IAEA-TECDOC-1132 XA0053626 ,

Control assembly materialsfor water reactors: Experience, performance and perspectives

Proceedings Technicala of Committee meeting held in Vienna, 12-15 October 1998

INTERNATIONAL ATOMIC ENERGY AGENCY /A

February 2000 31-08 J> The originating Section of this publication in the IAEA was: Nuclear Fuel Cycle and Materials Section International Atomic Energy Agency Wagramer Strasse 5 P.O. Box 100 A-1400 Vienna, Austria

IAEe Th A doe t normallsno y maintain stock f reportso thin si s series. However, electronic copies of these reports can be obtained from:

INIS Clearinghouse International Atomic Energy Agency Wagramer Strasse 5 0 10 x P.OBo . A-1400 Vienna, Austria

Telephone: (43) 1 2600-22880 or 22866 Fax: (43 2600-2988)1 2 E-mail: CHOUSE® IAEA.ORG Web site: http://www.iaea.org/programmes/inis/inis.htm

Orders shoul accompaniee db prepaymeny db Austria0 10 f o t n Schilling fore a th f m o n s i chequ r credieo t card (MasterCard, VISA).

CONTROL ASSEMBLY MATERIALS FOR WATER REACTORS: EXPERIENCE, PERFORMANCE AND PERSPECTIVES IAEA, VIENNA, 2000 IAEA-TECDOC-1132 ISSN 1011-4289 © IAEA, 2000 Printed by the IAEA in Austria February 2000 FOREWORD

The safe, reliable and economic operation of water cooled nuclear power reactors depends to a large extent upo reliable nth e operatio f controno l assemblie regulatioe th r s fo shutdowd nan e th f no reactors. These consist of neutron absorbing materials clad in stainless steel or zirconium based alloys, guide tubes and guide cards, and other structural components. Current designs have worked extremely well in normal conditions, but less than ideal behaviour limits the lifetimes of control materials, imposin n economia g c penalty which a stronact s a s g incentiv o product e e improved materials and designs that are more reliable.

Neutron absorbing materials currently in use include the ceramic boron carbide, the high melting point metal hafnium and the low melting point complex alloy Ag-In-Cd. Other promising neutron absorbing materials, such as dysprosium titanate, are being evaluated in the Russian Federation. These control materials exhibit widely differing mechanical, physical and chemical properties, which must be understood in order to be able to predict the behaviour of control rod assemblies.

Identification of existing failure mechanisms, end of life criteria and the implications of the gradual introduction of extended burnup, mixed oxide (MOX) fuels and more complex fuel cycles constitute firse sth t stesearca n p i improver hfo d material designsd san .

earle Inth y par f thio t s decade s recognizewa t i Internationae , th y db l Working Grou Fuen po l Performanc d Technologan e y (IWGFPT) that international conferences, symposi d publishean a d e reviewmaterialth n o s s science aspect f controo s l assemblie r betweenfa d san .werw fe e Consequently, the IWGFPT recommended that the IAEA should rectify this situation with a series of Technical Committee meetings (TCMs) devoted entirel materiale th o yt s aspect f reactoso r control assemblies e interveninth s n e firsheli n 199Th wa d .ti an 3g five years considerable progress ha s been made bringinn I . g together expert materiale th n si s scienc materiald ean s engineerin f controgo l assembly materials, the 1993 meeting and the current one are helping to fill a gap in the information exchange opportunities in this important branch of nuclear research and development.

This second TCM, entitled Control Assembly Materials for Water Reactors: Experience, Performance and Perspectives, was held in Vienna from 12 to 15 October 1998. Thirty-one participants from fourteen different countries attended the meeting and nineteen papers were presented and are reproduced in these proceedings together with a summary of the meeting.

The IAEA wishee participantth f thano t so l kal r thei fo s r contribution thio t s s publication, which summarizes experience to date with control materials and ongoing research in the field in the participating countries. I.G. Ritchi M.Jd ean . Crijns ,Divisioe botth f ho f Nucleano r Fuel Cycld ean Waste Technology, wer IAEe eth A officers responsibl organizatioe th r efo e meetin e th th f nd o gan compilation of this publication, respectively. EDITORIAL NOTE

This publication beenhas prepared from originalthe material submittedas authors.the viewsby The 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 delimitation ofthe or 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 partthe IAEA. ofon the CONTENTS

Summary ...... 1 LWR control assembly designs historica:A l perspective...... ? M. W. Kennard, J.E. Harbottle controR degradatiod PW ro l observes na post-irradiation di n examination Studsvin si k ...... 33 T. Jonsson, T. Andersson, L. Bjornkvist Status of control assembly materials in Indian water reactors...... 41 V.G. Date, P.O. Kulkarni Calculation modelling of the RCCA movement through bowed FA guide tubes...... 57 Razoumovsky,V. D. Yu.I. Lihkachev, V.M. Troyanov Material operatin controR g behaviouBW l B rodAB s f ...... 6ro 5 B. Rebensdorff, G. Bart Experience of CR and RCCA operation in Ukrainian WWER-1000: Aspects of reliability, safety and economic efficiency...... 77 A. Afanasyev Dysprosiu hafniud man m base absorber advancer sfo d WWER control rods...... 91 V.D. Risovaniy, A.V. Zaharov, E.P. Klochkov, E.E. Varlashova, D.N. Suslov, A.B. Ponomarenko, ScheglovV. A. Behaviou differenf ro t boron rich solid promisins sa g absorber PWR...... 10r sfo 3 D. Simeone, Deschanels,X. Cheminant,P. HerterP. Reinforcement against crack propagatio absorberR PW f developmenno y sb t of boron-carbon-hafnium composites...... 107 B. Provot, P. Herter NSC KIPT's experienc production ei absorbef no r materials, composited san product contror sfo l mechanism variouf o s s nuclear reactor types ...... 113 N.P. Odeychuk, V.F. Zelemky, V.A. Gurin, Yu.F. Konotop Damage analysi ceramif so c boron absorber material boilinn si g water reactors and initial model for an optimum control rod management...... 121 W. Schulz Fabricatio metallurgicad nan l propertie hafniuf so m alloy contror sfo l rods ...... 153 J.L. Bechade, P. Parmentier Irradiation behaviour of boron carbide in WWER-1000 ...... 167 Zakharov,V. A. V.D. Risovaniy, Fridmann,S.R V.B. Ponomarenko, A.O. Poslavskiy WWER-1000 control rod materials: Experience and development...... 175 V. Ponomarenko, A. Poslavskiy, A. Scheglov, A. Zaljetniy, V. Risovaniy FRAMATOME contro absorbed ro l r materials: Experienc development...... 19d ean 1 M. Monchanin, X. Thibault Measure of 11 -boron, 10-boron and 7-lithium concentrations in control rods using the nuclear microprobe technique ...... 203 D. Simeone, X. Deschanels, P. Herter The rol claddinf eo g materia performancr fo l controR LW lf eo assemblie s ...... 215 P. Dewes, A. Roppelt BWR control blade replacement strategies...... 229 M.W. Kennard,J.E. Harbottle High temperature study of the control rod behaviour under accident conditions...... 245 V. Troyanov, A. Pomeschikov, V. Sougonjaev, V. Ponomarenko, A. Scheglov

List of participants ...... 259 SUMMARY

1. INTRODUCTION

For the purposes of this meeting, "control assembly materials" included all of the solid materials used in the construction of control assemblies for the regulation or shutdown of water cooled, power reactors, i.e. neutron absorber(s), cladding, guide tube guidd an s e cards othed an , r structural components of the assembly. Liquid neutron absorbers and poisons were not considered and, wit exceptioe h th pape e on f rn o comparin behavioue gth borof ro n carbide (B4C dysprosiud )an m titanate (Dy203'TiO2) in severe accident conditions, the deliberations dealt with the behaviour of control materials in normal operating and shutdown conditions.

International conferences, symposia and published reviews on the materials science aspects of control assemblies have been relatively few and far between. Although the IAEA has held many meeting e safetth n yo s aspect f controo s l rodd controan s l systems n boti , h normal operating conditions and severe accident conditions, to our knowledge this is only the second meeting to focus materiale onth s aspect reactof so r control assembly materials firse Th .t Technical Committee meeting (TCM) entirel thin yo s intervenin e topi th hels n 199i cn dwa i d 3an g five years considerable progress has been made bringinn I . g together expert e materialth n i s s scienc materiald ean s engineerinf go control assembly materials, this meeting predecessos it , 199n ri theid 3an r published proceedinge sar helpin e informatiofilimportano n gt a th l n i p ga t n exchange opportunitie thin i s s branc f nucleaho r researc developmentd han .

The meetin s dividewa g d into three sessions: Sessio , Pas1 n t Experienc d Currenan e t Performanc f Controo e l Assembly Materials", chaire . DeweP y db f Siemen o sGermanyn i G A s ; Session 2, Absorber Materials Currently Used and Under Development, chaired by V.G. Date of Bhabha Atomic Research Centr Indian ei Sessiod an ; , Claddinn3 g Materials, Life-Limiting Criteria and Future Needs, chaired by G. Bart of the Paul Scherrer Institute in Switzerland.

The nineteen papers presented covered the control assembly materials currently used in PWRs, BWRs, WWER RBMKsd san wels a , improves a l d materials under investigatio their nfo r replacement r advanceofo r d water reactor designs e technicaTh . l discussions focusee performancth n o d f o e absorber materials, cladding, guide-tubes and guide-cards in reactor, keeping in mind issues affecting and limiting the working lifetimes of control assemblies, and hence their economic viability.

Past and current designs of control assemblies together with the stringent criteria followed by the utilities for their safe operation, inspection and replacement, has led to an outstanding safety record. However, mechanical mechanisms (fretting and wear), corrosion mechanisms (oxidation and hydriding), radiation damage mechanisms (enhanced creep and growth), and mechanisms such as irradiation assisted stress corrosion cracking (IASCC), involving synergism between two or more fundamental mechanisms l detracal , t fro e ideamth l behaviou f controo r l assembly materials. More importantly, this less than ideal behaviour limit e lifetimeth s f controo s l materials, imposinn a g economic penalty which acts as a strong incentive to produce improved materials and designs that are more reliable and approach closer to the ideal lifetime of a control assembly, which should depend solely on the neutronics of the chosen absorber. At the time of the last TCM, improved understanding of the damage mechanisms mentioned above was leading to improved behaviour of control assemblies through design changes, wear-resistant coating d surfacan s e treatments d bettean , r control over microstructures and compositions of the materials at the fabrication stage. One of the major goals of thif i se progresse o t s bees sha wa n maintaineM thisTC face f th increasin eo n di g demande th n so combined fuel/control assemblies, such as higher burnup, MOX fuel burning and load following.

Experience with control material waten i s r reactor s spreai s d over many different typef o s reactors in many different countries. Similarly, research, development and design of the different type f controso l assemblie carries severay si b t dou l different vendors. Presentation meetine th t sa g dealt wit currentlhe almosth f o l yal t used control material wate e mosn si th f ro t cooled, power reactor types operate t presentda , wit exceptioe hth f CANDUno additionn I . , presentatione somth f eo d san discussions were devote o alternativedt , advanced absorber material developmentd an s s aimet a d increasin lifetimee gth controf so l assemblies.

Following the formal presentations of the papers and some preliminary discussions, the meeting split into three groups to facilitate discussions and prepare a summary, recommendations for further work in the field and conclusions of the meeting. The first group considered PWRs, the second group considered BWRs and the third group considered Soviet designed reactors (WWER-1000 and RBMK). These discussions, which focuse progresn do 1993n si sinc curren e M lase ,th eth TC t t issues in the field of control assembly materials and the recommendations of the assembled experts for further R&D, are summarized in the following sections.

2. PWRs

. HerteP Thiy CEA-Saclayf b ro s d groule s pwa , France commos i s A . n practice amonR gPW operators, the control assemblies are referred to as RCCAs for simplicity in the following. This acronym stands for rod cluster control assembly. In addition, the neutron absorber, Ag-In-Cd alloy, is simply referre AICs a o dt .

2.1. Progress from 1993 to 1998

maie Th n problems see 199n i 3 were:

(1) Cladding wear and the rejection criteria linked with it; (2) Swellin AICf go ; and, (3) Potential problem timee th t ,sa suc swellins ha Bf go 4 C.

At the present time, the progress in these areas can be summarized as follows: (1) Good wear resistant coatings on the cladding materials are now in use. However, the possible transfer of the wear to the guide cards or other components (Zircaloy guide tubes) in contact wit coatee hth d claddin becomy kepe g b mindma n i to t proble ea s . ha d man (2) The mitigation of AIC swelling has been achieved by modification to the designs of RCCAs to include a larger gap between neutron absorber with gas filling, and improvement of the microstructur itselC AI f e (mainlth f o e y larger grain size with improved creep resistance). These improvements stem fro mbettea r knowledg thas w ne wa behaviouno th C f eo AI f o r available in 1993. (3) A better understanding of the behaviour of B4C shows that the problem of swelling, even though it occurs, is not as great as anticipated in 1993. Experience shows that B4C has to be washed out to be removed from the cladding and there is an irradiation threshold for the B4C to decomposee b coolante th y db .

2.2. Current issues

« Ther neea s ei investigato dt disposae eth f currentlo l y stored "spent" RCCA plao t r nd fo san disposae th thosf o l e that wil lfuture e aristh n ei . • Thers mentionea , is e d above potentiaa , r guidfo l e card guidd an s e wore tubet b ou no t s through contact with wear resistant coatings on the absorber cladding. • There is a need to analyse the existing criteria for the life of RCCAs and sub-assemblies. • Ther neea s develoei o dt p more efficient RCCA r corefo s s with highecontentX MO r s than those currently used. 2.3. Evolution in RCCA management

• Sinc weae eth r problems hav emose beeth r t nparfo t claddingsolvee th r dfo , furthethero n s ei r need for special constraints linked to cladding wear. This provides more freedom in the managemen RCCAsf o t .

• Ther neeo n chango ds et i currene eth t fluence B d limitation4C.an C AI n so • For RCCAs under development, however, it is hoped that the lifetime fluence can be extended.

2.4. Recommendations

Absorber materials: uppee th n ri parC 4 f controo tB r Fo fluencl w rodne sa • e criterion mus definee b t d because eth B4C will experience increased irradiation in the future. • Hafnium will have to be protected against wear and hydriding, which are linked because the Hf will hydride more rapidly if the protective oxide is removed by wear.

• Dysprosium titanate, hafnium boride (HfB otheabsorbe2and ) new r r materials must undergo more extensive evaluatio PWRsn i e us .r nfo

Other materials:

Wear has been considered only from the point of view of the cladding material. The wear of the guide cards and guide tubes must be studied also.

Development in inspection surveillance and end-of-life criteria:

• Clad swelling measurements performehave b o et d with increased precision. • Clad inspectee swellinb o t muca s n gdha o h larger scale.

3.BWRs

. RebensdorfB Thiy b sd groule s ABB-Atomf o fp wa , Sweden.

3.1. Progress 1993-1998

(1) Understandin e impacth f gf o fas o t t neutron irradiatio stainlesn no s steel (SS s increase)ha d significantly. However, there are still unanswered mechanistic questions associated with irradiation assisted stress corrosion cracking (IASCC). (2) From measurement n irradiateo s d samples r knowledgou , e abou e swellinth t d creean g p behaviour of control rods has increased significantly. (3) Current experience with "all-Hf' control rod quits si e encouraging. tipf shur H s(4fo t )dow n rod behavine sar predicteds ga . (5) The modelling of burnup behaviour (with respect to the control rod's working life) has been developed.

3.2. Current issues

(1) There is a need to compile recent B4C irradiation data, including its interaction with stainless comparo steelt d an , t witei h current model prediction attempn a n si validato t r improveo e eth models. (2) There is a need to study the benefit of Hf edges (strips and end plugs) on control rod wings.

(3) Ther neea s ei improvo dt knowledgr eou e abou interactioe Bd th 4tan C, f wherH f no e both absorbers are in direct contact, within a control assembly structure.

