in Ref. [5].

The main conclusions from the D-COM exercise were as follows:

Temperature: Temperature predictions showed a large spread. Fission gas release: Fission gas release during the base irradiation was in fair agreement with

experimental values. The fission gas release during the transient (bump test) was under-predicted by most of the codes.

Mechanical behaviour: Since the exercise concentrated on the thermal behaviour and gas release, many

participants did not provide dimensional data. Of those codes which submitted mechanical data, most predicted the cladding creep down reasonably well, mechanical data during the ramp were scarce and showed considerable spread.

The D-COM blind code exercise was considered by participants as being very valuable in promoting

discussions among modellers. A better knowledge of the centre line temperature during base irradiation was identified as an area of further development. It was also stated in the conclusions that basic phenomena such as gaseous swelling, transient gas release and grain growth should be better known during transients.

The subsequent experimental programmes both at Halden and Risø addressed these requests. Within these

projects it was demonstrated that the fuel thermal conductivity degrades with burnup and can be modelled by an additional phonon contribution. The effect of this degradation is a higher fuel temperature which partially explains the general under-prediction of fission gas release in the transient of the D-COM blind prediction. It is of interest to note that some modelling groups that participated in the D-COM exercise also participated in the FUMEX blind exercise. This list is shown in Table 2.

3

1.2. THE FUMEX BLIND EXERCISE

Following the D-COM exercise, the IAEA initiated a second code comparison exercise in 1993 addressing fuel thermal performance and fission gas release at high burnup as well as aspects of pellet–cladding mechanical interaction. There were a total of six cases, FUMEX 1–6 including 10 rods, which represented actual irradiations in the OECD Halden Heavy Water Reactor in Norway.

The FUMEX CRP was initiated by the IAEA following a recommendation of the International Working

Group on Fuel Performance and Technology (IWGFPT). It was conducted over the period 1993–1996. Fifteen countries took part.

The elements of the CRP were defined as follows:

A blind prediction carried out by the participants on data provided by the Halden Project, Norway, in the form of irradiation histories, in-pile measurements and post irradiation examination (PIE) of six experiments involving 10 fuel rods. Only after all the predictions were submitted were the measurements released;

A comparison of calculations carried out after code improvement on the 10 rods of the FUMEX blind exercise;

The definition of eight simplified cases, to assess code response to changes of single parameters such as internal gas composition, burnup, power steps, and a statistical analysis of two of the simplified cases.

Follow-up of code status, progress in modelling and modification made at research coordination meetings (RCMs), also providing a forum for discussion and interaction among participants.

The participants of the FUMEX CRP are shown in Table 2.

TABLE 2. PARTICIPANTS IN THE FUMEX EXERCISE

Country Organization Code

Norway/OECD Halden Experimental data provider

Argentina CNEA BACO

Bulgaria INRNE PIN micro

Canada AECL ELESIM.MOD11

CEC ITU TRANSURANUS

China CIAE FRAPCON-2

Czech Republic NRI Rež PIN/W

Finland VTT ENIGMA 5. 8f

France CEA/DRN METEOR-TRANSURANUS

France EdF TRANSURANUS-EdF 1.01

India BARC PROFESS

India BARC FAIR

India NPC FUDA

Japan CRIEPI EIMUS

Japan NNFD TRUST Ib

Romania INR ROFEM-1B

Russian Federation IIM START 3

Switzerland PSI TRANSURANUS-PSI

United Kingdom BNFL ENIGMA 5.2

United Kingdom NE ENIGMA 5.8 D

Note: Turkey joined the CRP at the time of the 3rd RCM in Bombay. Turkey used a version of FRAPCON-2.

4

In early 1993 the specifications of six experiments performed at the Halden Project (Norway) were distributed to the participants. The first research co-ordination meeting took place in Halden, 28 June–1 July 1993. During this meeting a description of the 19 codes was given and the preliminary results were released.

The second RCM took place on 15–16 September 1994 in Windermere (United Kingdom). Here, the outcome of the code predictions was discussed along with the future actions to be taken by the participants in code development and improvements. There was a general agreement that each participant should rerun the original FUMEX study, conduct a new study on simplified cases and a limited sensitivity study based on agreed uncertainties of power and dimensions to investigate the sensitivity of predictions.

The third RCM was held in Mumbai (India), 1–5 April 1996. The meeting focused on elementary model improvement, the impact of the FUMEX programme and the recommendations from the participating countries. In this meeting the role of quality assurance in developing and maintaining fuel performance codes was also introduced. The final report [6] provides a description of the experiments chosen, an overview of the codes used by participants in the exercise, and the improvements implemented as a consequence of FUMEX. A commentary is given regarding the various aspects of fuel behaviour tested and a detailed quantitative comparison is made between experimental data and code predictions. The report concludes with a discussion of the main findings of the exercise, the identified improvements and shortcomings in codes and modelling, and outstanding technical issues that require further attention.

1.2.1. Description of the codes used in the FUMEX exercise

Within the FUMEX exercise, blind predictions were submitted from 15 countries employing 19 codes or code variants. All the codes in the exercise used an axi-symmetric fuel rod representation and consisted of three main parts: Thermal analysis including gap conductance models which account for different pin pressures, gas

compositions and gap sizes; standard correlations for the thermal conductivity of fuel and cladding are used. Standard numerical techniques such as finite difference (FD) and finite element (FE) methods are applied.

Mechanical analysis including cracking and relocation of fuel pellets; in a few cases a simplified mechanical treatment of the fuel is adopted. However, most codes are based on an axi-symmetric, modified plane strain assumption. Two codes offer the capability of a two dimensional treatment. FD and FE methods are used.

A variety of physical models or empirical correlations are used for densification, swelling, fission gas release, grain growth, etc. The number of executable statements ranges from 2000 up to 30 000 and all the code descriptions claimed that the codes represented state-of-the-art modelling. Two codes were specifically designed and validated for heavy water reactors (HWR) with a collapsible cladding. As to be expected, these codes showed some deficiencies in predicting an open gap situation, and modifications were necessary when applied to the Halden irradiated rods.

1.2.2. Experimental data used for the comparison exercise

The FUMEX irradiations were all provided by the OECD Halden Reactor Project. They represented a selection of experiments from the Halden Project fuel testing programme, which focussed on the consequences of extended burnup on fuel operation. The six cases can be summarized briefly as follows: FUMEX 1 This data set represents the irradiation of production line pressurised water reactor (PWR) type

fuel under benign conditions. Temperatures remained low but increased slightly with burnup. FUMEX 2 This was a small diameter rod designed to achieve rapid accumulation of burnup. Temperatures

were estimated to remain low. The internal pin pressure was measured in-pile and an assessment of fission gas release (FGR) was also provided by PIE.