3.3. Recommendations

(1) For control cell positions the experts recommend the investigation of control rods with less reactivity worth. (2) Increase the number of control materials currently used in BWRs, choosing new materials, by comparin basie gth c physica chemicad an l l parameters (reactivity worth, materials density, etc.) taking cost into account. (3) Improve reactor water surveillance with respec (possiblH 3 o t t y also boro d lithiumnan o t ) improve our knowledge about control rod cracking. (4) Improv e technicath e l instrumentatio d managemenan n t acceptanc f dedicatef o eo T ND d control rod spenn i s t fuel pool optimizo s(t e contro managemend ro l t strategie reduco t d san e waste arisings). experte Th s ) note(5 d tha meetine th t g would have even been mor majoe th fruitfu l ral f playersi l , control rod vendors and utilities (especially from the United States of America, Canada and the r East) participatedFa d ha , . Therefore, they strongly suggest that future meeting f thio s s type should make every effor involvo t t largeea r selectio majoe th f nro players.

4. RUSSIAN DESIGNS

This group was led by V. Troyanov, Institute of Physics and Power Engineering, Obninsk, Russian Federation.

4.1. Progress 1993-1998

Since the TCM in 1993 a considerable amount of work on various absorber materials used in WWER reactor designs has been carried out. For example, the ultimate lifetime characteristics for 10 WWE controC 4 RB l rods have bee burnupB n studiedlifetime % th .50 d Thi limites o an ei ,t sw dno is 3 years operating as a control rod and 6 years as a shut down system rod.

numbeA f in-pilo r e testpre-irradiatiod an s n examination f dysprosiuo s m titanate have been carried out. DyiCyTiOa is now ready for industrial application. The design lifetime of the B4C/Dy2O3-TiO2 control rods is 10 to 15 years for WWERs. The long lifetime of the dysprosium absorbe confirmes ri year7 researc1 R operatiof y so d b MI e hth reacton i Dimitrovgradn ri .

RCCAs with hybrid hafnium/B4C absorber have been designed, manufacture loaded dan a n di WWER-100 199n 0i 7 (RovnoNPP, Ukraine).

e claddinTh g allo a fory( f stainlesmo s steel) wit e Russiahth n designatio s beeha n M XH n developed for control rod cladding application. There is experience of 17 years of successful operation with this steel irradiate waten di r cooled reactor formee th n si r USS higRo t h fluences.

4.2. Current issues

Control rod, high-temperature performance studies have e desigbegunth r nfo basis accident d (DBAseveran ) e accident analysis s alreadi t I . y clear thae higth t h temperature behaviouf o r Dy2O3-TiO2 is much better than that of B4C. Calculation models and computer codes for core structure mechanical analysis are being developed. Fuel assembly bowing is calculated by these codes during irradiation. Computer code DOR r analysiAfo f RCCo s A movement through bent fuel assemblie s beeha s n developed an d continue improvede b o st . RCCA drop-time excess possibld an , e incomplete insertio RCCAn a f no , may be calculated for the any RCCA in the core, for any period of operating time.

Other neutron absorber improvements are under investigation including:

• Hafnium; • Dysprosium (hafnium)/B4C hybrids; and, • Dysprosium hafniate.

Out-of-pil d in-pilan e e examination technique e beinar s g improved wit e objectiveth h f o s facilitating absorber material licensing.

In the multi-sectional control rods used in RBMK reactors, B,»C is the absorber and aluminium is the cladding material. Control rods based on dysprosium for RBMKs are under development, while basie th c desig beins ni g improved.

4.3. Recommendations

This group recommends that a common approach should be taken to the development of control assembly materials regardless of the reactor type. The guiding principle should be the goal of extendin lifetime gth f controeo l assemblie reactoo t p su r lifetime, while satisfyin requiremente gth s performind an A DB g f e safetwelo th n li y unde r severe accident conditions. These developments should be based on a careful comparison of the following characteristics of neutron absorbers:

• initial neutron efficiency and its evolution during operation; • chemical and structural stability; • nuclear density; • specific weight; • irradiation resistance (low swellin growthd gan , integrity, stabilit f physicayo mechanicad an l l properties); • chemical compatibility with structural material coolante th d san ; • mechanical compatibility wit core hth e structure, including abilit movo yt e within core structures without any problems; • thermal-physics characteristics; • structural integrity under accident conditions (including LOCA); • radioactivation of absorbers; • the possibility of re-fabrication or for other useful applications (for example as gamma-sources) or relative ease of disposal.

The group strongly recommended increasing the scope of international co-operation in this field, especially collaboratio fabricatioe th n ni r improvelicensind o nan w ne f gdo neutron absorbers.

5. OVERALL CONCLUSIONS

• The participants concluded that the opportunities for information exchange in the field of control limitedo materialto r thad behavioue fa an th t, e sar f controo r l assembly material botn i s h normal and accident conditions should be considered. • Vendors of CCRs are encouraged to pursue more vigorously the development of absorber materials less prone to swelling. participante Th • s underline neee dinvestigatth o dt disposae eth currentlf o l y stored "spent" RCCAs and to plan for the disposal of those that will arise in the future. Moreover, it was agreed that ease of storage and disposal should be taken into account during the development of new materials.

• Although the current designs of control assemblies have an outstanding safety record, their reliabilit economid yan c viabilit stiln enhancee yca b l improvementy db design i s materiald nan s behaviour that allow increased lifetimes in the core. • Wear-resistant coating surfacd san e treatment RCCAf so PWRn si performine sar g very wellt bu , the wear problem may just have been moved to the guide cards and guide tubes. • Unclad hafnium has excellent properties provided that it resides in the coolant flow where its oxide coating protect t frosi m hydrogen pick-up. « More efficient absorber materials based on enriched boron and hafnium will probably have to be develope advancer dfo d reactor types exampler ,fo , reactors designe fuels X buro dt n.MO • Dysprosium titanat s continueeha NPPdevelopee n i b showd e o an dst us sr greadfo t promise. According to the developers, it is ready for industrial application. In addition the special form of stainless steel developed for absorber cladding in The Russian Federation continues to perform well, even at high fluences. LWR CONTROL ASSEMBLY DESIGNS: XA0053627 A HISTORICAL PERSPECTIVE

M.W. KENNARD Stoller Nuclear Fuel, NAC International, Pleasantville Yorkw Ne , , United State Americf so a

J.E. HARBOTTLE Stoller Nuclear Fuel, NAC International, Thornbury, United Kingdom

Abstract

Control rod designs and materials have evolved in response to performance problems in both PWRs and BWRs. Irradiation-assisted stress corrosion cracking (IASCC absorbeo t e )du r swellin primarils gha y affecte controR dBW l rods with absorbersC 4 alss B ha o t occurrebu , PWRn di s with Ag-In-Cd absorbers primare .Th y problem somr sfo e controdesignR PW f lso rods have been rodlete weath f o rs against upper internal component swellind san g wit wea p crackingd hti an r . Competition amongst vendors for supplying control rod reloads has also resulted in design improvements. This paper provides an historical controR designsd BW ro ld an ,R revietheiPW rf w problemo remediesd san .

1. PRESSURIZED WATER REACTORS (PWRs)

1.1. Introduction

Vendors use different names to refer to their neutron absorbing components - Rod Cluster Control Assembly (RCCA) by Westinghouse (W), Fragema (FGA) and Siemens/KWU; Control Rod Assembly Framatom(CRAy b ) e Cogema Fuels (formerly Babcoc Wilcok& x Fuel Company, BWFC Controd )an l Element Assembly (CEA) by Combustion Engineering (ABB-CE). For simplicity in this part of the review e teron m, onls beeha y n e adoptewheus nr fo dreferrin l vendorsal o t g : Control Element Assembly (CEA).

1.2. Summar designf yo s

1.2.1. Introduction

The various vendors' CEA designs are reviewed in this section. Each vendor has one general design; however, different models exist to match the various rod arrays and fuel designs. The reloading of CEAs by one vendor in another's nuclear steam supply system (NSSS) has also produced new designs.

7.2.2. Mechanical design

1.2.2.1. ABB-COMBUSTION ENGINEERING

basie desigA Th cCE n offere r botfo dh d 16x114x1an 64 standard lattice CE-NSSS plante ar s similar with respect to the large diameter absorber rodlets or 'fingers'. In the CE fuel design, each guide tube displaces four cell locations in the fuel lattice as opposed to a single cell as in the other vendors' fuel designs. Typically CEAe th , s have fou fivr o r e fingers whic attachee har threadea y db d join centraa o t l hu r spide bo connectioe th s illustrate a rn i n pi n Figurn di A prevent . la e s rotatio e fingerth e f no Th . 16x16 System 80 plants (Palo Verde and Yonggwang) have a 12 finger CEA which is illustrated in Figure Ib. The fingers are attached to the spider by a bayonet-type fitting. 00 SPIDER ————_.

HOLDOOWN SPRING

SPIDER WITH NUMERICAL IDENTIFICATION

SERIAL No SPRING ON THIS WE

D FITTINEN — —— G SPRING RETAINER __— PLENUM -FULL LENGTH CONTROL —— HOLODOWN SPRING ELEMENT ASSEMBLY ——-STAINLESS STEEL END FITTING SPACER

HOLOOQWN SPRING

PLENUM

STAINLESS STEEL SPACER

B4CPELLETS 20 FEIT MfclAl & - REDUCED OlAMETLIi REDUCED DIA U,C B,,C PELLETS PCLUIG - STAINLESSSILLI SPACER V

SPACER b) 12-finger SILVER-INUIUM-CAOMIUM

SPACER

INCONELNOSfcCAl' FullFIG.1. length CEAs from ABB-CEfor a) 5-finger a) Standard 16x16 reactors and b) System 80 16x16 reactors (Reference: ABB-CE) The cladding and upper and lower end-fittings for all fingers are made from Inconel 625. With some exceptions Ag-In-Cd ,an botC 4 hB d absorber originausede e sth ar n .I l design 14x1r sfo 4 plantse ,th center finger was all B4C and the outer fingers had Ag-In-Cd in the tips with the remainder B4C. In all replacement CEA r 14x1sfo 4 plant originad san l CEA r 16x1sfo 6 plants wit C l finger4 hal ,B Ag-Ine sus - Cd tipsexceptioe On . ANO-centee n a th desig s A nf i o r 2 CE whicfingep n ni ti e hrth contains Inconel 625 in place of the Ag-In-Cd absorber. In another exception, the fingers of the System 80 CEAs contain all B4C. The tip has reduced diameter pellets surrounded by a low density (-20% TD) 347 SS liner (Feltmetal) purpose Feltmetae .Th th f eo l lineaccommodato t s ri e swellin absorbere th f go .

1.2.2.2. FRAMATOME COGEMA FUEL (BWFC)

BWFC manufacture NSSn dow Ss CEA it plant r r Westinghous sfo wels fo sa s a l e NSSS plants. With respect to the BWFC NSSS plants, the 15x15 plants utilize sixteen 'fingers' or rods fastened to a central cast spider assembly. The rods are attached to the spider using a threaded nut which provides a stiff connecting joint standare Th . d full length control rod BWFn si C plant 304Se sar S clad rods with type 304 or 308SS end plugs. The absorber material in the rods is Ag-In-Cd alloy. The designs for 15x15 plants are illustrated in Figure 2.

Full length control rod offeree sar d with Incone claddin5 62 l reduco gt eincreasen a wea d an r d absorber/claddin reduco t p swellingd gega ro . This desig s designateni s 'Planda t Life' control rods. More recently, two further design modifications have been offered:

• Hybrid absorbers consisting of a 36-40 inch (914-1020mm) region of Ag-In-Cd in the high flux region of the rod and B4C pellets to reduce gas release; • Cladding hardened with wear-resistant coatings: 304SS with Armoloy chrome plating; Inconel 625 cladding coated with chromium carbide Cr3C2; 316SS hardened with ion-nitriding.

BWFC offers 'black d 'greyan ' ' axial power stepping rods (APSRs) with reduced absorber lengths blace containd Th . kro inc6 3 hsa sectio) (914mm Ag-In-C f no d wherea 'graye containsth d ro ' s inc3 6 ha (160 sectio) 0 mm Inconef no l 600.

The CEAs offered by BWFC to Westinghouse NSSS plants contain all of the design permutations discussed above. The principal differences are that the control rod diameter, length and location in the fuel assembly are altered to be compatible with Westinghouse 15x15 or 17x17 fuel assemblies and Control Rod Drive Mechanisms (CRDMs).

1.2.2.3. WESTINGHOUSE

Westinghouse manufactures contro d assembliero l r theifo s r 17x17, 15x1 d 14x15an 4 NSSS plants. All designs are based on a central cast spider to which individual sleeves are brazed. The control rods are attached to the sleeves. This connection of rods to the spider is common to all W CEA designs allowd an s some additional flexur connectioe th n i e n during operation. Mitsubishi, unde licensa r e with Westinghouse, offer similaa s r product.

The standard 17x1containA 7CE s twenty-four rodlets l rodletAl . clae higsar n d i h purity Type Type 304Sus ed 308SSan plugsd Sen . Absorber materia availabls i l configurationso tw n ei . Full length standard rods utilize Ag-In-Cd absorber. Hybrid rods are also offered in which the lower 40 inches (102cm) of the absorber is Ag-In-Cd and the remaining 102 inches (259cm) is B4C. The standard and hybrid designs are shown in Figure 3.

There are 20 rodlets in the 15x15 fuel assembly array and 16 rodlets in the 14x14 array. Both cladding and absorber material, as well as construction, are identical to the 17x17 designs offered. COUPLING

CRA n r nn

SP10ER-

TOP VIEW

XEUTRCW ABSORBING MATERIAL- 134 inch (3400 mm)

CONTROL ROD -

F/G . BWFC2 . control assemblyd ro r 15x15fo plants

10 FIG. 3. Westinghouse standard and hybrid control rods

11 Hybrid absorbe currentlt r no rod e sar y manufacture either d fo e 14x1rth r 15x14o 5 designs. Some hybrid absorber rods were placed in service in the early 1980's in Westinghouse 17x17 plants. In 1987, Westinghouse introduced the Enhanced Performance Rod Cluster Control Assemblies (EP-RCCAs). The EP-RCCAs differ from the standard control rods in the following respects:

• Hard chrome, Armoloy-coated claddin mitigato gt e wear; • High purity 304SS cladding with increased resistanc irradiatioo et n assisted stress corrosion cracking (IASCC); • Increased diametra betweep ga l absorbee nth r materiacladdine th highese d th an n lgi t neutron flux locatio accommodato nt e absorber swelling. (See Figur. e4)

.381" Absorber Diameter .341" Absorber Diameter .336"

Rodlet ID .344"

FIG. Westinghouse4. increased designgap

12 EP-RCCAs have bee servicn ni mann ei y reactors ranging from 14x1 o 17x1t 4 7 core design ni Europe and the U.S., as well as in Japan. All 14x14 and 15x15 units are considered to be low-to- moderate wear environments 17x1e Th .7 unit classifiee , sar hige i.eT^ ar .h s a dwea r environments, and coincidentall havl Hf-RCCAyd al eha s replace EP-RCCAse th y db similarl.A y improved controd ro l design offere Mitsubishy db operatins i mosn gi t Japanese PWRs.

1.2.2.4. FRAGEMA(FGA)

The CEAs in EdF's 900 and 1300 MW plants in France are similar to the W 17x17 core CEAs. The 900 MW units have all Ag-In-Cd and the 1300 MW units have hybrid B4C/Ag-In-Cd absorbers. The Ag-In-Cd length is 40 in. (about 1 meter) at the rodlet tip, identical to the W design.

Harmoniw ne e designsth A n ™i Difference FG e ar : W d san betweeA FG e nth

• claddinTypL 6 e31 g tubing; • lon-nitrided tube surface; • Increased Ag-In-Cd absorber-to-cladding gaps with decreased absorber diameter.

stainlesL Typ 6 e31 s stee bettes ha l r mechanical properties, IASCC resistance formd an , s better nitride coatings than Typ use4 e30 d previously. controSomR eBW l rods have also switche Typo dt e ion-nitridee 316Th . d tube surface provides added wear-resistanc discusses i d ean d late thin i r s paper. The increased absorber-clad gap is to delay mechanical interaction, clad diameter swelling and potential crackin IASCCo t e gdu .

1.2.2.5. SIEMENS

Siemens/KWU manufactured CEA their fo s r standard 14x14, 15x15, 16x1d 18x1an 6 8 reactor cores desige rodlete th Th .assemblie d f no san s varies betwee differene nth t plants e layoue th Th . f o t rodlets within the spider for the 18x18 fuel is shown in Figure 5. By comparison the CEA for the 15x15 fuel, which has been the subject of the most significant wear, has 16 fingers, 12 of which have one rodlet hav 4 rodlet o d e tw finge an r d spe alss en a r e o th show t a Figurn no . e5

Siemens is the only vendor that uses stabilized stainless steel cladding (Titanium stabilized type 321) and no B4C in any of their rods. The Cr3C2 coating is offered as an option for wear resistance since the wear proble firss mwa t observe late th e n d1970sdiscussei s i d an , d later.

7.2.3. Nuclear design

In general e nucleath , r design objective vendor R e controth PW e r d systel ar sfo ro sal l f mo equivalent e .contro Thath , systed is tro l m must compensat e reactivitth r efo y effectd e fueth an l f o s water temperature changes that accompany power level changes over the range from full-load to no-load. In addition, the control rod system must provide at least the minimum shutdown margin under Condition 1 events (normal operation events musd capabl)e an tb makinf eo core gth e sub-critical rapidly enougo ht preven t exceedini t g acceptable fuel damage limits, assuming thae highesth t t worth contros i d ro l withdrawn at the time of a trip.

controR Al (clusterd PW l ro l ) systems whether they contain Ag-In-Cd, B r combination4Co , f so both absorbers are designed to meet the above objectives. Unlike BWR control blades, the loss of total contro wortd lro h (for safe shutdown materiao t e )du l irradiatio negligibles ni , since onl ysmala l number f controo l clusters (e.g. D-ban reactor)W n ki pard an ,t length rod somn s(i e cases, e.g., BWFC cores) may be in the core under normal operating conditions. There are no nuclear lifetime limits for PWR control rods.

13 BOLT SPRINP GCU

T SPIDENU R COMPRESSION SPRING

4620 short guide tubes l°ng guide tubes ___ -1 —— - ———————* i no guide tubes ——L J -fF \ f _-—- ^ 1 15 x 15 THREAD •&31I ... ^790 . k ...... - , - A7fl ... ——h- v

FIG. 5. PWR RCCA and control rod for Siemens Convoy plants (18x18) 1.3. Problem area remedied san s

1.3.1. Wear

The wear of control rods by rubbing against upper internal components, specifically guide cards, is a well known phenomenon today. Control rod wear is caused by several mechanisms. The predominant mechanism is rodlet vibrational contact against upper internal guide cards when control rod parkee sar thein di r fully withdrawn position. Wit e exceptiohth controe th f no l controrodW n si l bank D, or in the case of FGA also bank L, many utilities operate their control rods in the fully withdrawn position during 100% reactor power operation.

Secondary mechanisms such as sliding wear and rodlet tip wear can also contribute to the reduction of control rod lifetimes. Sliding wear occurs as a result of rodlet contact with fixed upper internal surfaces during contro movementd coree ro lwea p th f Ti . o r t occursou intd s resulsa o an f o t rodlet vibrational contact with fuel assembly guide tubes when control rod parkee ar s thein di r fully withdrawn position.

Figur showe6 distributiosa weaf no r CEApattern W rodle e weae r th o .t Th ra A st typicaFG a f o l ticrescent-shaps pi contaco t e du ts i wit fued e hth ean l element guide tube continuoue weae th Th .n ri s guid o stepping-inducet ee zondu s i e d frictios spreai d an dn axially ove fairla r y large span. Local discontinuous wear can occur there also from CEAs that are relatively stationary. The wear at the tip is against the fuel assembly guide tube and is generally crescent-shape.

Contro clad ro ld wea lean claca o rdt d failure d consequentlan s absorbeo yt r leaching. Clad failures can occur as a result of the high stresses imposed by differential pressure at localized wear scars. The differential pressure across the clad is caused by a combination of the external reactor coolant and internal gas pressures. These high stresses will either result in clad ovalization, which increases the possibility of a stuck rod, or brittle fracture which can lead to a dropped rodlet. Fatigue fracture during frequent load follow operation with rods could also occuweae th t rra location.

CARDS

RODLET

CONTINUOUS GUIDE

FUEL ASSEMBLY

Wear depth

FIG. 6. Typical Recorded Wear for a Low-Mobility RCCA (Reference: Framatome)

15 Absorber leaching occurs when through-wall wear exposes the absorber material of B4C control rods to the coolant. In this configuration, leaching leads to a loss of shutdown margin. Although leaching doe occut sno contron ri l rods that utilize Ag-In-C absorbee th s da r material, mechanical erosio lean ndca to an introduction of undesirable isotopes into the coolant (e.g., AgHOm).

variablee Th s that determin amoune eth typd weaf tan eo r are:

• hydraulic forces in the upper guide structure; • dimensions and weight of the rodlets; • dimensions and configuration of the structure designed to guide the control rodlets; • contro motiod lro n relate operationao dt l mode.

Al thesf o l e variable havn sca esignificana textent e effecth typed n o tweaan ,f o ,accound ran r fo t differences between vendor betweed san n various plantsame th f eso vendor.

The hydraulic forces are the most complex and difficult to characterize of all these variables. In typW ee guidth e structure instancr sfo e ther bots ei h downwar dupward floan wp dto fro floe mth w from bottoe th m with cross-flo middlee th n wi , whil Siemensn ei coolane 'th plantf o s upflow i l t sal e Th . rodlet typ Siemend W ean n s i s plant guidee sar intermediaty db e cards plantW , & whils theB e n ei yar enclose continuoun di s tube their sfo r entire length.

1.3.2. Wear remedies: Development of wear resistant coatings

1.3.2.1. Introduction

Several options exis reduco tt eliminatr eo e wear:

• modif hydraulice yth s • modify the upper internal structures • modify the absorber rodlets to provide a wear resistant surface for Type 304 steel • replace Typ wit4 morhe30 a e wear resistant material.

e easiesTh f theso t wear-resistana et optionpu o t s i s t coatine rodleth n tga o surface us r o e different vendore claddingth providf w o l sno Al . e improved r enhanceo , d designs dato t ed mosan , f o t them have performed satisfactorily. A hard wear-resistant coating on the rodlet may transfer the wear problem to the guide cards or guide tubes. Increased monitoring of the coating surfaces is needed as coating introducede sar . Another consideratio mechanicae th s ni l integrit corrosiod yan n resistance th f eo coatin reactore th varioun ge i .Th s type coatingf so discussee sar d below. 1.3.2.2. Chromium carbide Cr3C2

Coating f Crso 3appliee C2ar d eithe plasmy rb detonatioy b ar sprao n (LC1C)n ygu ngu procesa , s patented by Union Carbide. The carbide is in a NiCr binder and applied at a 150 °C tubing temperature. This typ coatinf eo favores gi d primaril BWFy yb Siemensd Can coatinge Th . s use aboue dar t 40-5m 0u

thick. There is concern that the very hard surface of a CrC2 coated rod could result in increased guide 3 card wear (Cr3C2 is harder than chrome-plated or ion-nitrided surfaces) but early measurements in USA after two cycles showed no evidence of this. Siemens has not detected wear or flaking in plasma-sprayed

CrC coated rod mann i s y year f servico s a lea n di e assembl Germann i y morn i d e yan tha0 25 n 2

accumulate3 d cycle f loweso r duratio aboun no CEAs0 4 t dato T onl.e eth Crys i Ccoatin g whics hha 2

had no reported performance problems. Its3 high cost is its major disadvantage; the current process requires that each tub coatee eb d individually.

16 1.3.2.3. Chromium plating

Cr-plateo Tw d (Armoloy) CEAs fabricate BWFy db C were irradiate reactorS U a n d.i Aftee on r cycl f irradiationeo maximue exceede) th , um mil 9 4 e 2. ms(7 dth d weaan ) r um depthmil 3 8 2. s(5 f so complete coating thickness of 0.05 mil (13 um). In the next cycle standard CEAs replaced these two. The wear in these standard CEAs were, after one cycle, 14.7 mils (370 um) in the location where the Cr- t significan no weae s mil3 otheth e um)8 2. d rwa th s (5 d n an rplate,i t locationha t clead y dno ro rs wh i t I . e BWFth C chrom applicatio s effectivW a e e t platth no s experiencs a eei W n( s lesei l s mi tha 5 n0. threo t wea) p (1eu 3cycles)um n i r . This experienc BWFd ele concludCo t e that chrome platint no s gi very effective. The W EP-RCCA, described above, uses a chromium plated (Armoloy) SS 304 clad. The nomina um)3 lfirse thicknes(1 thesf l Th o .t platinmi e e 5 th rod f 0. so s gi were delivere 1987n di .

1.3.2.4. Nickel plating

The Framatome tests showed that this performed well. However, nickel solution and activation in e coolanth t could hav enegativa e effec activitn o t y transfer. f coursThio s i s ei alloy N tru r f o efo s claddin wells ga .

1.3.2.5.Nitriding

Nitride extremele sar commerciae y th har d dan l nitridin stainlesf go s steel surface varieta y sb f yo processes for improved wear and fatigue resistance is common. One of the potential disadvantages is poor corrosion resistance.

The Framatome have selected the ion-nitriding process for nitrogen enrichment of the steel surface of their Harmoni™ design CEAs. Electrical discharge in a nitrogen atmosphere produces a plasma which depassivate heatd surfacse an sth convertd ean s molecular nitroge ionio nt c nitrogen nitrogee Th . n nca then diffuse into the steel surface.

nitrogee Th n resides interstitiall thicsurfaca m u n yi k 0 2 (Figure range o layet th . 2 n 1 e7) i r

Chromium nitrides (Cr r Cr No fory 2Nmma ) with excess nitroge r thickeno d an r ) coatingum 5 (3 s degrade the corrosion resistance of the coating. This was a problem with earlier coatings until the process and final product were optimized.

Ion-nitriding is performed in a furnace as shown in Figure 7. Tubes with their tips welded on are furnacee th n suspendei p . to Afte sealed e an dth r t vacuuda m evacuatio mixturna f nitrogeo e d nan hydrogen is introduced and a high voltage generator (300 - 1000 volts) forms the plasma. The corrosion resistance is closely related to the process temperature and after evaluation of several process conditions, currene th temperaturw lo t e treatmen s foun wa tpreven o t d t nitride formatio alsd an on provide good corrosion resistance.

1.3.3. Absorber swelling andlASCC

1.3.3.1. Introduction

Mechanical interaction between the absorber and the cladding, which results in cracking of the clad, is one of the mechanisms which governs the service life of control rods. The phenomenon can occur slightln i , y different forms, wit l absorbehal r materialse case th th f eo n I :. Ag-In-CdHf r o C 4 B , first two the cause of the intergranular cracking of the clad is a combination of some or all of the following:

• absorber 'swelling', which stresse clade sth ; • clad wall thinning, due to wear; • irradiation effects on the clad mechanical properties; • coolant chemistry.

17 Pressure regulation X I

200-250 Tube r Furnacspe e Charge

20um Coating

Type 304L Type 3I6L

F/G 7. /OK nitrided control rodlet coatings by Framatome

problee Th m generally occurregioe th f highesn si no t neutron fluence tipd , ro nea, whice rth s hi also vulnerable to sliding wear against the guide tube. In some instances the cracks remain hairline (i.e. very tight continued )an d operatio possibles ni . Thi particularls si y tru f Ag-In-Ceo d rods. Deterioration crackee oth f d conditio lean leachinno ca dt f solublgo e B,jC intcoolane oth t and severn i , e caseso t , complete failure of the rod resulting in the loss of absorber pellets. In the case of low solubility Ag-In- Cdcombinatioe th , f exposurno absorbee th f eerosioo d an r n processe resulisotope n th sca n i t e Ag110m appearing as a contaminant in the primary coolant.

18 The problems with Hf absorber rods fabricated by W are related to hydrogen diffusion through the cladding wall causing primary hydriding of the Hf. The manufacture of this design has been discontinued by W but some of the originally supplied Hf CEAs may still be operational.

1.3.3.2.Ag-In-Cd

swelline transformatioe Th th f Ag-In-Co gt o e du s df i indiu o n m (In) int (Snn oti neutroy b ) n capture. The change in chemical composition and the basic transmutations are shown in Figure 8. The mechanism leading to IASCC clad cracking in Ag-In-Cd rodlet tips is as follows:

• Absorber - clad gap closure due to creep and slump of Ag-In-Cd under its own weight, and compressio rodley nb t sprin present)f g(i ; • Increased creep rat absorbef ei r temperatur increaseds ei ; • Swelling of Ag-In-Cd; • Afte closurep ga r , swelling continues; • The clad fracture strength, or ductility limit, is reached and the cladding cracks longitudinally; • Stresses are relieved and swelling continues.

The main operational effect of swelling is the potential interference within the dashpot region, leading to incomplete or slow rod insertion.

2. BOILING WATER REACTORS

2.1. Introduction

The primary mechanical life limit of BWR control rods has been due to IASCC cracking of containers for the B4C absorber. The B4C irradiation induced swelling is the source of stresses and the coolan e sourcth t f corrosiono e . cruo Strest e d du saccumulation s have also contribute IASCCo dt . Improved design materiald an s s have been introduced periodically sincC 4 e discoverB th e e th f o y swelling proble life eth design e limid m197th n i an f 8s beeo t sha n extended e improveTh . d mechanical nuclead an r currendesigne th d tsan performance problem e discusse ar se thre th r e dfo curren R BW t contro vendorsd lro : SiemensABBAd an E G , .

2.2. Summar designf yo s

2.2.7. Introduction

All BWR control rods are cruciform in shape and the four wings fit between the channeled fuel assemblies. Three types of control rods are in use:

• GE and Siemens: each of the 4 wings has 18-21 tubes filled with vibro compacted B4C powde bundled ran d together usin gU-shapea d stainless steel sheath U-shapee Th . d sheaths weldee ar centea o dt r post lowea o t , rhandla castino t ford o et housine gan me th th r gfo absorbeC 4 B r rods; • GE: the newer 'Marathon' blade wings are composed of longitudinally welded tubes with a square outer envelope filled with encapsulated, vibratory-compacted B4C powder or Hf- rod seale d endse san th t d;a • ABB: The wings of the ABB control rods are solid pieces of stainless steel with gun-drilled

horizontal holes thafillee ar t d wit powdeC hB vibratory b r y compaction holee e Th . sar 4 covered wit hstainlesa wine th seas gi d s lan steecolumn C weldedr 4 B ba l e alse sar Th o. connected by a vertical gas plenum along the outer edge of each wing.

19 100 Silver

107 Cd108 stable 109 110 ->Cd110 stable In 115 stable 113 cd Cd114 stable

10.0

o O

1.0

.5 1.0 1.5 2.0 2.5 3.0 3.5 Fluencex }21 nvt-

Calculated8. FIG. changes compositionin ofAgln-Cd alloy functiona as affluence (Reference: WAPD-214, December 1959)

In orde o extent r e mechanicalth d lifetime bladeth f o ,e each vendo s developeha r d hybrid absorber contro e higl th rod hn i exposur sf whicH e hus e section LASCd san C resistant steel absorber tubes. Almost all control rods currently supplied are of this type.

20 2.2.2. Mechanical design

2.2.2.1. ABB ATOM

The currently offered control rod designs are all of the general type described in the previous sectio diffed nan r primaril e absorbeth n yi r desig s listena d below wit commerciae hth l designation ni parenthesis:

• All B4C absorber (CR-70); rodf tipholeH e n si th , t standarsa • d diameter absorbers (CR-82); • Same as the CR-82, but smaller diameter absorbers (CR-82M); • Hf rods in holes at tip and plugs at rim, standard diameter absorbers (CR-85).