FUMEX 3 This case consisted of 3 short rods equipped with centreline thermocouples each with a different gap and fill gas composition. After steady state irradiation to approximately 30 MW·d/kgUO2, they were given a severe increase in power (power ramp).

FUMEX 4 Two rods filled with 3 bar He and 1 bar He/Xe mixture were irradiated to approximately 33 MW·d/kgUO2. Both rods experienced a period of increased power part way through the irradiation.

FUMEX 5 The test case comprised a single rod base irradiated at low power to 16 MW·d/kgUO2 with a power ramp and a hold period at the end of life. The main purpose of this case was to assess pellet clad mechanical interaction (PCMI) and FGR under ramp conditions.

FUMEX 6 Two rods were base irradiated at low power. The rods were refabricated to include pressure transducers. Rod internal pressure was monitored during power ramps, one fast, one slow.

5

The response to the FUMEX programme was very encouraging with a high degree of participation from Member Countries. All agreed that it was a worthwhile exercise and that the cases chosen were stringent tests of model and code performance. The exercise was useful in demonstrating the strong points of the codes as well as highlighting deficiencies where improvements were necessary. As a consequence most of the codes underwent some development during the programme. It was also apparent that many of the codes had been developed on only a limited database and that the FUMEX cases provided a valuable addition. As a result of the FUMEX exercise, the following points were noted. It was universally recognized that the fuel conductivity decreased significantly with burnup, and at the end of the exercise, all codes included a treatment of this phenomena. It was in the area of thermal performance that the greatest improvements were made. The exercise showed that difficulties still remained with modelling fission gas release. However, through

refining existing models and the introduction of new models there was a general improvement in predictive capabilities;

It was apparent that the major lack of progress was in the area of mechanical interaction. This was considered to be an important omission with adverse consequences on many aspects of fuel modelling;

The exercise showed that modern codes could be run on state-of-the-art PCs without difficulty. Despite the complexity and degree of difficulty of the experimental cases chosen for the comparison, in general, the codes could handle the volume of data and required mathematical convergence without difficulty;

Quality assurance (QA) was recognized as an essential part of the code development process. A further conclusion was that there was a need for technical meetings or workshops on specific technical issues, and over the next few years, these were held at CEA Cadarache as follows: Thermal Performance in Light Water (High Burnup) Fuels, 3–6 March 1998, [7]; Seminar on Fission Gas Behaviour in Water Reactor Fuels, 26–29 September 2000, [8]; Pellet–Clad Interaction in Water Reactor Fuels (PCI-2004), 9–11 March 2004, [9].

1.3. FUMEX-II COORDINATED RESEARCH PROJECT

In response to requests from participants of the earlier code comparison exercises, the new CRP, FUMEX-II, was launched in December 2002. The general purpose of this exercise was to expose code developers to a wide ranging database of information, namely, the IFPE Database, and, through a large number of participants, to assist compilers of the IFPE Database to correct errors, detect missing data and search for additional datasets. More specifically, it was agreed that the FUMEX-II CRP would concentrate on the predictive capabilities of codes at extended burnup, i.e., under conditions where restructuring of the pellet rim had been observed by PIE. Unlike the former exercises which required ‘blind’ predictions, all of the data were released at the start of FUMEX-II.

The 19 participants of the FUMEX CRP exercise were requested to prioritize the topics they wished for

inclusion in the new CRP. Fifteen answers were received and the topics listed in Table 3 were identified as important points for code improvement. TABLE 3. QUESTIONNAIRE FOR PREFERRED TOPICS FOR FUMEX-II CRP

Topic Number of answers

A. Availability of a comprehensive database for code validation 15

B. Influence of the high burnup 'rim' structure on thermal performance and fission gas release 13

C. Transient data on reactivity insertion accidents (RIA) and loss of coolant accidents (LOCA) 13

D. The influence of densification and swelling on thermal performance 11 E. Mechanical treatment of fuel pellets and PCMI 10 F. Data on mixed oxide (MOX) fuel 10 G. Data on intra-granular microstructure 10

6

1.3.1. Description of FUMEX-II The key elements of the FUMEX-II CRP were defined by a panel of experts at a meeting held in Vienna

from 26–29 November 2001. This group drew up a list of potential cases for participants to use to calibrate and compare their predictions. From this list, participants in FUMEX-II were requested to perform calculations for the six cases identified as high priority and a minimum of a further 4 cases at their discretion.

Subsequent to constructing the original list of cases it was found that a number of cases were not to be

available. In particular, the CEA GONCOR dataset and the WWER fuel irradiated in the Kola-3 reactor and ramp tested in the MIR test reactor. Data from the HATAC and REGATE experiments were offered as substitutes for the GONCOR dataset and the Russian data was finally made fully available after the end of the CRP.

As well as real experimental irradiation histories, the participants of FUMEX-II were also asked to model

a group of idealized cases, designed to test the capabilities of the codes and to test their capabilities at very high burnup. The seven prepared cases included two idealized commercial power histories and corresponding nominal fission gas release measurements based on the experience of the fuel vendors BNFL and AREVA; two cases explored the range of power and burnup of Canadian PHWR (CANDU) fuel operation, extended to high burnup; two cases were very simple power histories extending to 100 MW·d/kgU and the final case required participants to mimic the experimental Vitanza threshold for 1% fission gas release [10].

The cases agreed by the participants at the end of the first RCM are given in Table 4 where the six highlighted cases were considered to be priority cases for all participants to complete. Further priority cases were selected at the second RCM, the notes of this meeting are given in Annex II, and these are identified with an asterisk. The table includes the Kola/MIR data that was only available after the end of the project and lists two severe transient cases; a reactivity insertion accident (RIA) and a loss of coolant accident (LOCA). These cases were not finalized.

7

TABLE 4. LIST OF CASES FOR FUMEX-II, HIGH PRIORITY CASES ARE IN BOLD

No. Case identification Measurements made for comparison 1. Halden IFA 534.14, rod 18 EOL FGR and pressure, grain size 22 m,

Bu 52 MW·d/kgUO2 2. Halden IFA 534.14, rod 19 EOL FGR and pressure, grain size 8.5 m,

Bu 52 MW·d/kgUO2 3. Halden IFA 597.3, rod 7 Cladding elongation, at Bu 60 MW·d/kgUO2 4. Halden IFA 597.3, rod 8 FCT, FGR at Bu 60 MW·d/kgUO2 5. Halden IFA 507, TF3 Transient temperature during power increase 6. Halden IFA 507, TF5 Transient temperature during power increase 7. REGATE FGR and cladding diameter during and after a transient at Bu

47 MW·d/kg 8. HATAC FGR and cladding diameter during and after a transient at Bu 49

MW·d/kg 9.a Kola-3, rod 7 from FA222 FGR, pressure and creepdown at Bu 55 MW·d/kgUO2 10. Kola-3, rod 52 from FA222 FGR, pressure and creepdown at Bu 46 MW·d/kgUO2 11.a Kola-3, rod 86 from FA222 FGR, pressure and creepdown at Bu 44 MW·d/kgUO2 12. Kola-3, rod 120 from FA222 FGR, pressure and creepdown at Bu 50 MW·d/kgUO2 13. Risø-3 AN2 Radial distribution of fission products and FGR-EOL,

Bu 37 MW·d/kgUO2 14. Risø-3 AN3 FGR and pressure-EOL, FCT,

Bu 37 MW·d/kgUO2 15. Risø-3 AN4 FGR and pressure-EOL, FCT,

Bu 37 MW·d/kgUO2 16. HBEP, rod BK363 FGR-EOL, Bu 67 MW·d/kgUO2 17. HBEP, rod BK365 Fission products and Pu distribution, FGR-EOL,

Bu 69 MW·d/kgUO2 18. HBEP, rod BK370 Fission products and Pu distribution, FGR-EOL,

Bu 51 MW·d/kgUO2 19. TRIBULATION, rod BN1/3 Pressure, FGR, cladding creepdown, Bu 52 MW·d/kgUO2 20. TRIBULATION, rod BN1/4 Pressure, FGR, cladding creepdown, Bu 51 MW·d/kgUO2 21. TRIBULATION, rod BN3/15 Pressure, FGR, cladding creepdown, Bu 51 MW·d/kgUO2 22. EDF/CEA/FRA, rod H09 Fission products and Pu distribution, FGR-EOL,

Bu 46 MW·d/kgUO2 23. Kola-3 + MIR test Temperature during ramp, FGR-EOL, Bu 55 MW·d/kgUO2 24. Kola-3 + MIR test Pressure-EOL, Bu 55 MW·d/kgUO2 25. RIA to be specified (real data or simplified case) 26. LOCA to be specified (real data or simplified case) 27. Simplified cases (1) Temperature vs Bu for onset of FGR

(2a) FGR for constant 15 kW/m to 100 MW·d/kgU (2b) FGR for 20 kW/m at BOL decreasing linearly to 10 kW/m at 100 MW·d/kgU (2c) FGR for idealized history supplied by BNFL (2d) FGR for idealized history supplied by FANP (3a) FGR for CANDU idealized history (3b) FGR for CANDU idealized history

a Further priority cases selected at the second RCM. Note: FCT — fuel centre temperature; Bu — burnup; BOL — beginning of life; EOL — end of life;

The list of participants and their codes and affiliations are given in Table 5. Several participants used

more than one code and often code variants and development versions were used. The Indian participation was split into two teams using different codes.

8

TABLE 5. LIST OF CODES AND ORGANIZATIONS PARTICIPATING IN THE FUMEX-II EXERCISE Country Institute Code

Argentina CNEA BACO Belgium Nuclear Research Centre, SCK CEN FEMAXI-PLUTON

FRAPCON 3.2 MACROS-2

Bulgaria INRE PIN w99 TRANSURANUS (WWER)

Canada AECL ELESTRES China China Institute of Atomic Energy METEOR Czech Republic Nuclear Research Institute, Rež FEMAXI-V PIN/PIN2FRAS EC JRC Institute for Transuranium

Elements TRANSURANUS

Finland VTT ENIGMA IMAGINE

Germany/France FRAMATOME ANP GmbH FANP Development code (COPERNIC-3)

India BARC FAIR FUDA PROFESS

Japan NUPEC FEMAXI JNES Republic of Korea Korea Atomic Energy Research Institute INFRA

Romania Institute for Nuclear Research TRANSURANUS DCHAIN5V

Russian Federation A.A. Bochvar Res. Institute of Inorganic Materials

START-3

Switzerland PSI PSI version TRANSURANUS

UK BNFL ENIGMA-B 7.7 Norway OECD Halden Reactor Project Data provision and support OECD/NEA IFPE data bank

1.3.2. Conclusions from the FUMEX-II CRP

The final report of the FUMEX-II project [11] showed that the results of code predictions against the priority cases showed that fuel temperature modelling had been much improved since the previous FUMEX CRP. Fuel centre temperature predictions were generally good, and matched the data well, up to burnups of around 60 MW·d/kgU. The agreement was good for both normal operation and during power ramps.

The modelling also showed good agreement for fission gas release at burnups close to current commercial

limits (around 50 MW·d/kgU). However, it was recognised that standard models did not account for an increase in fission gas release rates observed at high burnups and the teams used various options and additional modelling in their codes to try to account for this phenomenon. Three distinct approaches were tried: Allowing fission gas release directly from the rim structure seen at the periphery of pellets at high burnup.

Modelling choices include varying the retentive capacity of this region and in determining how to define the extent of the rim region. Evidence for this mechanism comes from the existence of the rim structure, which seems to initiate at the same time as the additional release;

Allowing release of additional gas from saturated regions of the fuel, where the saturation is temperature dependent and the additional release comes from the pellet interior. Modelling choices here lie in determining the saturation level and the temperature dependence of this effect;

Allowing an additional burnup dependence on the diffusion and re16-solution parameters used in standard models. Release of fission gas is enhanced in the pellet centre with this approach. Experimental data on fission gas distributions in high burnup had not been sufficiently clear to allow the

teams to be able to positively distinguish between these models, though each approach predicted different distributions of retained fission gas. The reason why release from the rim was excluded in the latter two approaches was due to the fact that the available experimental data seemed to show significant retention of

9

fission gas in this region. Difficulties in interpretation arose from determining the concentration of fission gas retained in bubbles in the rim structure, which were not well measured by standard electron optical techniques. Additional data together with further interpretation and detailed examination of existing data was seen to be necessary.

The modellers noted several important issues and uncertainties for high burnup modelling which were

addressed in part during the later stages of the FUMEX-II CRP. These included: Accurate calculations of the burnup dependent radial power profile, i.e. plutonium build-up at the rim. What is the effect of the high burnup structure (HBS) seen at the rim? What is its impact on the thermal performance and FGR behaviour of the fuel rod? Is a separate treatment of this HBS region required for successful modelling? What are appropriate conditions for the formation of the HBS? At what burnup does the enhanced release begin? What temperature limits should apply to the models? What are the effects of pressure, grain size, dopants or other details of fuel rod manufacture?

The additional priority cases allowed the participants to attempt a consensus on some of these issues. In

particular the modelling of the grain size effect would appear to be in reasonable agreement with the scale of the effect in the experimental data.

The modellers found difficulty with the Risø data in particular, with most under-predicting the measured

FGR, however it was not possible to discriminate between the models for high burnup release, though detailed examination of the results continued. The overall results showed no diminution in the predictive capability of the codes with burnup, it would appear that all of the modelling approaches tried to date were adequate to explain the high burnup releases available in the CRP list of cases. Whilst some modellers relied heavily on the rim structure to enhance release at high burnup, others did not, and there was not sufficient experimental data available to allow discrimination between the competing models in their release predictions, though there is experimental evidence of fission gas retention at the rim.