The designs are shown schematically in Figure 9 and the reasons for the differences discussed below.

controC 4 B l CR-70lal rode originae th th , s ,i lplants B supplAB .o yt Sale f thio s s model essentially stopped in 1984 because it was superseded by the hybrid designs with improved mechanical lifetimes.

e CR-8Th 2 contro d desig ro s lABB' wa n s first progression froe basimth c C 4 CR-7B l al 0 containing rod. The top 152 mm (6 inches) or 22 holes of the absorber region in each of the four absorber wings were filled with hafnium pins remaindee Th . absorbee th f ro rl regioal Bl n i al 4 Cs s ni . A the ABB control rods the four wings are a solid piece of stainless steel with gun-drilled horizontal holes that are filled with BC powder, by vibratory compaction. The CR-82 blade material was until recently

high purity 304L stainles4 s steel with control of P to <0.025%, Si to <0,3%, S to <0,02%, B to <5 ppm and C to <0,03%.

Early irradiation results indicated that further mechanical life extension woul desirable db e which developmene leath o dt modifiee th f o t d CR-82 CR-82Mr o , CR-8e Th . recommendes 2i shutdowr dfo n rod applications. The material change from Type 304L to Type 316L stainless steel, made for the CR- 82M, has been applied to this as well as all other control rods.

The CR-82M has two features designed to extend the mechanical lifetime:

• A modified absorber configuration to minimize strain at the blade outer surface achieved by: smaller absorber holes; thinner ligaments between holes; thicker outer wall; • more IASCC resistant steel: Type 316L instead of high purity Type 304L (L = low carbon).

The comparison of absorber configurations is shown on Figure 10. Finite stress analysis comparison showe decreas% 40 d a modifie e strain ei th r nfo d desig shows na Figurn ni Ex-reacto. e11 r materials tests have shown that Type 316L stainless steel has better IASCC resistance than Type 304L. The difference between the two steels is 3% molybdenum contained in Type 316.

The CR-85 control rod design has hafnium plugs at the wing tips of the horizontal absorber holes. The remainde f thoso r e holes contai standare nth d vibro compacted powder extensive Th . e cracking e initiafounth n i ld lead assemblie s associatewa s d with secondary hydridin d swelline an g th f o g hafnium standare CR-8e th Th .d 5ha d diameter absorber hole structura L Typd 4 san e30 l steel. Currently there is no further interest in this design.

21 The CR-85M was then marketed as the improved version which has the same design modifications as the CR-82M. This design has been used by Japanese utilities, however since the absorber diameter change would requir a licensine g change, the ye standar th opte r fo dd absorber diameter and only the material change to Type 316L steel. The design is identified as CR-85M-1.

Boron carbide powder

WXK!^mmismrm

CR-70

Hafnium rodlets

CR-82

F'^T^^^T^^^^^^^^i'TKifiiaiiinHMinBiijr

Boron carbide powder Hafnium rodlets

CR-85

Hafnium plugs

Boron carbide powder

controlFIG.ABB 9. blade types

22 High worth control rod versions of the above designs have also been produced. The higher worth is achieve increasiny db diametee glengte absorbee th th th f d ho potentia ran e rth holest a t l bu ,cos f o t higher stresse blade th t esa surface. Wit trende hth fuen si l development towards higher burnup, longer cycle highed san r power densities, control rods wit increasen ha d contro wortd ro l h compared wite hth original equipment rod desirable sar e for:

• Improved shutdown margin • Flexibility in reload pattern designs • Possibilit exteno yt d fuel cycle lengt burnud han p • Operational flexibility.

CR-82 CR-82M

CR82

© Hf-hole Max. stress hol- C e B4 ©

Geometry changes (mm)

Blade material: Type 304 S originallLS y

Type316LSS currently

CR82M

Reduced stress Max.stress

Blade material: Type 316L SS originally and currently

FIG. Modifications10. of ABB's CR-82 control blade

23 ABB control rods are called 'high worth' (HW) control rods if they have 5-15% (relative) higher rod worth than the original equipment rods. The CR-82 HW design has a larger diameter absorber (0.217 in., 5.5 mm) and a longer length absorber (4.09 in., 103.9 mm).

(Differential gas pressure Ap=12,l MPa) 2,5c-4

-5,Oc-5 -

-l,Oc-4 30 35 distance [mm]

FIG. 11. Outer-wall strain analysis

2.2.2.2. GENERAL ELECTRIC

GE produces two general types of advanced control rod designs. The first type is the Duralife Serie controf so l rods which primarily differ fro originae mth l cobalequipmenw lo f to e tus desige th y nb alloy higd san h purity 304SS tubing additionn I . , later design Duralife th n si e series contain hafniun mi various high exposure portion rode th .f so

The second, mechanically dissimilar design is the Marathon control blade which uses square tubing welde foro dt m each blade wing. Hafnium absorbe alss ri o use hign di h exposure regions, sucs ha the edge tubesnecessarf i d addee b an , platn s da y ca e materia blade th t ea l tip. Individual blade hafnium loadinvariee b y dg ma wit h contro requirementsd ro l .

distinguishine Th g design feature eacf so h Duralife control blade typ summarizee ear d below:

• Bl Duralif4C-absorbeAl - 0 12 e r design utilizing high purity 304S non-cobald San t pind san rollers. featuree Duralif e • th f th wito f 0 additiol e so Duralif h12 eth Al - p hafniu0 f to no 14 e th t ma of one absorber section. • Duralife 160 - All of the features of the Duralife 120 with the addition of hafnium absorber highese rodth n si t burnup locations (outside edg eacf eo h wing). • Duralife 190 - All of the features of the Duralife 160 with additional hafnium absorber in the second highest burnup location (top of each blade wing). In addition, a lighter weight velocity limiter is also used.

• Duralife 215 - All of the features of the Duralife 190 with additional B4C absorber material in thicker blade wings. The Duralife 215 is specifically designed for C-lattice plants. • Duralife 230 - Similar to the Duralife 215 but designed for D-lattice (BWR 4-6) plants.

24 Marathon Marathoe Th n radicalla contro s i d ro ly different concept when compare Duralife th o dt e series. maie Th n differenc Marathoe th n i e replacemene nth contros i e absorbed th ro lf o t r tubsheatd ean h arrangement with an array of square outer envelope tubes (Figure 12), which results in reduced weight and increased absorber volume. In addition, the full length tie rod used in previously approved designs is replaced with a tie rod segment, which also reduces weight (Figure 13). The outer corner sections of the individual tubes are used as wear surfaces against the channels. They are fabricated from a high purity Rad Resist Type-304 stainless steel that provides resistanc o irradiation-assistet e d stress corrosion cracking. Each wing is comprised of either 14 or 17 absorber tubes. If needed for enhanced life, provisio incorporatioe mads ni th r efo hafniua f no m plat rectangulaa n ei r section (pocket) beloe wth controe th f o l p rodto handl beinp ,e ti wit th e gt h eth a variabl lengtn ei h dependent upon individual customer requirements.

U.S. Patent 4,929,412

P., ,70

OiQCiQOGiGtQGC-iQO

InternalFIG.12. features Marathon ofthe control blade (U.S. Patent 4,929,412)

25 Hafnium blade In co-operation wit, Hitachh GE Toshib d an i a have introduce all-hafiiiuo dtw m absorber control blade designs for use in BWR control cell locations. The two designs, one with Hf-plates and the other with Hf-rods, are shown in Figures 14 and 15 respectively. With both designs, the control rod drop rate is approximately 0.85 m/sec, or less than the maximum allowable rate of 0.95 m/sec, and the time to 90 full-if %o n positio approximatels ni seconds6 y2. r leso , s tha maximue nth m allowable6 tim3. f eo seconds. U.S. Patent 4,929,412

H

r

FIG. 13. Marathon control blade (U.S. Patent 4,929,412)

As with the early GE version, the plate design utilizes two hafnium plates on each face of the wing of the cruciform blade. The absorber plates are fixed to the sheath by stainless steel pins which also provide separation of the plates. The water-filled gap between the absorber plates forms a flux trap in which epitherma fasd tlan neutron moderatee sar absorbedd dan savo T . e weigh providt tbu requiree eth d shutdown margin, four hafnium plates with different thickness are used with the plates at the top of the blade (most likel peae yth k exposure location) havin maximue gth m widt f approximatelho y (0.08 inch) 2 mm. Proceeding down the blade, the thickness is reduced in three stages to approximately (0.04 inch) 1 mm.

26 rod-type Inth e blade structure th , essentialls ei ydesigC 4 identicaB e n th (Duralif o t l e 120) with the hafnium absorber rods replacing the B4C tubes. The absorber rods are suspended from the top of each wing. The rods are thicker (5 mm) in the top half of the wing and thinner (3 mm) in the bottom half. The all-hafnium designs have been in service since 1989.

Handle

Roller

Blade

Sheath Tie rod Sheath Hafniura plate

\ Hafnium plate Pin Water gap SectioA nA-

Velocity limiter

Socket

FIG. 14. Plate type Hafnium control blade

2.2.2.3. SIEMENS

The Siemens Hybri controd4 desigd ro l s simila ni Duralife th o t r e typcontroE eG l rodse Th . earlier Hybri edgf desigd1 H ed rodnha s only. Ther bees echangha o ndesign e th n thesf ei no e rods sa Siemen longeo sn r manufacture controR sBW l rods.

2.2.3. Nuclear design

The key nuclear criterion governing the nuclear design of a core's worth of BWR control blades is the cold shutdown margin specified in the Technical Specification, with the most worthy blade stuck out of the core at any point in a fuel cycle. A second criterion for replacement blades is to maintain or improv e scraeth m reactivity functio s definee Technicaa n th y b d l Specifications. Meetin scrae gth m reactivity criterion is primarily a mechanical design concern (blade weight) since blades are generally designe havo dt reactivitea y hold down characteristi t leasca t equivalen thoso t e the replacinge yar .

27 Tie rod Sheath Hafniud ro m

Hafniud ro m

SectioA - A n

Velocity limiter

Socket

typeRod FIG.Hafnium15. control blade

controR BW l blades normally operat goocora e r th edfo n e i portio theif no r histor thereford yan e the absorber material is subject to depletion, and loss of worth. The end of nuclear life is defined as a 10% relative reductio n coli n d reactivity highesy wortan n i ht exposed axial quarter segmente Th . relative worth reductio takes ni n fro cole mth d initial controworte th f ho l blad value% design10 e e Th . most probably evolve dconservativn i froe mus e core design analysis BWRn I . a significans t axial segmen defines i t quartea s da r axial segment (generall quartep to e yrth Siemend controls)an E G s n I . plants the quarter axial segment criterion is also compatible with the original plant process computer tracking the control rod fluence. In modern process computer systems the control rod tip is also tracked for fully withdrawn control rods.

Worth reductio quarted nan r segment average depletio determinee nar d through detailed neutronic analysis of a blade as mechanically designed. If Hf is present in the blade design, an equivalent average depletion expressed in terms of 10B, is used:

Equivalent Avg. Depletion= Total absorptions in BaC + Hf_____ Total Boron atoms if entire blade were B4C

28 This allow plane sth t process computer control blade trackin usee b do gt directly with multiple

absorber control blade designs. BWR control blades generally can be designed with extended nuclear life (more snvt 1 to reach 10% worth loss) by increasing the number of absorber atoms per unit volume. If this is done by increasing the absorber surface density (absorber surface per unit volume of absorber), initial worth will increase, so that lifetime increases could be limited by the 'matched worth' criteria if applicable.

3.3. Problem areas and remedies

3.3.1. swellingC 4 B irradiatedand assisted stress corrosion cracking

The B10 isotope in boron carbide (640) reacts with neutrons to form He and Li in equal 10 proportions. As an example 30% B depletion produces 3 w/o Li. The B4C lattice accommodates the He t bul bu s goooccurk anC , 4 ha swellinresulda Li B d C s e 4 shigshowa s B a tth f he gFiguro n n i Th . e16 temperature strength d sincan ,t i operatee t relativela s temperaturesw lo y e swellinth , g resultn i s significan containertC 4 stresseB e th addition n sI .i release s i e H nportioa e d th producin f no g additional pressurs ga depictes ea combinatioe Figurn di Th . e17 thesf no e stresses wit oxygenatee hth d cooland an t irradiation e irradiatioresulth n i t n assisted stress corrosion cracking (IASCC e stainlesth f o ) s steel container. The B4C is slowly washed out through the crack by the coolant.

effece firss Th wa t discovered durin examinatioe gth bladeE G f nso fro m197German a i R 8 nBW and since the s beenha n establishe mechanicae th s da l bladeslifR e developmene limiBW Th .f o t f o t improved blades focused on better IASCC resistant steels and alloys as well as substitution of the much lower swellin Br g 4fo ratC f .eH

3.3.1.1 remedieB AB . s

For mechanical life extension ABB is relying primarily on the current improved features of:

• smaller, more closely spaced absorber hole reduco st e blade surface stresses • steeTyp L improve r 6 lfo e31 d IASCC resistance.

More advanced remedies are being irradiated in assemblies to test high density (99% of theoretical) B4C pellets as a substitute for powder.

believeB AB s that standard satisfactors i Typ 6 e31 9x10o yt 21 nvt, whic beyons hi nucleae dth r lif f 6xl0eo 21 nvt. Nevertheless testina , g progra initiates mwa hign do h purity versions a f Typ so 6 e31 well as various heats of the same composition. Grain boundary segregation has been studied for evidence of embrittling elements Fe, Cr, Ni, Mo, Si, O, S, and P. Work on Type 304 has shown the presence of Ni, S, and P peaks, as well as Cr depletion, at the grain boundaries. The latter effect is believed to accelerate IASCC.

3.3.1.2 remedieE G . s

Substitution of Hf for B4C in high exposure positions in the Duralife series eliminates some of the B4C swelling. This also extends nuclear lifetime. In addition, encapsulating the B4C in the Marathon design reduces the impact of B4C swelling on the square tubes. Increased void volume in the square tubes of the Marathon design accommodates B4C gas release at higher exposure. Substituting Hf for B4C obviously reduces the problems associated with B4C.

snvt - smeared thermal neutron fluence based on the exposure of the four fuel assemblies adjacent to the control blade of interest, unit f 10so 21 n/cm2

29 Crevice corrosio causbelievee s th nwa f crackine eo b o dt originan gi controE G l l blade designs, becaus a erelativel y stagnant crevic s formei e spac th n ei d betwee e handle e sheatth th nd n I an h. combination with a relatively high impurity content in the coolant, the concentration factor in the crevice will accelerate corrosion. Subsequently GE revised all of their designs to eliminate locations with stagnant coolan modificatioe th y b t additiod nan f flono w holes nea sheath-to-handle rth e weld wels a , l as modifying the top weld configuration. Modified lead control rods have been operating since 1986 and these changes have largely eliminate crevice dth e corrosion problem.