Not many codes have good mechanical modelling capabilities, and the results that were obtained were

limited. Diametral swelling was a difficult area for many codes and their models for this were often very sensitive to small changes in the modelling assumptions. The initial rod elongation in a transient due to PCMI was reasonably well predicted by a few codes, but relaxation of the rod growth was rarely modelled. 1.4. THE FUEL PERFORMANCE EXPERIMENTS DATABASE

At the same time as the FUMEX exercise was being undertaken, the OECD/NEA Nuclear Science Committee (NSC) Task Force recommended the compilation of a public domain database on fuel performance for the express purpose of fuel performance code development and validation. In the light of the experience during the FUMEX exercise, the Agency actively supported this initiative and made available both data and funds for what is now known as the International Fuel Performance Experiments (IFPE) Database.

The database was further developed as a public domain database and was utilised and expanded during

the FUMEX-II exercise. The aim of the IFPE Database is to provide, in the public domain, a comprehensive and well-qualified database on zircaloy-clad UO2 fuel and recently MOX fuel for model development and code validation. The data encompass both normal and off-normal operation and include prototypic commercial irradiations as well as experiments performed in material testing reactors (MTR). To date, the database contains over 1200 individual cases, providing data on fuel centreline temperatures, dimensional changes and FGR either from in-pile pressure measurements or PIE techniques, including puncturing, electron probe micro-analysis (EPMA) and X ray fluorescence (XRF) measurements. This work in assembling and disseminating the database is carried out in close cooperation and coordination between OECD/NEA and the IAEA and the IFE/OECD/Halden Reactor Project.

The data sets are dedicated to fuel behaviour under thermal reactor irradiation, and every effort has been

made to obtain data representative of boiling water reactor (BWR), pressurised water reactor (PWR), WWER, commercial advanced gas-cooled reactors (CAGR) and PHWR conditions. In each case, the data set contains information on the pre-characterisation of the fuel, cladding and fuel rod geometry, the irradiation history presented in as much detail as the source documents allow, and finally any in-pile or PIE measurements that were made. Special emphasis is given to data relevant for current issues such as behaviour at high burnup. The

10

were made. Special emphasis is given to data relevant for current issues such as behaviour at high burnup. The database contains, besides the compilation and evaluation of the experimental data, also the detailed primary documents from which the data were derived. The compilations contain for user convenience a synthesis with the data required for model development and validation. The IFPE contains all cases investigated both in the FUMEX and FUMEX-II exercises. Through the FUMEX exercises, feedback from modellers could be used to improve the content by removing some inconsistencies or errors. The IFPE database is now widely used in about 100 institutions in more than 30 countries. Feedback from users has been essential to ensure that the database improves with its use.

This database is restricted to thermal reactor fuel performance; principally with standard product Zircaloy

clad UO2 fuel, although the addition of advanced products with fuel and clad variants is not ruled out. Emphasis has been placed on including well-qualified data that illustrate specific aspects of fuel performance. Of particular interest to fuel modellers are data on: fuel temperatures, FGR, fuel swelling, clad deformation (e.g. creep-down, ridging) and mechanical interactions. Data on these issues are of great value if measured in-pile by dedicated instrumentation and in this respect, the IFPE database is fortunate in having access to several diverse experiments.

In addition to direct in-pile measurement, every effort is made to include PIE information on clad

diameters, oxide thickness, hydrogen content, fuel grain size, porosity, EPMA and XRF measurements on caesium, xenon, other fission product and actinides.

2. FUMEX-III

In response to requests from Member States and previous participants of the FUMEX and FUMEX-II CRPs, and following a recommendation of the Technical Working Group on Fuel Performance and Technology, the IAEA decided to coordinate a further fuel modelling exercise. This exercise was designed to be more focussed on mechanical interaction between the pellet and cladding and the effects in transient and accident conditions than the previous CRPs. A further consideration was to try to bring together modellers working on different reactor systems and from a wide range of organizations so that experience and ideas could be shared.

The first RCM was held in Vienna, 10–12 December 2008, the second was held in Pisa, Italy,

1-4 June 2010, and the final meeting in Vienna, 5–8 December 2011. Consultants meetings were held in Vienna from 30 April–2 May 2008, on 2 October 2009, and on 9 December 2011. 2.1. OBJECTIVE

The objective of the CRP on “Improvement of Computer Codes used for Fuel Behaviour Simulation” (FUMEX-III, 2008–2012) was to improve the predictive capabilities of codes used in fuel behaviour modelling for extended burnup and under transient conditions. In extending the previous CRPs on this subject (FUMEX, 1993–1996 and FUMEX-II, 2002–2006), the focus was on the topics of fission gas release, pellet to cladding interaction and dimensional changes at extended burnups for LWRs of above 60 MW·d/kgU and up to 20 MW·d/kgU for PHWRs. In addition, the CRP addressed the performance of codes used for transient analysis such as RIA and LOCA at extended burnup. 2.2. REQUIREMENTS FOR FUMEX-III

This CRP used well qualified data from both research and power reactor programmes from several countries against which the participants were asked to compare their code predictions. Much of these data were already available in the IFPE Database, some having been added during the FUMEX-2 project. Further additions to the database were made during the project. This CRP promoted interaction and discussions amongst fuel modellers which has resulted in a better understanding of physical processes and phenomena, and has allowed improvements to be made in both codes and their models.

11

2.3. PARTICIPANTS

A list of the participating organizations and the codes they used is given in Table 6. TABLE 6. PARTICIPATING ORGANIZATIONS OF THE FUMEX-III CRP

Country Organization Code Argentina CNEA BACO Argentina CNEA DIONISIO Belgium Tractabel, GDF-SUEZ FRAPCON FRAPTRAN Brazil CDTN and ETN FRAPCON FRAPTRAN Brazil IPEN-CNEN/SP FRAPCON FRAPTRAN Bulgaria INRNE TRANSURANUS (WWER) Bulgaria INRNE FEMAXI Canada AECL ELESTRES China CIAE METEOR FTRANAC China NPIC COPERNIC China CNPTRI COPERNIC Czech Republic NRI TRANSURANUS FEMAXI-6

FRAPCON FRAPTRAN Finland VTT ENIGMA Germany ITU TRANSURANUS Germany/ France Areva GALILEO Hungary AEKI FUROM

TRANSURANUS FRAPTRAN India BARC FAIR Italy ENEA

Politecnico di Milano TRANSURANUS

Italy University of Pisa TRANSURANUS FUELSIM

Japan JAEA FEMAXI-7 RANNS

Japan JNES FEMAXI-JNES Japan NFI THERMEX-N

FORTE Republic of Korea KAERI INFRA Romania INR ROFEM

CAREB Russian Federation Bochvar Institute START-3 Russian Federation SRC TRINITI RTOP Switzerland PSI FALCON United Kingdom BE ENIGMA 5.14 United Kingdom Nexia ENIGMA-B Ukraine Kharkhov Institute PAD FEMAXI