Helium release [ %]

measured data

0 10 20 30 40 50 Averag B-1e% 0 depletion

Pressure (MPa)

30

measured data

0 10 20 30 40 50 Average % B-10 depletion

ABB-AtomFIG.16. helium release controland blade pressure data (Reference: ABB)

30 + 1n-—>7LU4He diffuse throug crystallite hth e frea eo t surface ^Hegas precipitat dislocatione th t ea s graid an n boundarief so boron carbide -> Boron carbide swelling

0.1 urn Boron carbide crystal TEM image of irradiated boron carbide and helium gas bubbles

HeliumFIG.17. migration (Reference:B^C in ABB)

REFERENCES

. Sasaki Y . Got K d oan , ] 'Counter-measure[I ControR Degradatiod PW r Ro l fo s Japan'n ni e b o t , published. . ShirayanagiH [2] . FukumotT , . ShigaS d o,an 'Advanced Control Rod r Japanessfo Plants'R eBW , publishede b o t . [3] Kodama et al. 'IASCC Susceptibility of Austenitic Steels Irradiated to High Neutron Fluence'. Sixth International Symposiu Degradation mo f Materialno Nuclean i s r Power System s- Wate r Reactors Diegon Augus, PublicatioSa . A ( CA , 3 '9 t TMs)f no . . BruemmeS [4] . 'Graial t re n Boundary Chromium Concentration EffectIGSCe th IASCn d o sC an C of Austenitic Stainless Steels'. Ibid. [5] H. Chung et al. 'Grain Boundary Microchemistry and Intergranular Cracking of Irradiated Austenitic Stainless Steels'. Ibid. Kasahar. e EffectS Th . Minof ] al o s t [6 ae r Element IASCn so C Susceptibilit Austenitin yi c Stainless Steels Irradiated with Neutrons'. Ibid. Katsur. R . 'Effecal t ] ae Stresf [7 o t IASCn so IrradiateCn i d Austenitic Stainless Steels'. Ibid. . CooksoJ . 'Th al ] et n[8 e Rol Microchemicaf eo Microstructurad an l l Effect IASCe th n i sHigf C o h Purity Austenitic Stainless Steels'. Ibid. [9] A. Jacobs et al. 'The Correlation of Grain Boundary Composition in Irradiated Stainless Steel with IASCC Resistance'. Ibid. . Garzaroll[10F ] . 'Deformabilital t e i Higf yo h Purity Stainless Steel Ni-Basd san e Alloy Cor e PWR'a th f en o si . Ibid. [II] C. Vitanza 'In Pile Test Programme in the Halden Reactor'. Ibid. NEXT PAGE(S) left BLANK

31 PWR CONTRO DEGRADATIOD LRO OBSERVES NA DN I POST-IRRADIATION EXAMINATIONS IN STUDSVIK XA0053628

T. JONSSON Studsvi, kAB Nykoping

T. ANDERSSON . BJORNKVISL , T Vattenfall AB, Ringhals, Varobacka

Sweden

Abstract

controR ThePW l rodlets examine commo e Studsvin di th f o e nk ar typ e with Ag-Cd-In allo absorbes ya stainlesn i r s steel cladding. The main degradation mechanisms for this type of control rods are diameter increase due to absorber swelling and claddin e weath f ro g agains componentse tth , which guid rodlete eth s abov coree eth . Some example degradatiof so e ndu to these mechanisms are shown and discussed. The consumption of neutron capturing isotopes is not life-limiting for this type of control rods.

1. INTRODUCTION

Totall rodlets9 y1 controR fro massemblied PW ro l s from hav 4 Ringhald e an undergon3 , s2 e hot cell examinations to varying extents in Studsvik. The main purpose of the examinations has been studo t degradatioe yth rodletf nlono o t e g sterdu m exposur reactoe th o et r environment.

2. WEA CONTROF RO ABSORBED LRO R

The control rods are withdrawn from the reactor core during the main part of the reactor operation. Onl lowee yth rrodlet e endth f exposee so sar higa o dht neutron fluxcomplete Th . e length of a rodlet is however exposed to the water coolant above the core. This will cause some vibration of the rodlet wead san r agains differene tth t components observatione Th . weaf so r agree essentially with previously reported observations e.g. [1].

The lower ends of the rodlets are supported by the guide tubes in the top end of the fuel assemblies. Some e bulleweath f o rt nos normalls ei y observe t celho l n examinationdi s (Figur. 1) e The lowee wea th ofted t ra ren n extends irregularly aroun complete dth e circumference depte Th .f ho observee b weae n th ca r dimensionan di l measurements (spiral profilometr depte Figurn i yTh . h e2) and extent of the wear near the lower end has in no observed case been large enough to threaten the integrity of the rodlets. It should however be observed that the wear makes the bullet nose less suitabl calibratior efo f equipmentno , when dimensional measurement f rodletso performee sar a t da power station.

A guide region above the core also supports the rodlets. Wear occurs also in this region. A quantificatio f thio n se disturbe b wea n ca ry interferencdb e wit diametea h r increas e samth n ei e regioabsorbeo t e ndu r swelling e spiraTh . l profilometry curv Figurn ei showe3 s e loweweath n i r part of a rodlet with very small diameter increase due to absorber swelling. The wear at the guide cards highe alonp u r rodlete gth oftes si n localise smalo dt l marke depte regionth th n f d ho sca an s amoun larga o t claee parth d f thicknesso t bee s case ha non t e i observe n .I d thafrettine th t t sucga ha locatio resultes nha penetratina n di g defect (Figurspirae Th . l profilometre4) Figurn yi showe5 e sth axial extension of wear at guide cards. The fretting can in some cases have a peculiar appearance, which is not easily explained by pure fretting (Figure 6). The mechanism is unknown to the examiners, but some kind of contact corrosion is suspected. The wear of the corresponding guide card has not been examined.

33 FIG. 1. Lower end ofrodlet with wear

m fi 0 10 +

FIG.Spiral2. profilometry of lower ofrodletend with axial crack, swelling wearand

34 " . . -.....: .^..—* »-. muni....-•"'.••";'..._,: 1 !.LJ.l :i|, + 100

. Spiral3 . profllometry of lower ofrodletd en with axialo wearn t bu crack swellingr o

FIG. Penetrating4. fretting wearto againstdue guidea card

200mm

p- jlfjji,^ . 1'

' ' *

/7G. 5. Spiral profllometry showing wear against guide cards

35 FIG.Fretting6 wear to due against guidea card, irregular shapescontactto due corrosion and/or material loss in both rodlet and guide card

3. SWELLING OF ABSORBER

Gamma spectrometry measurements confirm thae neutroth t n dose receive controe th y db l rodleta ver s y ha sstee p axial gradient e swellin(Figur e Th absorbe th . f 7) eo g r increases with increasing dosee swellinTh . g shoul de absorber consequentlth f e o Th d . en e largese b yth t a t maximum diameters however, occur about 30-6 e rodletsfro m e endth 0m th f importann o s A . t factor for an explanation of this phenomenon is the plasticity of the absorber. The axial cut shown in Figur showe8 s that plastic deformatio swellina f no g Ag-Cd-In absorbe occun Ag-Cd-Ie ca r Th r n absorber has been extruded out into the open space between the end plug and the cladding The free space in the lower end of the rodlet and the deformation of the absorber will result in a delayed increase of the outer diameter of the rodlet. More free space in the lower end of the rodlet will therefore increase this delay

-§. 3,OOE+10 ca

1,OOE+10

0,OOE+00 50 100 150 200 250 300 350 400 Distance from lower end (mm)

FIG 7 Gamma measurement shoving Co-60 activity in end plug andAg-IlO activity in absorber

36 As:-Cd-In absorber

Cladding

FIG. Metallography;8. Axial throughcut plug absorberend and

4. DIAMETER INCREASE OF RODLETS

Significant diameter increases of rodlets have only been measured for rods with axial cracks in the cladding. Typical examined defect rodlets have axial cracks only on one side of the rodlets. It is evident that a starting crack will decrease the stress in other parts of the circumference. In most cases ther s onli e eaxia on y l crack. Inspectio e outside rodletth th f f o no e s show n somi s e cases irregularitie propagatioe th n i se axiath f lno cracks e propagatio(FigurTh . e9) n axiae frontth f lo s cracks extend further on the inner than on the outer side of the cladding. The observed short interruption interpretee cracke b branchino th t n n e i ssca du s da g during axial growt crackse th f ho . Without evident experimental proof it is suspected however, due to the environmental conditions, that the cracks start fro outee mth claddinge r sidth f eo e observation conclusione Th . th t no t sbu s agree with published data [2]. The geometry of the crack in Fig. 10 supports that the crack at that axial position has propagated inwards. Examinations of the propagation of cracks has been performed withi EPRn na I project [3]presentee Th . d results illustrat axiae eth l variatio crackse e th th t f no bu , author EPRe th f I so t conclud reporno o d t e whethe cracke rth s start fro innee m th outer ro r surfacf eo the cladding.

FIG. Optical9. photo ofrodlet, showing irregularity axialin crack

diametee Th r increase estimatee b n sca d with reasonable accuracy openine fro size th mf th eo g crackse oth f difference Th . e between measured diameter increas diameted ean r increase calculated

37 fro e cracmth k width correspond unifora o st m strai crose f 0,1-0,no Th s. sectio6%' Figurn ni 0 1 e shows thinner claddin positioe th cracke t gth a causef e no .b Thiy dsma eithe locay rb l deformatior no by thinnin wearo t e g. mechanisdu a claddin e Wea e th b f y ro glocalisinr ma mfo g axial cracke on o st part of the circumference of the cladding.

FIG 10. Metallography; Transverse cut through axial crack in the cladding

The content of hydrogen can influence the mechanical properties of stainless steel. The compositio s measurewa amound s nan ga d f o insidt intace on e s tpressur rodletga e t rooTh a e. m temperature was 1.4 bar and 86 % of the gas was hydrogen. It seems likely that hydrogen diffuses fairly quickly through the cladding material at reactor temperature and that the measured hydrogen content correspond equilibriun a o st m pressure wit reactoe hth r coolant.

The correlation between the axial extent of the cracks and the diameter increase is weak. Among the rodlets examined in 1997, there were several rodlets where the axial cracks extended downwards below the Ag-Cd-In absorber. The axial cracks were in several of these cases divided into brancheo tw lowee th (Figured t increasee 12)sa d en r an Th . diametee 1 th 1 s f so r were also largen i r these cases and it is evident that the resistance of the cladding will be smaller in such cases The fast e absorbeth f o p rswellin ti result e th t sga more directly int odiametea r increase with less plastic deformation of the absorber. The branching of the crack will consequently result in a larger diameter increas rodlete th f reactor eo spe r cycle.

5 CORROSION OF ABSORBER

One cross section of a defect rodlet was examined. The material loss from the absorber was small in the studied cross section. There was some material loss close to the crack in the cladding and some intergranular corrosio surface absorbee th th t nf a o e r (Figure 10) microprobA . e examination showed that Sn, which is formed by transmutation from In, is inhomogeneously distributed in the probabls i materia n S d yan l concentrate graio dt n boundarie absorbere th n si .

6. DISCUSSION

From the examinations it is concluded that the diameter increase of control rodlets without cracks is very small. The swelling of the absorber results in cracks in the cladding starting about 40-

calculatioe Th 1 n assumes tha l deviational t s fro maximue mth m weao t diametetha d maximue e ran th t du e rar m diamete unaffectes ri wearsmalley e db b straie y Th . nrma than calculate ovahsatioe th f di mainls ni y causey db deformatio weary b t .no d nan

38 rodletse diametee th fro f ratth e m lowe5 e o f m0m eth o .Th d ren r increas probabls ei y determiney db the average swelling along some cm length of the absorber. If the irradiation of the rodlet continues there is some danger that the crack propagates below the lower end of the absorber and divides into brancheso tw resistance Th . claddine th f eo g against absorber swelling will the muce nb h smalled ran continuee th d diameter increas probabls ei y determine averagn a y db e swelling rat onlf eo y somm em of length at the tip of the absorber.

1 ;: ;US3Sii«»'»>rss.?;»i£»ft^.ft^Ss^^*^^*^ *^.* * - '••"» "~'•'••'"~ " ':«& £. f

FIG. Axial11. crack with branches lowerthe end in

FIG. 12. Axial crack with branches in the lower end

REFERENCES

[1] DE PERTHUIS, S., RCCA's Life Limiting Phenomena: Causes and Remedies, (Proc. Technical Committee Meetin Advancen go Contron i s l Material r Watefo s r Reactors, Vienna Nov9 Dec2 ,2 o t . . 1993), IAEA-TECDOC-813, Vienna (1995) 61-78. ] [2 MATSUOKA t al.e , Nucl. ,J T. , . Sci. Technol. , (Augus,35 t 1998) 564-578. [3] SIPUSH, P. J., et al., Lifetime of PWR Silver-Indmm-Cadmium Control Rods, NP- 4512, (March 1986).

NEXT PAGE(S) left BLANK 39 STATUS OF CONTROL ASSEMBLY MATERIALS IN INDIAN WATER REACTORS

V.G. DATE, P.O. KULKARNI Bhabha Atomic Research Centre, Trombay, Mumbai, India

ABSTRACT

India's present operating water cooled power reactors comprise boiling water reactor f Tarapuo s r Atomic Power Station (TAPS pressurized an ) d heavy water reactors (PHWRs t Kota ) a (RAPS), Kalpakkam (MAPS), Narora (NAPSd an ) Kakrapara (KAPS). Boiling water reactor TAPf so boroe Sus n carbide control blade contror sfo powef o l shur wels fo ra t s a l down (scram). PHWRs use boron steel and cobalt absorber rods for power control and Cd sandwiched shut off rods (primary shut down system) and liquid poison rods (secondary shut down system) for shut down. In TAPS, Gadolinium rods (burnable poison rods) are also incorporated in fuel assembly for flux flattening. Boron carbide control blades and Gadolinium rods for TAPS, cobalt absorber rods and shut down assemblies for PHWRs are fabricated indigenously. Considerable development worr evolvins carriefo wa kt ou dg material specifications, componen assembl& t y drawings d fabricatioan , n processes. Detail variouf so s contro shud assemblief an ltof s being fabricated currentl highlightee yar papere th n di .

1. INTRODUCTION

Reactor control systems and reactor shut down systems are the most important engineering safety systems of any reactor. The first man made reactor CP-1, in 1942, had the manual shut down system called SCRAM-Safety Contro Accelerated Ro l d Movemen proteco t t- reacto e th t r from over power [1]. The 1st generation of pressurized heavy water power reactors (PHWRs) use moderator level control and moderator dumping for control and shut down of reactor. Since then lot of technological developments have taken plac improvo et reactoe eth r contro shud an lt down systems. New and improved designs, materials, fabrication techniques, advancements in reliability analysis, instrumentation and control all have contributed to the present reliable and safe control and shut down systems in various power reactors.

The current meeting is aimed at reviewing mainly the material aspects of control and shut down assemblies used in water-cooled power reactors and the behaviour of the materials under normal operating conditions attempn A . followinmade s i t th n ei g paragraph preseno st statucontrota e th f so l assemblies use Indian di n water cooled reactors with respec materialo t s used, fabrication techniques use assemblien di operationad san l experience e absorbeTh . r materials use thesn di e assembliee sar B4C, Cd sandwiched tubes, boron steel, Co absorbers and gadolinium rods.

2.1. Control blades use Tarapun di r Atomic Power Station (TAPS)

Tarapur Atomic Power Station (TAPS) consists of two General Electric (GE) BWR units, oldese th f o t operatine whicon worlde e hth ar n gi . TAPS went into commercial operatio 196n ni d 9an r completefa hao s s 9 year2 d f safd economio s an e c commercial operation loggin 8 year5 g f o s valuable reactor operation experience.

Manoeuvring of the reactor power and reactor shut down is carried out by inserting/withdrawing control blades from reactor core with the help of control rod drive (CRD) mechanism. contro9 Ther6 e ear l blades, whic stile har l operative TAPe eacn th i , f Sho reactorl Al . control blades ente core th r e fro bottome mth . Botto e followinth m s e controentrha th gf d yo ro l advantages: (a) It provides axial flux shaping for greater fuel economy; t driveinterfere no Th o sd e) wit(b h off-line refuelling operatio whicn ni reactoe hth re vesseb o t s i l removed for the access of the core.

41 2.2. Descriptio controf no l blade

Contro llon bladM g4 s assemblei consistd yan sheathef so d cruciform arra f stainlesyo s steel tubes filled with Boron carbide (BC) powder. The main structural members of the blade (Fig.l) castingd en o f 304stw o consise Lth gradf o t e stainles sincorporatinp steelto e ;th e ghandla th d ean 4 bottom incorporating drive connecting hub, a central hollow cruciform bar and 4 U shaped tube sheaths. There are 17 poison rods containing B4C powder, in each wing of the cruciform and they are encase shapeU n i d d stainless steecastingd e centra en lweldin th G e sheaths TI o th t s lf e go Th . cruciform bar and the spot welding of the U sheaths onto the arms of the cruciform bar provide a rigid structure to the blade. Spherical rollers are provided in the top castings, for movement of the blade in between fuel channel disd san c shaped roller providee sar botto e th n di m castin r movemengfo t inside the guide tube.