TRANSURANUS United States of America INL BISON United States of America PNNL FRAPCON

FRAPTRAN Norway Halden Reactor Project Data support and provision France OECD/NEA Coordinator of the IFPE

The participants had a wide range of reasons for joining the FUMEX-III CRP and some were more active

than others. Several participating groups were working in collaboration with each other, sharing development and using codes. In Brazil, a group of six organizations collaborated to use the FRAPCON/FRAPTRAN codes and there is another group of developers of these codes based in Belgium and the United States of America. The TRANSURANUS user group includes the code owners, the European Commission, and developers in Bulgaria working on the WWER version of the code and three organizations in Italy carrying out basic research, there are other users in Hungary and the Ukraine. The Japanese code FEMAXI is still under development in Japan and standard versions are in use in many countries.

12

Many of the participants used the FUMEX-III CRP to provide a validation database for their codes or to help in development. Some were using commercial codes and used the CRP to help develop an understanding of the code and to train users. The wide range of participants and their needs contributed to informative discussions and widespread cooperation between the participants. 2.4. CASES

At the consultants meeting in April 2008 a review was carried out of the remaining issues in the field of fuel modelling and of the availability of experimental data that might help fuel modellers to address the problems. The consultants proposed that a matrix approach would be useful, with a range of topics to be studied across the main reactor systems that exist; PHWR, LWR (BWR and PWR) and WWER. It was recognised that there were different concerns for the different reactor systems and it was suggested that there should be a group of priority cases, but that it would not be expected that all participants would carry out all of these cases, but that they would choose from amongst these cases according to their needs. It was expected that participants would consider cases from other systems to test the limits of their codes applicability and validation.

A matrix of the proposed issues and the potential cases to address them is given in Table 7. A list of the

priority cases agreed at the first research coordination meeting is given in Table 8. A final list of cases that were made available to the project is given in Table 9, which provides a brief

overview of the issue that each case was intended to address. Fuller details are provided in Section 3. Participants were asked to provide output data for these cases in a manner that allowed easy inter-comparison, though it was recognised that the output of the codes would be different and that some codes did not have the capability to model all the topics.

The cases included tests from the SUPER-RAMP and INTER-RAMP series where around 50% of the

rods had failed during a power ramp. Participants were asked to use these tests to determine a pellet–cladding interaction (PCI) failure threshold rather than to simply model the behaviour of the fuel during the test in terms of fission gas release or temperature during the ramps.

2.5. ADDITIONAL INFORMATION

The discussions at the first RCM are recorded in Annex I. The information includes the original case list and the revised version agreed at the meeting.

The discussions held during the second and third RCMs were important for the exchange of ideas and

discussion of the issues. The records of these meetings are given in Annex II and Annex III respectively. Participants’ reports to the project are appended in the CD attached to this publication.

13

TAB

LE 7

. PR

OPO

SED

MA

TRIX

OF

CA

SES

FOR

FU

MEX

-III

Pl

ant t

ype

MO

X

Mec

hani

cal i

nter

actio

n Se

vere

tran

sien

ts

FGR

; Tem

pera

ture

FU

MEX

-II

PCM

I PC

I LO

CA

R

IA

Load

follo

w

trans

ient

s Tr

ansi

ents

G

ad/N

b N

orm

al

oper

atio

n FG

R

CA

ND

U

CN

EA 5

rods

1

undo

ped

BU

15

2 lo

w B

U ro

ds

from

Ris

ø-2

GE-

m a

nd II

-3

for p

aram

etric

stud

y - f

ind

failu

re

thre

shol

d IN

TER

RA

MP

10G

20

G

SUPE

RR

AM

P 3

5G 50

% o

f rod

s fai

led

for e

ach

burn

up

FIO

131

CA

ND

U

(MT)

IRD

MR

FIO

-118

; FI

O-1

19 (7

rods

in

tota

l 2+5

)

Pi

test

i RO

89

RO

51

AEC

L N

RU

(one

ro

d fr

om e

ach

ring)

LWR

IF

A 6

29.1

IF

A 5

14

IFA

565

IF

A 5

91

Ris

ø-2G

ei G

eh

(GEi

firs

t)

Ris

ø-3

AN

8 G

E7

IRD

MR

FIO

-118

FIO

-119

O

SIR

IS -

2 ro

ds

NEA

B

ench

mar

k IF

A

650.

1/2

MT4

, MT6

A

NSR

R

3 B

WR

ca

ses

FK 1

,2,3

IFA

519

.8/9

Ja

pane

se

BW

R lo

ad

follo

w

IFA

535

5 an

d 6

(?

FUM

EX-I

/6)

Ris

ø 3

rod

II5

GA

IN ro

d 11

7%

G

d

GA

IN 4

rods

, 3%

and

7%

Gd

(one

tran

sien

t te

sted

)

US

16×1

6 P

WR

(d

o 7

rods

) (G

rain

si

ze, l

iner

, re

fere

nce)

A

REV

A id

ealiz

ed

case

G

inna

AN

2 A

REV

A c

ase

Vita

nza

thre

shol

d

WW

ER

MIR

Ram

p 9

rods

U

S 16

×16

PW

R

annu

lar p

elle

ts

Lovi

isa

Gin

na h

ollo

w

pelle

ts

Mat

eria

ls

IF

A 5

65 Z

r2

Zr4

cree

p

AEK

I dat

a tu

be

burs

t E11

0 (+

Zr4

stan

dard

s)

14

TAB

LE 8

. PR

IOR

ITY

CA

SES

IDEN

TIFI

ED A

T TH

E EN

D O

F TH

E FI

RST

PR

OJE

CT

MEE

TIN

G

Pla

nt

type

M

OX

M

echa

nica

l int

erac

tion

Seve

re

trans

ient

s FG

R; T

empe

ratu

re, e

tc.

PCM

I PC

I LO

CA

R

IA

Load

follo

w

trans

ient

s Tr

ansi

ents

G

ad/N

b N

orm

al o

pera

tion

FGR

CA

ND

U

CN

EA 5

rods

1

undo

ped

BU

152

low

BU

from

R

isø-

2 G

E-m

and

II

-3

IRD

MR

FIO

-118

; FI

O-1

19

For p

aram

etric

stud

y –

find

failu

re

thre

shol

d IN

TER

RA

MP

10G

20

G

SUPE

RR

AM

P PK

6 an

d PW

3 35

G

FIO

13

1

IR

DM

R F

IO-1

18;

FIO

-119

(7 ro

ds in

to

tal 2

+5)

Pi

test

i RO

89 R

O51

A

ECL

NR

U (o

ne ro

d fr

om

each

ring

)

LWR

IF

A 6

29.1

PR

IMO

rod

(BD

8)

Ris

ø3 G

E7

OSI

RIS

-2 ro

d H

O9

IFA

65

0.2

FK1

and

FK2

IFA

519

.8/9

R

ods D

C a

nd

DK

IFA

535

5 ro

d 9

Ris

ø 3

rod

II5

52G

G

AIN

Gd

701

and

301

US

16×1

6 P

WR

TSQ

002

TSQ

022

AR

EVA

idea

lized

cas

e

WW

ER

MIR

Ram

p ro

ds 4

1 48

50

51

US

16×1

6 P

WR

TSQ

022

Gin

na X

O3

segm

ente

d st

anda

rd a

nd a

nnul

ar p

elle

ts,

nom

gap

, sta

ndar

d cl

ad

15

TABLE 9. LIST OF CASES AVAILABLE FOR CALCULATION AT THE END OF THE PROJECT

Condition Case Comment

System of particular interest

CA

ND

U

LWR

WW

ER

Normal operation

US PWR 16×16 Solid and hollow pellets, burnup ~55 MW·d/kgU

x x

Areva idealized case Commercial fuel history to 80 MW·d/kgU with indicated FGR

x x

FUMEX-II cases Range of cases for newcomers to use x Risø-3 AN2 AN3 Radial distribution of fission products and

FGR-EOL, Bu ~40 MW·d/kgU x

R. E. Ginna Solid and hollow pellets, burnup ~55 MW·d/kgU

x x

RO51 RO89 CANDU variants, burnup to 160 MW·h/kgU

x

AECL BDL-406 Natural UO2 fuel irradiated at low linear powers (30-40 kW/m) to burnups above 450 MW·h/kgU in the NRU reactor

x

Gadolinia Gain rods 301 and 701 3% and 7% gadolinia doped rods irradiated

to 50 MW·d/kgU x x

Transients IFA 535.5/6 Experimental rod with low internal

pressure, burnup ~48 MW·d/kgU. x

Risø-3 II5 Power ramp to 40 kW/m at 50 MW·d/kgU x x GAIN rod 701 7% gadolinia rod ramp tested x IRDMR FIO 118/119 Diameter measurements of ramped

CANDU type fuel x

MIR WWER fuel ramped at up to 40 kW/m at burnup of up to 60 MW·d/kgU

x

Load follow CANDU JC NR CANDU fuel bundles (37 elements) tested

at the NRU reactor x

IFA 519.8/9 DC DK Very high burnup (100 MW·d/kgU) Halden test, some data problems

x x

PCI criterion INTER-RAMP

SUPER-RAMP PWR and BWR ramp tests; 50% of rods failed during ramps, PCI failure predictions required

x x x

PCMI Risø-2 GE-m Risø-3 II-3 Power ramp test on low burnup rods

(~17 MW·d/kgU) x x

Risø-3 GE7 Ramp test to 35 kW/m at 42 MW·d/kgU. Good data clad ridging.

x x

OSIRIS H09 (J12) Diameter measurements before and after ramping

x

CNEA BU15 MOX fuel ramp x

16

TABLE 9 LIST OF CASES AVAILABLE FOR CALCULATION AT THE END OF THE PROJECT (cont)

Condition Case Comment

System of particular interest

CA

ND

U

LWR

WW

ER

RIA BWR RIA FK1 FK2 Two pulse irradiation tests, intact at 544 and 293 J/g (130 and 70 Cal/g)

x

LOCA IFA 650.1-2 Scoping LOCA studies at Halden x x FIO 131 LOCA test of CANDU type fuel x MOXa IFA 629 Refabricated MOX fuel, irradiated in

Halden x

PRIMO MIMAS MOX fuel irradiated to burnup of 30.1 MW·d/kgHM and then ramp tested

x

a MOX is an acronym for mixed oxide fuel and refers to plutonium bearing oxide fuel.

3. DESCRIPTION OF CASES

The cases made available to the participants of the FUMEX-III CRP are detailed below. 3.1. NORMAL OPERATION CASES 3.1.1. US PWR 16×16

The US PWR 16×16 lead test assembly (LTA) extended burnup demonstration programme was conducted during the 1980s with the objective of demonstrating improved nuclear fuel utilization through more efficient fuel management and increased discharge burnup. The use of the 16×16 LTAs with Zr-4 cladding in this programme demonstrated the capability to achieve peak fuel rod average burnups of ~60 MW·d/kgU. Both poolside (non-destructive) and hot cell (destructive) post irradiation examinations (PIE) of selected rods from the two LTAs were conducted. These examinations included rods irradiated for 3 and 5 cycles. Two of the 16×16 LTAs were irradiated in a commercial US PWR.

The standard fuel rod design consists of enriched UO2, solid cylindrical pellets. In addition to the standard

design fuel rod, three additional design concepts were included in a limited number of rods in the two LTAs. These were:

An annular fuel pellet design; Large grain size pellets (35 µm as opposed to the nominal 7 to 12 µm standard pellet design); Cladding with graphite coating (~ 8 µm thickness) on the interior surface.

Two priority cases, TSQ002 (standard fuel rod) and TSQ022 (fuel rod with hollow pellets), were chosen for the FUMEX-III calculations to help modellers distinguish between solid and hollow fuel pellets. Hollow pellets are used in WWER plant and solid pellets in Western PWRs. The basic characteristics of fuel rods and selected EOL data are summarized in Table 10.

17

TABLE 10. BOL AND EOL CHARACTERISTIC DATA OF FUEL RODS TSQ002 AND TSQ022 Fuel rod No. TSQ0002 TSQ022 Inner fuel diameter, mm 0 2.34 Outer fuel diameter, mm 8.255 8.255 Pellet length, mm 9.91 9.91 Chamfer and dish Yes Yes Fuel stack length, mm 3810 3810 Enrichment, %235U 3.48 3.48 Cladding inner diameter, mm 8.43 8.43 Cladding outer diameter, mm 9.7 9.7 Fill gas pressure, MPa 2.62 (He) 2.62 (He) Initial free volume, mL 25.42 37.22 Rod average burnup, MW·d/kgU 53.24 58.12 EOL free volume, mL 17.8 31 Δ free volume, mL -7.62 -6.22 Δ gas volume (EOL - BOL), % 5.7 2.6

3.1.2. AREVA idealized case

This case is an idealized commercial operation, which is based on measurements for three rods operated for 3, 4 and 7 cycles in a commercial French PWR reactor. Those rods experienced very similar power histories, i.e. the 3 and 4 cycle rods match very closely the first 3 and 4 cycles, respectively, from the 7 cycle rod, allowing three FGR 'measurements' for a single power history. The maximum fuel rod burnup is about 81.5 MW·d/kg(HM) with an FGR of about 9% (see Table 11). The given FGR uncertainties allow for measurement, fabrication and irradiation uncertainties. Figuress 1- and 2 represent the power history of the idealized case. TABLE 11. EXPECTED FGR VALUES

End of cycle Insertion time Burnup Expected FGR value (d) (MW·d/kg(HM)) (%) 3 916.4 36.6 0.5 +0.5/-0.2 4 1239.1 49.7 1.9 +1.0/-0.7 7 2141.9 81.5 9.0 +2.5/-2.0

18

FIG. 1. Power history for an idealized rod irradiated to over 80 MW·d/kgU.