A shaft passes through the central hole in cruciform bar to the bottom of which a spring loaded lock plug is attached. By means of the handle in the top casting the lock plug can be operated from the top of the reactor for latching and delatching of the blade from the control blade drive.

FUEL. CHANNEL

GRID

HANDLE POISON RODS TOP CASTING

CENTRAL CRUCIFORM WITH HOLE

BOTTOM GRID PLATE. CASTING BOTTOM ROLLER GUIDE TUBE SOCKET COUPLING A CONTRO VIEF O W L BLAD SPUN I ED TAPS REACTOR

FIG. 1. View of a of control blade in TAPS reactor

42 2.2.7. Poison rod

Each smal a poiso s i ) stainles ld diametern ro WT 0.6 x m L 380 sx 3m stee m D 0m O (4. l m 7m 304 L grade tube containing boron carbide powder and welded with end plugs at both ends. Fig. 2 show typicasa l poison rod. Boron carbide powde packes i requirere f th o o dt % d75 densito t 5 y(6 T.D) by vibro compaction. The powder is confined longitudinally into several independent compartment crimpiny sb g over steel balls located insid tubee eth t regulasa r intervals purpose Th . f eo this compartmentalizatio avoio t densificatios e ni dth botto n vibrationi o t C e 4 mB du f no operation si n and deceleratioy avoidinb p to o gt forcinC n4 forcesB f go . Fig .give3 s flow-shee fabricatioe th r fo t n of poison rods.

r- FIRSrir\o PLUD TiEN G C.INr-POISOU ruuo N rTUB— E ^COMPARTMENTAL BALL

MV/ 11(f 11 \jjjjjiiii j r; j 7 j jj \i it i,t jjj ieszzzai——\ (————vi.^>y.;.cgzasaa t xtjjf >

LrPiMCRIMPP V-R.r\-B4C. POWDEPOWDFR SECONSFfiONl PI PLUH D DEN FN G

Poison2. FIG. rod

2.3. Developmen materialf to fabricatios& n technology

Considerable developmen e indigenouth r fo t n effor i d bee st ha fabricatiotpu n f controo n l blades. The most important task was to generate detailed specifications and fabrication drawings for componentse th . Extensive development wor alss kha o been r standardizincarriefo t dou g fabrication parameters like vibro compactio weldinG plugd TI Bf n4 en Co Argo n , f si g o Heliumn& , spot welding f sheatho assembld san f bladeyo . Special fixtures were designe fabricated fillindC an 4 d B r gan fo d welding of assembly. Fig. 4 shows a flow-sheet for the fabrication of control blade. The prototype blades were tested successfully in out-of-pile tests and then loaded in reactor. Subsequently about 35 blades have been made so far towards requirements of replacement of blades in TAPS reactors.

2.3.7. Boron Carbide Powder

Natural Boron Carbide powde bladese uses ri th maie n di Th .n specificatio powdee th r e nfo r ar given in Table I. The boron carbide powder is prepared by reacting boron anhydride with carbon at high temperatur graphita n i e e resistor furnace [2]. Boron carbid s formei e s densa d e crystalline chunks which are subsequently ground to powder and graded into different size fractions. A special leaching treatmen powdee gives i t th o n t contro o t r 620e moisturd th l 3an e content within specified limits powdee sizTh . 3 f ero e mesfraction 5 0 -20th d 8 mesh32 n h0 i + an 0 2 + 0 + s-6 , 6 viz-1 .- proportion of 40:20:40 gives the required tap density of 1.7 gm/cc. Fig. 5 gives a sketch of graphite resistor furnace used for production of Boron carbide and Fig. 6 gives flow sheet for the production of Boron Carbide s capabli C 4 f soakinB .eo g large volume f hydrogeo s n releasee whicb n hca d upon irradiation therefors i t I . e pruden preveno t t t absorptio By powdeb 4C C4 2 C B .H 0 f dries no i r20 t da beforh fo 2 r e filling int fille C beforopluh d 4 tubeB 8 en d gr d I etubeI fo san degassee C sar 0 20 t da welding.

2.3.2. Other materials controln i blades

The majority of components of the blade except the rollers and pins in top and bottom castings f stainlesmade o t ar eou s steel grade 304L gradL . uses ei avoio dt d sensitization during weldingr Fo . rollers r stellite-fa o s , 3r pin allo fo uses sd Hayneywa dan s alloy-2 proposew no s used s i 5wa t o dI .t

43 TABLE I. MATERIALS SPECIFICATION FOR BORON CARBIDE POWDER

1. Chemical analyses boron 76.5% minimum boric anhydride 0.1% maximum iron 0.8% maximum titanium maximu% 1 . 0 m total chlorin fluorind ean e maximu10m 0pp m cobalt 50 ppm maximum copper maximum pp 0 10 m manganese 100 ppm maximum sodium maximu10m 0pp m carbon remainder 2. Boron isotopic content the boron naturaf shalo e b l l isotopic content0 1 B % , . i.eat . 3 19.0. 6± 3. Particle density the material shall have an average particle density of not less than 2.38 g/cm 3theoretica e (95th f %o l density)

chang inconeo et l X-750 allo rollerr yfo condition0 PH-13. d 10 san l alloo H- pins 8r M n yi fo , vien i , w hige oth f h radioactivit Co-6o t ye 0buildu whilp du e baso usinC ef go stellit Haynee& s alloys. Similar replacements have been incorporte [3]E G . y db

2.4. Contro drivd ro l e mechanism

basie Th c drive mechanis doubla ms i e acting mechanically latched, hydraulic cylinder using reactor feed water as operating fluid. The use of reactor feed water as operating fluid eliminates the need for special hydraulic system. The drive uses simple piston seals with leakage into the reactor

SWAGED,MACHINED INSPECTED AN D S.S.304LTUBES

CLEANING 1 MACHINED I-END PLUG WELDING INSPECTED AND • BY TIG PROCESS CLEANED 304PLUGD EN L S

INSPECTION (VISUAL, GLYCOL, RADIOGRAPHY)

VI8ROFILLINF GO BORON CARBIDE POWDER i CLEANIN TUBF D GO EEN

D PLUEN G1 3 WELDINE TH N GI HC WELDING CHAMBER

INSPECTION (RADIOGRAPHY, GLYCOL,H6LEAKJ

FIG. Flow3. sheet fabricationfor of poisonrod

44 PLUD 1sEN tG WELDING POISOF O N TUBES

LOA C POWDEC D RN I 4 CENTRAL CASTINGS POI'ON TUBES U SH2ATHS I CRUCIFORM 2nd END PLUG BAR WELDING I POISON RODS LOAD 17 POISON n WELDIHTi F U G 30DS.N FAC'1 CRUCtrOR'lO T l CASTINGS

SPOT WELDING AND-TIG-WELDING — OF LOADE SHEATHDU S

INSERTION OF CENTRAL SHAFT 1 FIXING HANDL SHAFN O E T I FIXING VALVE, LOCK PLUNUTD AK N GSO SHAFT AND VALVE SEAL IN BOTTOM CASTING I .MOUNTIN "SOCKEF O G D TAN LOADIN F SPRINGO G

CHECK LATCHING, DELATCHING OPERATION

TACK WELD VALVE, LOCK PLUG D NUT SHAFN AN O S T

DEBUR D INSPECAN R T

FIG. 4. Flow sheet for control blade fabrication

GRAPHITE CORE GRAPHITE LINING CERAMIC INSULATION TERMINAL BLOCK

*;// i-*'^,^ii^i^!r''ilii7'in'V--^-v. /;V-£tfii «'l|iir^"«'"«Mi|liV-^: J^'Sll...! I.II«l|j,,|J>*,|l||i,,| I'AXl^J

WATERCOOLE* D JACKET' ' MILD STEEL SHELL

POWER SUPPLY

FIG. 5. Graphite resistance furnace used for the production of boron carbide

45 UNCONVERTED CHARGE FROM f BO PREVIOU N SRU ?ON PETROLEUM CAR BIDE WEIGHING I /~\j COKE POWDER J ^ * L ' -H-.| FURNACE -jl\ . —' | = MIXEF R 3 T Td_ BORIC ACID GRINDING, WEIGHING r ! i Pnwpp ^MP PI v -

MAGNETIC SEPARATION

CLASSIFIER

GRADED

CHEMICAL TREATMENT

FIG. 6. Flow sheet for the production of boron carbide

minimizing contamination and helping to cool the mechanism. The drive is capable of inserting and withdrawing the control rod at a slow, controlled rate as well as providing rapid insertion in an emergency condition. The hydraulic fluid to control rod drive system for its rapid insertion on a scram signa supplies i l d from three redundant sources: storee Th d energ accumulatore th n yi ; High pressur hydrauliD eCR c systemd ;an Reactor pressure itself.

With this arrangement, during total loss of power (TLOP) to the station all the 69 control blade t insertesge d int reactoe oth r core withi requiree nth d tim storey eb d energ accumulatorse th n yi , failing reactowhice th y hb r pressure itself. Contro drived ro l replacee sar followinn di g conditions: After 10 years of service; Whe drive nth havins ei g persistent high temperature problem; Whe t movindrive nno th s e i applicatio n go normaf no l hydraulic pressure.

2.5. Operational experience with boron carbide control blades

Right from the commissioning of TAPS units in 1969 till 1996 TAPS Unit I has experienced 152 scram experiences unid ha san 2 t scrams9 d10 , becaus f systeeo m transient losd f powessan o o rt the station s seewa n t I tha. reactoe th t r coul broughe db safo t t e shut down sub-critical stat fasy eb t acting shut down mechanism alone, without necessitating slow acting liquid poison syste mactuato t e r coolinFo [4]. g contro d drivro l e seals continuous cooling wate f higheso r t purit s maintainei y d through mini-filters of 25 microns with 99% efficiency. This is to prevent damage of seals due to high background temperatures. During the operation of the reactor, control rod high temperature, problem

46 appears to be due to clogging of the mini-filters. Rectification of high temperature problems is carried out by isolating the drive and replacing the mini-filters. Clogging of the mini-filters also poses problem controllen si d movemen controe th f o t l blade noteworthe b sy coree in/ouma th t f I .o ty that syste D desige th CR m f no doe jeopardizt sno fase eth t acting reactor shut down mechanism function even with a clogged mini-filters.

During the operation of last 28 years no failures have been observed with regard to materials of control blades. Problems suc swellins ha g under irradiatio reactiod nan n with stainless stee t higa l h temperatures have been reporte boror dfo n carbid literaturn ei e [5] boroo N . n contaminatio bees nha n observe waten di r samples take t regulana r intervals confirming tha tgroso thern e s ear defect y an n i s poisoe th f o n rods.

No PIE has been carried out so far for any of the control blades removed from reactors after completing design life which is based on 10% reduction in control worth or 34% reduction in boron content. 3.0. CONTROL & SHUT DOWN ASSEMBLIES FOR PHWRs

3.1. Control assemblies

Indi operatins aha g pressurised heavy water power reactor Mw(es0 eac22 f h)o capacit Kott ya a (RAPS), Kalpakkam (MAPS), Narora (NAPS Kakrapar)& a (KAPS).

Basically for the control of reactivity in PHWRs three types of devices are provided (Fig. 7). Regulating rods (RRs) or central adjusters for adding and removing reactivity quickly (± P), the absorber rods or corner adjusters (ARs) which are provided to add positive reactivity (+P) for

LEGEND: O A-ABSORBER ROD NOZZL. E E-N VERTICAL PLANE O FM-FLUX MONITOR NOZZLE jrjni| ^077P L_ n R" a ^H'I1F F Tn^ M-MECHANir.AIO N077D , RO F SHUI F OF T 14 O R-RFfiUl.flTINR ROD N077I F . 2 O ,SH- SHIM ROD N07ZI F 2

M W Rl pb $J ii >jw 1 457.2 A t fl^Mk ^ FM 1 J<1 rM | { N-S VERTICAL q_ PLANE •*, i «< ^JUJ5_a.a 228.6 914.4 >* i?* «£ -< 3>511^

200.0 TTYK) 405.0 (TYP) 600.O (TYP.) BIOO 100.0

U6O.5

LocationFIG.7. of reactivity control devices PHWRin

47 providing xenon override followin compensato t triga r po daile eth y burn-un pcase no loath f eo n di availability f fuellino g machin d shian em rods (SRs r providin)fo g backu o regulatinpt g rodn i s bringing negative reactivity (-P) for power set back or step back [6]. In RAPS & MAPS the set back action is done by moderator level control while in other PHWRs it is done by shim rods (SRs). Normall kepe AUTn ar o t keps e d partiallar Oy an RR t s ) controlfull (abouIN AR N N yI I e % Th .80 t on manual control wit hmechanicaa l stopper called ratchet which prohibits e withdrawalar s SR e Th . always kept on auto control outside the core. If they enter the core for any set back action for more than 10 seconds they are pushed back by addition of about 1 mk equivalent of boron added by the chemical boron addition system [Control Addition Mode (CAM) system]. All these control devices are fully inserte triy pan signaln do .

PHWRe MW firs e 0 sth I tn50 unit whicf so comine har t Tarapur a p gu three th , e control actions are respectively done by the Light water zonal power control system consisting of 14 zonal control compartments (ZCCs) adjuste7 1 , r rod foud san r control rods (CRs).

3.1.1. Control assembly materials fabrication&

3.1.1.0. Absorber rod corner so r adjuster rods

Cobalt assemblies are used as absorber rods at corner locations of the regulating system of PHWRs. However stainless steel assemblies can also be used at these locations. Eight absorber assemblies corne4 e eac2 th , rn hi positions , provid ereactivitk aboum 1 1 t y incremen overrido t t e xenon for approximately 30 minutes after shut down from full power equilibrium operation of the reactor. Another purpose of the cobalt absorbers is to produce Co-60 isotope in the form of Co slugs of specific activit curies/gra0 y5 fore th pelletf mo n i r mspecifif o s o c activit curies/gram0 y20 .

Each absorber assembly assemblieb consistsu 4 f so s mounte centraa n do l support tube with retention support assembl t topya . Therelement6 e ear r sluassemblb pe s g su element8 d yan r pe s pelle b assemblysu t . Each element contain cobal5 1 s t 3 pelleslu r o g t capsules encapsulaten i d aluminium.

3.1.1.1. Encapsulation

Considering the end use and design requirements, Co slugs and pellets are to be encapsulated with Al as cladding. The stringent specifications stipulated for slug/pellet capsules are: i. The capsule should be made as per close dimensions; ii. Seal welds shoul defece db t free; iii. There shoul e intimatdb e mechanical contact betwee cobale nth e claddintth slud gan g material; spacpellee e th Th n ivei t .capsule s shoul fillee db d wit gase hH . Specification cobalr fo s & t aluminium to be used in absorber assemblies are given in Table II.

After several trials encapsulation routes were standardised to get acceptable slug/pellet capsules maie Th .n step fabrication si thesf no e assemblie givee sar Figuren . i 8 & s7

3.7.7.2. Cobalt elements

A slug element contains 15 Co slug capsules and pellet element contains 3 pellet capsules clad in zircalloy-2 tubes e capsule-elemenTh . t inter-space contain e fabricatio th e gas H r so .C Fo f o n elements, resistance welding technique under He cover gas proved successful.

After fabricating the elements, the elements were autoclaved in steam at about 400°C and 53 Kg/cm2 pr. for 10 hrs. Radiographic examination of these elements revealed collapsing of pellet capsules and the analysis of fill gas showed hydrogen as the major content in pellet capsules and also in element/capsule inter-space.