FIG. 2. Detailed power history for the idealized AREVA case.

0

50

100

150

200

250

0 10 20 30 40 50 60 70 80 90

Ave

rage

rod

LHG

R [W

/cm

]

Burnup [MWd/kg(HM)]

19

3.1.3. Risø-3 AN2, AN3

The Risø National Laboratory in Denmark have carried out three irradiation programmes of slow ramp and hold tests, so called 'bump tests' to investigate fission gas release and fuel microstructural changes. The third and final project, which took place between 1986 and 1990, bump tested fuel re-instrumented with both pressure transducers and fuel centreline thermocouples. The innovative technique employed for re-fabrication involved freezing the fuel rod to hold the fuel fragments in position before cutting and drilling away the centre part of the solid pellets to accommodate a new thermocouple.

The fuel used in the test AN2 was ramped in the unopened state. During base irradiation the power profile

was insignificant. However, during the bump test, there was a significant axial power profile. The fill gas and dimensions remained as fabricated and the bump testing was carried out 6-9 November 1987.

AN3 was refabricated from segment CB8 with pressure transducer and fuel centreline thermocouple. The

fill gas was 15 bar helium and bump testing was carried out 8-11 January 1988. The rod AN3 had been a priority case in FUMEX-II and two participants attempted this case during

FUMEX-III and only one modelled the AN2 rod. 3.1.4. R. E. Ginna rodlets

The objective of this programme was to develop a fuel design with increased margin to PCI failure threshold and with an increased potential for higher burnup. The means by which this objective was to be attained was with annular pellets and zirconium barrier cladding. This feature makes this case especially suitable for WWER applications.

The annular pellets used have a void volume of approximately 10% higher than comparable solid pellets.

Annular pellets are believed to lead to (a) less hour-glassing, which would result in less clad ridging; (b) lower fission gas release by virtue of lower maximum pellet temperatures; and (c) lower end-of-life internal pressure by virtue of larger available volume for the fission gasses released, leading to significantly lower end-of-life gas pressure for the same fuel rod energy generation. Barrier cladding consists of Zircaloy-4 tubes with an integral inner layer of unalloyed zirconium comprising approximately 10% of the total wall thickness. The overall cladding dimensions are the same as the standard cladding. The zirconium barrier cladding is a design developed to provide resistance to PCI, and improve the capability of reaching a higher burnup while the fuel is subjected to variations in local linear heat generation rates resulting from control rod movements, overall core power manoeuvring, and fuel assembly shuffling.

Four Siemens Power Corporation 14×14 lead fuel assemblies were inserted into the R. E. Ginna reactor,

in May 1985, under a cooperative programme jointly sponsored by Empire State Electric Energy Research Corporation, Rochester Gas and Electric Corporation and Siemens Power Corporation. The programme included segmented rods containing rodlets with different combinations of cladding, pellet types, and pellet-to-clad gaps. The test segments were approximately 66 cm long and located symmetrically at the centre of the fuel rod. Power and burnup accumulated in the two central rodlets were nearly the same for all segmented rods and nearly uniform along the length. The segmented rods achieved a burnup of 52 MW·d/kgU for solid pellet rodlets and 57 MW·d/kgU for annular pellets. The power histories of the two rodlets are shown in Fig. 3.

20

FIG. 3. Power histories of rodlets 2 and 4 from R. E. Ginna irradiation.

For FUMEX-III the interest is for the comparison between the solid and hollow pelleted fuel which has

relevance for WWER and PWR comparisons. Two rodlets (rodlet 2 and rodlet 4) were chosen as priority cases. PIE data includes dimensional measurements and fission gas release from pin puncturing.

3.1.5. Experimental fuel elements (EFE) 51 and 89

These experiments were performed in the Nuclear Research Institute Pitesti, Romania with the objective of gaining information about the behaviour of CANDU fuel elements within the limits of the design parameters. Two rods with different plenum sizes, instrumented with pressure transducers, were tested in the C2 irradiation device designed for the TRIGA reactor (14MW). The pellets of the EFE 89 element had a density of 10.54-10.62 Mg/m3, a diametral gap of 0.084 mm and an enrichment of 3.92 wt%. For the EFE 51 element, the values were 10.70-10.75 Mg/m3, 0.100-0.130 mm and 7.04 wt%, respectively. Figure 4 shows the irradiation histories of these two CANDU elements, EFE 51 and EFE 89.

FIG. 4. Real and simplified power histories of experiments No.89 and No.51.

0

5

10

15

20

25

30

35

0 5000 10000 15000 20000 25000 30000 35000

Rod

aver

age

pow

er (k

W/m

)

Irradiation time (h)

Power history for Ginna rodlets

Rod 2

Rod 4

0 20 40 60 80 100 120 1400

100

200

300

400

500

600

EFE No 89

linea

r pow

er (W

/cm

)

time (days)

simplified history real history

0 20 40 60 80 100 120 140 160 1800

100

200

300

400

500

600

EFE No 51

linea

r pow

er (W

/cm

)

time (days)

simplified history real history

21

The linear power was in the range 55030 W/cm during most of the irradiation for both experiments with the sole exception of the abrupt fluctuations at EOL in EFE 51. The final burnup was 137.6 MW·h/kgU for EFE 89 and 159.3 MW·h/kgU for EFE 51. The coolant pressure was 10.7 MPa for both experiments; the pH was in the range 9.5-10.5 for EFE 89 and 6.2-6.8 for EFE 51. 3.1.6. AECL BDL-406

Bundles XY, AAH and GF were CANDU fuel bundles irradiated in the NRU (National Research Universal) reactor at the Chalk River Laboratories as part of an experimental programme on performance of natural UO2 fuel irradiated at low linear powers (30-40 kW/m) to burnups above 450 MW·h/kgU (18.8 MW·d/kgU). The bundles were of the Bruce Nuclear Generating Station (NGS)-A “first charge” design and were similar to the Bruce bundles currently in use at Bruce NGS-A, except that they contained outer-element gas plenums to accommodate fission-gas release. They were successfully irradiated in the NRU reactor between 1975 and 1990 to outer element burnups that ranged from 570 to 900 MW·h/kgU (23.8 to 37.5 MW·d/kgU) [12].