48 TABLE II. SPECIFICATIONS FOR COBALT AND ALUMINIUM USE COBALDN I T ABSORBER ASSEMBLIES

Cobalt Cobalt material for the manufacture of slugs and pellets shall be of a high quality suitable for irradiation Chemical composition shall be as follows 99.8% minimum cobalt 0.15 maximum nicke iro& l n 0.01 maximum other impurities 0.15 maximum gases ______Material density shall be______8.4 g/cc minimum Aluminium The aluminium alloy material to be used for the encapsulation of cobalt pellets and slugs shall be indal-is or equivalent having the following mechanical properties minimua Mp 3 6 m U.T.S. minimu% 25 m ) elongatio% mm 0 5 n n(i Material shall be aluminium 99.5% minimum copper 0.05% maximum iron 0.4% maximum silicium 0.3% maximum ______manganese______0.05% maximum

After close examinatio fabricatioe th l al f no n procedure elaboratd san e checkin s founwa t dgi releases wa 2 H d thae froth t i platemN pelleto dC s during autoclavin elemente pellee th th f go d t san capsules collapse pressuro t e elementde du eth n builresuli a p s sdifferentiau ta f o t - expansiol l A f no zircaloy-& 2 during autoclaving eliminates . Releaswa ^ F f bakiny eo db cobale gth t pellet vacuun si m and the collapsing of pellet capsules was eliminated by increasing the free space in the elements.

3.1.1.3. Co assemblies

e desigTh f assemblo n s finalisewa y d after severa lf pilo trials t e ou testing, d alsan o, performance trials in the reactor. The initial design, caused a severe spinning effect resulting in element rotation and subsequent severe wear/fretting - wear on end plugs, spacers, anti torque pins and central suppor t desigw e tubess satisfactorne founb wa no A t .d y which incorporated modifications like: Improved latch body; Provisio holef no centran si l support tube; Tapered end spacers; Provisio sprinf nbottoo e th t g a mgiv o t compressivea sub-assemblye th Ibsen 0 o forc. 10 f eo ; Anti torque pin differenf so t designs.

3.1.2. Regulating rods and Shim rods

There are two regulating rods made in stainless steel 304 grade and two shim rods made of borated stainless stee r PHWpe l R reactor. Borated stainless steel rods containing abou boro% 2 te nar . ID m m 0 3 hollo d an wD rodO sm witm 0 h6

3.2. Operational experience of control assemblies of PHWR

e adjusteTh r rod f RAPo s S offered trouble free operation till 198e centrath 6 f o whel e non adjuster rod got struck due to ball screw failure. During the rehabilitation of adjuster rods that had failed during 1986, it was noticed that the coolant flow provided to the adjusters is inadequate and

49 consequently these mechanis e runninmar g e drybalTh l. screws thus were deprive f wateo d r lubrication. The failure of the ball screws was attributed to lack of lubrication.

l ENCAPSULATIOA o SLUGC F NO S I SPINNING

PRESSING I DRILLING I WELDING

D PRESSIN2N G

MACHININ SLUGF GO S

FINISHED SLUGS

LOADING IN Zr-2 TUBES AND RESISTANCE WELDING 1 ASSEMBLY INTO BUNDLES

ASSEMBL BUNDLEF YO N SO THE EB WELDED CENTRAL SUPPORT TUBE ASSEMBLY

FINISHED Co SLUG ASSEMBLY

FIG. 8. Steps in fabrication of Co slug absorber assembly

A typical problem of NAPS moderator system is a high Co-60 activity coming from adjuster rod ball screws [7]. The stellite-3 balls of the ball screw assemblies have been found heavily eroded. The e ballCo-5th f s o 9e reactotrave th o t lr core with coolan getd an st activate d leadan do t s contaminat equipmene eth pipind an tincreased gan d activit entirf yo e moderator system.

3.3. Shut down assemblies

firse Th t four unit t RAPsa MAPS& S have onlshue yon t down system (SDs) viz. moderator dump system engineeres i t I . havo dt e high reliabilitlarga s eha deptt i y& negativf ho e reactivityt Bu . it has a slow initial insertion rate of negative reactivity. Further mechanically the dump action puts a lot of stress on the core components. More importantly the moderator which provides a large heat sink would be absent after shut down action. These lead to the concept of two independently fast acting

50 systems in the standard PHWRs designs from NAPS onwards. These two independently operating fast acting systems are ~ primary shutdown system (PSS) and secondary shut down (SSS). PSS consists of 14 stainless steel clad cadmium shut off rods, which get inserted from top of the core in about 2.2 consistS zircalo 2 SS worta 1 . seconds f f o sabouha h o Mk y d 5 empt3 tan s y tubes through which Lithium pentaborate solution is inserted from the bottom of the reactor under the action of high pressure helium gas. It is grouped into 4 banks having a total worth of about 32 Mk.

The 500 MWe PHWRS have two independent SDS in line with the CANDU philosophy. The first SDS, SDS-1 consist of 28 stainless steel clad cadmium shut off rods having a total worth of about 72 Mk. The second system, SDS-2 consists of 6 poison injection nozzles which inject curtains of gadolinium nitrate solution under high pressure gas action directly into the moderator providing greater than 300 Mk of negative reactivity.

3.3.1. Materials and fabrication of shut down and assemblies

e primarTh y shut down syste f Indiamo n PHWRs, from NAPS onwards s alsi , o calles a d mechanica systed l shuro f mof t (MSRs) systee .Th m consist shu 4 rodf 1 of tf so locate locatio4 1 t da n reactoe th f o r p ovessento arrangemenle (FigTh . .7) t locatioa f o t n consist verticaa f so l tubulad C r shut off element sandwiched between austenitic stainless steel tubes. The element is hung from the drive druth men i mechanis typm6 wit31 e hel e a austeniti hf th p o c stainless steel wire rope rope Th .e from the drum runs to pulleys to locate the shut off element centrally at its site. The schematic sketch movemen d gives i typicaro a shu e f f FigR n no i Th of t . .MS l9 reacto n i t guides i r zircaloy-a y db 2 guide tube (Fig. 10)thermae Th . l neutron absorbing materia shue f elementh of treactoe n i lth s i t r grade cadmium. The cadmium is contained between two co-axial stainless steel tubes. (Fig. 11).

AITIIBI'.S I S

) PLUKM I G RESISTANCE WELDING * VIBROPACK1N PELLETo C F Gl TUBEO A N SI S

PLACE FOIL & WARM AT 120°F FOR 30 MINUTES 4 II END PLUG RESISTANCE WELDING UNDER He

MACHIN .PES EA R DRAWING

RANDOM RADIOGRAPHY

He LEAK TEST

ACCEPTABLE PELLET CAPSULES

LOADING CAPSULES INTO ONE END WELDED Zr2 TUBES

II h'.ND PLUG Ri:siSIAI\CK\VKI.I)IN(;UNI)i:KUcOI''/,r2TUlU:S

AUTOCLAVIING ATf>()0°F& INSPECTION

FINISHED Co PELLET ELEMENT

ASSEMBLY INTO BUNDLES (SUBASSEMBLY)

ASSEMBL BUNDLEF YO CENTRAN SO L SUPPORT TUBE

FINISHE PELLEo DC T ASSEMBLY

FIG. 9. Steps in fabrication of Co pellet absorber assembly

51 SHU ELEMENTF O T .

COMPRESSION SPRING

FIG. 10. Mechanical shut-off rod for PHWR

52 P EXTENSIOTO N

WELD

SHUT-OFF ELEMENT

WELD W3

BOTTOM SPRINP TO G

3760mm PROTECTIVE CAP 13

WELD W2

BOTTOM PLUNGEP

WELD W5

BOTTOM SPRING HOUSING

WELM W D

Cd-SANDWiTCHED BOTTOM PLUNGEP HEAD TUBE ASSEMBLY

sandwichedCd FIG. 11. tube assembly

53 Cadmium sandwiched with stainless steechosewas l neutroas n n absorbing elemenhigto h due t neutron absorption cross section, ease of fabrication and low vapour pressure. Sandwiching is required for augmenting strength. Cd sandwiched absorber element is further welded to other SS components bottos ha R m sprinMS . g which absorbs kinetic energy whe touched nro s "stop platef "o guide tube. Upper end of MSR has wire rope fixing arrangement. Fabrication of zircaloy-2 guide tube assembl MSRr y fo (Fig) s12 .involve s intricate machinin weldinB E d gcomponentf an g o meter6 s5- s long fabricatioe Th . MSRf no s involve weldinG sTI componentf go s using special fixtures.

3.4. Operational experience of shut off rod assemblies in PHWRs

performance Th shut-off eo assemblied fro reportes si smoote b o goodd t han d till date, as far as shut off elements are concerned. Some minor problems were faced during commissioning due to mechanical design deficiencies such as

Rope snapped during partial element drop in many mechanisms during commissioning period. Malfunctionin positiof go n indicating micro-switches during commissionin operationsg& . leakage H e fro flange mth e jointstane th do st pipe lower body

TOP SLEEVE GUIDE TUBE STOP PLATE BOTTOM SLE EVE

FIG. Zircaloy-212. guide tube assembly PHWRfor

4.0. BURNABLE ABSORBER RODS IN TAPS

4.1. Gadolinium rods used as burnable poison rods in TAPS

1.5% Gd2O addes i 3 UOo dt 2( 2.66% enriched thesd )an e UO2- 1.5% Gd2O3 pellet usee sar n di 2 rods of the 6 x 6 fuel bundles of TAPS . These burnable poison rods help in reducing flux peaking and in obtaining flux flattening in fuel assemblies in the core.

e generaTh l processing scheme preparatioth 2 r —1.5fo e %UO f no Gd2O3 pellet s givei s n i n Fig. 13. Gadolinium rods are made and used regularly in the fuel bundles manufactured by NFC (Nuclear Fuel Complex), Hyderabad for TAPS. Operational experience of TAPS shows that gadolinium rods are giving the desired performance and no problems have been faced.

5.0. OTHER MATERIALS UNDER DEVELOPMENT

Development wor plannes ki controtaken e o b p materialo d nt u ro l s like Cr-Hf borides, ceramicsmixture& C 4 B f so , Gadolinium compounds vien i , f theiwo r potential application waten si r reactors.

54 U02

BLENDING I PRECOMPACTION I GRANULATION

FREE FLOWING GRANULES I FINAL COMPACTION i GREEN PELLETS OF 50% THEORETICAL DENSITY I SINTERIN 3 HOUR2- R 162T SGA FO 0C

SINTERED PELLET 93-96F SO % THEORETICAL DENSITY I CENTRELESS GRINDIN PELLETF GO R SFO CORRECT DIAMETER & SURFACE FINISH

FlowFIG.13. sheet UOfor 2-Gd2O3 pellets

6.0. CONCLUSIONS

e indigenouTh s capabilities have enable requiremene cateo th t s o t dru f controo t shud an lt down assemblies of the operating water cooled power reactors. The performance of the assemblies in the reactors has been reported to be satisfactory. However, this has to be substantiated by poolside dischargew fe n inspectioo E AHW d setting-ufue& d PI X assembliesd an R l nan RMO PW f o f po e .Us reactors envisaged in our programme will require control assemblies with improved design, materials, fabrication and hence improved performance. Efforts are hence required to be put in this direction.

REFERENCES

] [1 Reactor shutdown system (Proc. Workshop, Kalpakkam, 1997), IGCAR, Kalpakkam. ] [2 BOSE, D.K. t al.e , , Productio f higno h purity boron carbide high temperature material& s processes , (19863 ,& Vol.72 ) s 133-140,No . [3] BWR Control Rod Cobalt Alloy Replacement, EPRINP-2329, (March 1982). [4] BHASIN, B.K., et al., "Control & shutdown mechanism, operating experience of TAPS" (VI. 1.1., Proc. Workshop on reactor shutdown system, Kalpakkam, 1997) IGCAR, Kalpakkam. ] [5 BART , "HigG. , h temperature boron carbide interaction with stainless stee contron i l l rods" (ProcAdvanceM TC . contron i s l assembly material r watefo s r reactors, Vienna c 1993De , ) IAEA. TECDOC-813, Vienna, (1995) 177-186. [6] SRTVENKATESAN, R., et al., "Physics requirements of shutdown systems for Indian PHWR" (1.2.1, Proc. Workshop on Reactor shutdown system, Kalpakkam, 1996) IGCAR, Kalpakkam. ] [7 JAIN, A.K. t al.,e , "Malfunctio failurd an n regulation i e protectivd nan e system f PHWRsso " (VI.8, Proc. Worksho Reacton po r shutdown system, Kalpakkam, 1997), IGCAR, Kalpakkam.

NEXT PAGE(S) left BLANK ' 55 CALCULATION MODELLING OF THE RCCA MOVEMENT XA0053630 THROUGH BOWE GUIDA DF E TUBES

D.V. RAZOUMOVSKY, Yu.I. LIHKACHEV, V.M. TROYANOV State Research Centre IPPE, Atomic Energy Ministry, Obninsk, Kaluga Region, Russian Federation

Abstract

Rod control cluster assembly movement throug bowee hth d guide tube considereds i s e movemenTh . t equatios i n presented with assumptione somth f eo speciad san l attentio determinatioe pais th i n o dt mechanicae th f no l friction force. The numerical algorithm is described and some results of parametric studies are presented.

1. INTRODUCTION

numbee Th f operationao r technologicad an l l factors influence coverlesa n o s s fuel assembly (FA resultd deflectioA an )F n si core a n i n . Thi witP s hNP fac s e WWERprovei ttru e th b t ea o dt - type reactors. Sometimes this phenomenon produces Safety and Control System (SCS) malfunction, t suitabl operationP safe thano th s e NP i tr same efo facte th f eTh . o skin reportee dar numbea t da f o r Francand PWReUS desigof n [1,2].

Both required time limi tcontro d excesro d lsan cluster assembly (RCCA) incomplete insertion classyfiee malfunctionar S SC s da core th en I RCC. typeo tw affecteAs f resistanc i so e th y db e forces. That are the hydrodynamic forces (namely viscous friction on the rod walls, the back flow of the displaced water, the coolant flow, the spider hub flow separation) and the mechanical friction between the guide tubes (GT) walls and rods. The last force is brought into existence after FA deflects at the value of the difference between inner GT calibre and rod diameter. Time limit excess may be accounted for the hydrodynamic forces increase, but it is not true for incomplete RCCA insertion.

The paper [3] deals with the problem. The slide friction force is said to be a principal cause of incomplete insertion. But they suggested simplyfied approach for the force determination, also it was necessary to carry out the rod drop experiment in the concrete core. This paper offers more universal techniqu predico et t RCCA behavio bendeonle A th F ye n basro dth f axiseo .

2. RCCA DROP DYNAMICS MODEL

In a core RCCA is affected by the gravity, the viscous friction, the spider hub resistance and the mechanical friction. The RCCA acceleration is defined by these forces sum:

e RCCth Ms i A mass s graviti g , y constant s viscoui k , s friction coefficient e RCCth s i A x , e spide th mase resistancb e GT)lengt s th th i hu r s f s , i ,o hk centr L e , e L] coordinat , [0 e x ( e coefficient frictioe th s i . n\j , coefficient, N(x normae )th l force betwee. RCCe GT nth d Aan dx ) (2 Th . e0 initia= - , l: condition0 = / ; 0 s = loo x k Q: like = / : at

In the equation (1) is ignored: the pressure reduction due to the coolant flow across the FA, the back flow of the displaced water, the pressuresation of the water ahead of the RCCA and the damping devices affection afte bottoe th r m design positio reacheds ni .

57 Both incomplete RCCA insertion (Xiy=0 wil assumee b l malfunctiona s da failurS . TheSC e nfunctioth followingne wilth e b l :

3. MECHANICAL FRICTION DETERMINATION

The main resistance cause was said to be the friction force, that is proportional to the normal force between RCCA and GT. Let us apply the relationships:

X N(x~)=\\q(x)dx (3) o

4M-EJ&& (4)

q(x) is the rod lateral load, y(x) is the rod axis deflection, E is the Young modulus, J is the moment of inertia for the rod (it is assumed the fixed characteristics along the rod length).

roughls u t Le y suppos abscence eth annulue th f eo s between rod e inne th wal,T d tharG an ls i t equal to the situation when rod is deflected according to the GT form.