Post-irradiation examination of the fuel elements from the bundles included visual examination, sheath

profilometry measurements, gas-puncture analysis, metallography, fuel grain size measurements, a graphite coating survey and chemical burnup measurements.

Typically, natural UO2 fuel is discharged from CANDU reactors at a bundle-average burnup of about 200

MW·h/kgU (8.3 MW·d/kgU). The performance of CANDU fuel at extended burnups is of interest to operators who wish to understand the operational limits of the fuel and to fuel designers who wish to understand the parameters that affect fuel performance.

This case was made available late in the project and no participant had undertaken this case by the end of

the project. 3.2. GADOLINIA DOPED FUEL 3.2.1. GAIN project

The two rods provided for the FUMEX-III CRP came from the BelgoNucleaire GAIN Project. There were four rods in the main experiment, but the interest for the participants was to compare the behaviour of rods with two different levels of gadolinia doping

The full dataset is for four Gd2O3 doped UO2 rods irradiated in the BelgoNucleaire GAIN Project. The

rods are identified as rods GD0301, GD0302, GD0701 and GD0702. GD0301 and GD0302 contain pellets doped with 3wt% Gd2O3 and rods GD0701 and GD0702 contain 7wt% doped pellets.

All rods were irradiated in BR3 to a final burnup of approaching 40 MW·d/kgU. In addition, rod GD701

underwent 2 overpower transients in BR2. The sequence of irradiation is given below: (1) All rods irradiated in BR3 cycle 4C; (2) GD0701 transient tested in BR2; (3) All rods irradiated in BR3 cycle 4D1; (4) GD0701 transient tested in BR2; (5) All rods irradiated in BR3 cycle 4D2.

Table 12 shows the maximum and rod average burnup (MW·d/kgM) of the GAIN rods at end-of-cycles. Non-destructive PIE was performed at the end of each irradiation measuring: creep-down, rod and fuel length changes. After final discharge, destructive PIE was performed either at PSI in Switzerland or at Windscale in the UK.

22

TABLE 12. MAXIMUM AND ROD AVERAGE BURNUP (MW·D/KGM) OF THE GAIN RODS AT END-OF-CYCLES:

Rod BR3/4C BR3/4D1 BR3/4D2 Ave. Max. Ave. Max. Ave. Max.

GD0301 13.7 18.3 26.2 34.2 38.8 49.7 GD0302 13.9 18.5 26.2 34.2 37.9 48.6 GD0701 13.0 17.7 25.9 33.9 38.9 49.9 GD0702 12.7 17.2 25.7 33.6 38.9 49.8

Two rods, GD301 and GD701, were chosen as priority cases for FUMEX-III.

3.3. TRANSIENTS 3.3.1. IFA 535.5 and IFA 535.6: Effect of pre-pressurization on fission gas release

In order to investigate the effect of fuel rod pressurization on fission gas release, two power ramp tests were performed with two pairs of re-instrumented high burnup fuel rods in the Halden Boiling Water Reactor. The fuel rods identified with the numbers 809, 810, 811 and 812 were base irradiated in the upper cluster of the two-cluster test instrumented fuel assembly (IFA) from May 1973 to June 1985 up to the average burnup level of 43.4 MW·d/kgUO2.

The neutron flux was monitored with vanadium neutron detectors located at different axial and radial

positions in the assembly, but other parameters were not measured in this IFA. After the base irradiation the fuel rods were equipped with pressure transducers and clad elongation sensors in order to monitor the rod internal gas pressure and the mechanical behaviour during a slow and a fast power ramp in IFA-535.5 and IFA-535.6 respectively.

The re-instrumentation technique permits the attachment of pressure transducers and pressurization of the

rods without losing the fission gas already accumulated in the free volumes, since a completely sealed instrumented head containing a micro drill is welded on to the fuel rod. After welding, the micro drill is driven by a pair of external rotating magnets in order to merge the gas plenum and the end plug. Prior to application, the instrumented heads for rods 810 and 812 were pressurized to 50 bar with He in order to provide higher internal pressure after drilling. On the other hand, vacuum was created in the end plug of rods 809 and 811.

The re-instrumented rods were ramp tested in IFA-535.5 (rods 809 and 810) from November 1985 to

February 1986, and in IFA-535.6 (rods 811 and 812) from December 1986 to May 1987. Both rigs were loaded into separated high pressure loops simulating PWR conditions thus minimizing the risk of cladding failure due to large pressure difference. The on-line rod pressure, and consequently FGR monitoring, made the observation of particular phenomena related to power variation possible. Beyond the in pile pressure monitoring, PIE was performed for rods 809 and 810 in support of calculated FGR. Although the fuel centre line temperature was not measured during the tests, this essential parameter for code validation can be supported on the basis of IFA-533.2 data also presented in the IFPE. IFA-533.2 contained the remaining two rods of IFA-535.4 upper cluster, i.e. rods 807 and 808, re-instrumented with fuel thermocouples. The irradiation conditions of the rig were very similar to those of the present cases. Hence the tests complement each other with respect to fuel temperature and FGR. The power history is shown in Fig.5.

23

FIG. 5. Power history of IFA-535.5.

The low pressure, slow ramp case, IFA-535.5 rod 809, was the priority case chosen from this experiment.

3.3.2. Risø-3 bump test II5

The Risø National Laboratory in Denmark have carried out three irradiation programmes of slow ramp and hold tests, so called 'bump tests' to investigate fission gas release and fuel microstructural changes. The third and final project, which took place between 1986 and 1990, bump tested fuel re-instrumented with both pressure transducers and fuel centreline thermocouples. The innovative technique employed for re-fabrication involved freezing the fuel rod to hold the fuel fragments in position before cutting and drilling away the centre part of the solid pellets to accommodate a new thermocouple.

The fuel used in the test II5 was from IFA-161 irradiated in the Halden BWR to 46 MW·d/kgUO2. The

base and ramp power histories for the test segment are shown in Fig. 6. The ramp test comprised a series of steps at increasing power levels and a hold at the final power of 40 kW/m for around 26 hours.

(a) (b)

FIG. 6. Power history of the Risø-3 test II-5 showing (a) the base irradiation and (b) the bump test at Risø.

0

5

10

15

20

25

30

35

40

45

50

0 10 000 20 000 30 000 40 000 50 000 60 000 70 000

Rod

aver

age

pow

er (k

W/m

)

Irradiation time (h)

Power history of IFA-535.5

IFA 535.5

0

10

20

30

40

50

60

0 10000 20000 30000 40000 50000

Rod

aver

age

ratin

g (k

W/n

m)

Irradiation time (h)

Risø test II5: Rating history

Rod averagerating kW/m

0

10

20

30

40

50

60

48550 48570 48590 48610 48630 48650

Rod

aver

age

ratin

g (k

W/n

m)

Irradiation time (h)

Risø test II5: Rating