For the two-dimensional solution FA axis is described with the (X, Y) set, Y is the deflection in definTo laterathe eLJ. l [0, loa analyticae dthe poinX X, t l representation y*(x) is (y*(X Y) = ) necessary particulas it e genera f Th o . m lr su solutio solutio e equatioe th th s i f no n) a ny(4 pd (xan ) general solution of the congruent homogeneous equation yg(x):

) (5 y(x y= )g (x y+ )p (x)

2 3 4 3 6 yg(x) = C0 + C,x + C2x + C3x ; yp(x) = C4x + C5x + C6x + C7 cos(x) + Cs sin(x)

absenc e momente th forcee th o th t d f e eso an forcef snecessaro Du s i t i e s d acrosth r ro yfo e sth function y(x) to be continuous and to possess its continuous derivation up to the fifth one. At the ends f intervao conditione th l s y'(0 cons)= e y'(Lt lengt A cons)= F addeddivide s t e hu a tar th t l eal Le . adjacene a numbe e relief theth b o n mo y t o df steps o ran t spaceo s , r grids e solutioe Th .th f o n equatio f ) functionwil o representene (5 t b l se e sth y/x)y db , which describ propee axith A t eF a s r step. The function y,(x) must be equal to the two inner point axis deflections and equal to the both step end deflections. Moreover, at the step ends this function must have continuous derivations up to fifte th h one.

Coefficient e equatio th e strictl n ar i s) n(5 y defined fro systee mth f thesmo e conditionse Th . system is solved by the Gaussian method modification. This technique allows to increase the number of steps in order to refine the FA bended axis without complication of the algorithm.

necessars i t I defino yt e lateraeth l load y thao orthogonaq

Text continued on page 63.

58 20 -, 2E-7 -, Y, MM F*0,9-10g 6k

Guide Tube Deflection

Axia Drod pRo l 10 - 1E-7 - Resistance

I T i

2,000 4,000 0 2,000 4.000 5E-1- -T 1 ..... 10.0

X, M

3E-11 5.0 - RCCA movement

Rod Lateral Load displaeement x(t) velocity v(t)

2.5E-11 0.0 • n——i——r i———T Is 2,000 4,000 0.0 0.4 0.8 1.2

FIG. 1. RCCA insertion in "cosine"—shaped FA

20 -^ 2E-6 -, I Y, MM FxO,9-106 kg

4.000

RCCA movement

dispjacement x(t) velocity v(t)

-1E-9

2,000 4,000 0.0 0.4 0.8 59 1.2

FIG. 2. RCCA insertion in "dollar"-shaped FA

59 20 — 4E-6 _, Y. MM FxO,9-106 kg

\ Guide / Tube \ Deflection /' i ' Axial Rod Drop \ 2E-6- 10 - Resistance

2,000 4,000 0 2,000 4,000 4E-9 -, . _ 0 ....4. . x, M

0 - zo.

Rod Lateral Load

•4E-9 0.0 ; 2,000 0 4,000. 0 0.4 0.8 t 1.2

FIG. RCCA3. insertion "bow"~shapedFAin

20 -, 2E-6 -, Y, MM FxO.9-106 kg

Guide Tube Deflection

Axia Drod Ro lp -K Resistance

-20

0 2,000 4,000 0 2,000 4.000 1E-9 -r- •-— 8.0 -,

0 - 4.0 -

RCCA movement

Laterad Ro l Load displzeelfient x(t) velocity v(t)

-1E-9 0.0 -|——i——r

2,000 4,000 0.0 0.4 0.8 61 1.2 FIG. 4. RCCA drop. Viscous & mechanical friction influence

60 1.0 —i F, kg

Resistance Forces

viscouF s— frictio— n 0.8 — ...... p sliding friction

0.6 —

0.4 —

0.2 —

0.0 i———i———i———i———|—

0.0 1.0 2.0 3.0 X, M 4.0

FIG. 5. Mechanical & viscous friction forces comparison

20 -| 2E-6 -, Fxo,9-106kg

4.000 1E-9-,

RCCA movement

displacement x(t) velocity v(t)

-1E-9

2,000 4,000 0.0

FIG. 6. RCCA drop. Hub resistance & friction influence

61 1.0 F, kg

Resistance Forces 0.8 ...... F sliding friction ——— F hub resistance

0.6 —

0.4 —

0.2 —

0.0

0.0 1.0 2.0 3.0 x, M 4.0

resistanceFIG.Hub 7. friction& comparison

20 -, 2E6 -, Y, MM F*0,9-10e kg

Guide Tube1 Deflection

Axial Rod Drop •1E-6 - Resistance

-20

0 2,000 4,000 0 2,000 4.000 1E-9 —,- . _ 0 ....4. .

X, M

0-1 ZO-

ovement

Rod Lateral Load cement x(t)

-IE-9 - • 0 0. I ' ' ' s , t 2,000 4,000 0.0 1.0 2.0 3.0

FIG. 8. RCCA drop. All forces influence

62 1.0 —i F.kg

Resistance Forces

0.8 — F sliding friction viscouF s friction resistancb hu F e F total hydrodynamlcal resistance 0.6 —

0.4 -

0.2 —

0.0

0.0 1.0 2.0 3.0 X, M 4.0

FIG. 9. All forces comparison

4. APPLICATION AND RESULTS

This algorith realises mcomputee wa th n di r code "DORA", thaprograpara a s f i to t m complex "RANDEVU-3" - "TEREMOK" - "DORA", intended for prediction of FA axis bowing and SCS behaviour (drop time and insertion depth is determined) for the WWER-type core at the arbitrary moment of fuel cycle.

A number of parametric results are listed below. The influence of FA form, friction coefficient and hydrodynamic forces on the RCCA drop was investigated. The mass center velocity v(t) and insertion depth was calculated.

The study of FA shape influence on RCCA movement has been carried out on the three basic variant , snamel 3) (Fig , 2 , y.1 so-called "cosine", "dollar "bow"d an "e FigureTh . s represenT G t deflection laterad ro , l load, axia drod ro lp resistance displacemend ro , velocityd an t "bowe Th . " shape mose th seem e t b incline RCCe o st th r Adfo sticking.

In the case study of friction force influence the incomplete insertion arose while the doubling of its value had happened over the initial one.

e investigatioTh hydrodynamie th f no c viscous frictio s alswa on carrie . t Thi(Fig5) dou , s4 . force doe t affec no se RCC th t A movement very much generan I . t becami l e valuabl e seconth n i e d half way, when rods are almost inserted at full length. In comparison, the spider hub resistance force produces the specific velocity profile (Fig. 6, 7). It tends to influence at the beginning, when the RCCA gets its velocity maximum.

Unde influence rth forcel al f hydrodynamie o s th (Fig ) 9 , .8 c RCCA resistanc composes ei f do approximatele th y equal impact f botso h viscou sresistanceb frictiohu beginnine d th nan b n I .hu e gth resistanc viscoue th e affectd sen frictioe sth moren i nt magnitudbu , e exceedoneb hu .e sth

63 REFERENCES

[1] Specialist Meeting on Nuclear Fuel and Control Rods: Operating Experience, Design Evolution and Safety Aspects, Madrid, 5 — 7 November 1996. [2] International Topical Meetin Lighn go t Water Reactor Fuel Performance, Portland, Oregon- 2 , Marc6 h 1997. [3] BJORNKVIST, L., KJEE, E., Application of a semi-empirical rod drop model for studying rod insertion anomalie t Soutsa h Texas Projec Ringhald an t s uni . Internationa4 t l Topical Meeting on Light Water Reactor Fuel Performance, Portland, Oregon 6 Marc2- , h 1997, 81-89.

64 MATERIAL OPERATING BEHAVIOUR OF ABB BWR CONTROL RODS

B REBENSDORFF ABB Atom AB, Vasteras, XA0053631 Sweden

. BARG T Paul Scherrer Institute, Villigen PSI, Switzerland

Abstract controR BW boroe le rodus Th nB s carbidmadAB y eb e (B4C hafniud )an absorbes ma r material withi claddinna g of stainless steel The general behaviour under operation has proven to be very good ABB and many of their control rod customers have performed extensive inspection programs of control rod behaviour However, due to changes in the material properties under fast and thermal neutron irradiation defects may occur m the control rods at high neutron fluences Examinations of irradiated control rod materials have been performed m hot cell laboratories The examinations have revealed the defect mechanism Irradiation Assisted Stress Corrosion Cracking (IASCC) to appear in the stainless steel cladding For IASCC to occur three factors have to act simultaneously Stress, material sensitization and an oxidising environment Stress may be obtained from boron carbide swelling due to irradiation Stainless steel may be sensitized to mtergranular stress corrosion cracking under irradiation Normall e reactoyth rs oxidisin i environmen R e BW Th g a m t presentation focuse n findingo s s fro t celmho l laboratory worcontroR n irradiateo kBW l B rodd studieAB dan s f o s irradiated contro I Apart cellmateriald PS ho t ro lsa e t froth n mi s physical, mechanica microstructurad an l l examinations, isotope analyses were performe describo dt locae eth l isotopic burnu borof po n Consequences (suc possibls ha washC eB4 - undea out f o ) r BWRB operatioAB a , afte n e i occurrencn th r craca f eo s discusseki d base neutron o d n radiographic examination controf so l rods operated with cracks

1 INTRODUCTION

ABB Atom has delivered almost 4000 control rods to Boiling Water Reactors (BWRs) around the world. More than 100 f thes0o e rods have been inspected after varying tim f operatioeo t a d nan varying levels of neutron exposure. Besides the visual pool-side inspections, control rod wings with cracks have been analysed by neutron radiography to reveal the behaviour of irradiated boron carbide behaviour in case of cracking and water intrusion. Irradiated control rods and irradiated control rod material samples have been examined in hot cell laboratories to understand the materials behaviour and properties due to fast and thermal neutron irradiation.

2 ABB BWR CONTROL ROD DESIGN

controR basie BW Th cl B roddesig AB shows si f no Figurn ni Fou1 e r stainless steel plates are forming a cruciform shaped rod. The steel quality initially used from start was AISI 304L SS which has in later manufacturing been replaced by AISI 316L SS. Boron carbide (B4C) powder with a void volume of 30% is used as absorber material together with hafnium.

3 NEUTRON IRRADIATION OF CONTROL ROD MATERIALS

Neutron irradiation is known to the change properties of the materials used in control rods Some of these changes have an influence on the control rod behaviour.

3.1. Stainless steel

Neutrons of energies above 1 MeV cause hardening, increase in yield and ultimate strength, segregatio nstainlese etcth n i . s steels used. Segregatio f chromiuno m fro graie mth n boundarief so the steel means that the material becomes sensitized for stress corrosion attack in the grain boundaries. The impact, for the operational behaviour, of other changes in the stainless steel material structure and properties is under investigation since not yet fully understood.

65 Handle

Guidd epa

Wing

Shaft

Extension

FIG 1. Basic design of the ABB BWR control rod

The environmental assisted cracking occuring in control rods is intergranular and called Irradiation Assisted Stress Corrosion Cracking (IASCC). IASCC needs three simultaneous factors to occur:

1. Stresses - in a control rod, these can be caused by the swelling boron carbide, 2. Material sensitization - chromium depletion in grain boundaries is such a factor caused neutroy b n irradiation, there migh t understoodothere tb ye t sno ; 3 Oxidising environment - IASCC due to chromium depletion in the grain boundaries require oxidisinn sa g environment There seems, however othee b o t r, factor t highea s r irradiation eliminatin requiremene gth oxidisinf o t g conditions. crace Th k mod bees eha n observe severan di l investigation picture th d Figurn sean i fros i e m2 Pos a t Irradiation Examination (PIE) of a ABB BWR control rod at the Paul Scherrer Institute (PSI). The material is Type 304L stainless steel. In Figure 3, a grain boundary STEM analysis (made by Magnox Electric elemenf )o t profiles, fro same mth e material show depletioe sth chromiuf no m from the grain boundary. The neutron fluence was 3 1021 n/cm2 (E>lMeV).

66 FIG 2 Inter granular crack mode irradiatedin type 304L3 10 21 SS, n/cm2 (E>lMeV)

IrracfiatecS Rn» Grain Area

Boundary \ Position 1

78

70 -

£ *9-

I* -

U - JO - 15 - 20 - 1S - 10 -

c o

o ft. o o

3•

1 S « ! - 0 -I.1 S 10 Pftsibwv frunl

FIG 3 Concentration profiles of elements typein 304L irradiatedSS. 3 102'to n/cm2 (E>lMeV)

67 Figure 4 shows the chromium profile of the material before irradiation. In this non-irradiated condition there is an enrichment of chromium to the grain boundary. The material was in a solution heat treated condition prior to irradiation.

Unlrrofiated Fine Grain Area

Boundar Positioy3 n2

IS -

TO - IS 10 - »s - 341

c o

o c. o o

- 2

-JO -li -2S -'5 -1C -J

FIG. Concentration4. profiles of elements non-irradiatedin type 304LSS

Studie y Ljungberb s d Jenssean g ] hav[1 n e shown that Type 316 LS materiaS s lesi l s susceptible to IASCC after irradiation to relevant BWR neutron exposure levels, 1021 - 1022 n/cm2 (E>lMeV) Figure se , . e5

Contro experienced ro l s with this material sho wsignificana t dela IASCn yi C susceptibility for Type 316compareS S L Typo dt e 304 . HoweverLSS , immunity against IASCC seem bese b to st obtained by avoiding high stresses in the steel.

3.2. Boron Carbide (B4C)

IO By absorption of thermal neutrons, the B isotope in the B4C gets transmuted to helium and lithium according to the reaction:

10B + n -» 7Li + 4He + 2,8 MeV

This mean swellins i s neutroo that C 4 e B t gdu n irradiation sinc ee heliu mostth f o m atoms formeneutroe th y db n absorption reactio trappee nboroe ar th n dni carbide crystals.

68 .ffsl X«

c o ozT Q. C

O « IgI* £ >. 'x -0 ECD

2468 10 Fluenc " n/cmto ex 2 (E>1 MeV)

FIG. Comparison5. of susceptibility IASCCto of type type 316Land 304LSS SS

Secondl e chemicayth l propertie e changeb y structurma sdC 4 B whe e s nth disturbei e y b d formation of Li and He in the crystals, replacing some boron atoms. Smaller amounts of tritium (3H) are also formed during irradiatio othery nb , less frequent transmutation reactions impaco . n Thi s sha t controe ousee nth b behavioumonitod o dn t ro l ca t controstatue e i t th th r f inventord rbu so ro l n yi the reactor core, as discussed later in this paper.

Swelling of the boron carbide as well as the helium gas release are important parameters when designin controe gth l rod. Experiences from contro operatiod ro l shows nha n that swellin bees gha n the service lifetime determining parameter but not the helium release. A hot cell examination of boron carbide samples, irradiated under relevant conditions in the Forsmark 1 BWR, has been performed at PSI [2]. The study showed that the fraction released helium from the irradiated boron carbide was much lower than the helium release curve presented in [3].

3.3. Hafnium

HafniucontroR BW ln m i rod uses si locationn di s with high local neutron exposure. Hafnium is heavy and expensive which explains the limited use of the material in the rods. The benefit of using hafniu absence th s mi f swellineo g during irradiation. This behaviou s beeha r n verifiea y b d controR irradiaten a witd PI BW n ro Elo hB hafniudAB t PSI a p m,ti [4] same .Th e study revealed that that hydrogen content of hafnium in a BWR control rod is low even after operation, as long as the contro freis e frorod l m cracking. This resul agreemenin is t t with calculation [5]s mad.ABB eby Thu showns i t si , tha hydrogee th t n concentratio environmentR BW e th n ni , als casn oi hydrogef eo n water chemistr o caust y w operationelo significano to s i , t concentratio f hydrogeo n n withie th n contro althougd ro l stainlese hth s steel claddin higs gha h permeabilit hydrogenr yfo . Tritium formed boroe ith n n carbide during irradiatio s retaineni boroe th n ndi carbide structur d doet an no s] [3 e influence the hafnium. The latter has also been proven by inspection of hundreds of intact control rods with hafnium, showing no swelling.

Icracfa k occurs that lead wateo st r intrusion hydroge formee b y corrosioy db nma n reactions a sreactioy b wel s a l n between lithiu d waterman sucn I . h cases hydridin f hafniugo s beemha n experience shows controdR a . Experience Figure7 BW n n i hafniuf d lB o rod an e AB s6 us sn mi y sb show tha significano n t t hydrogen concentration e obtaine sar e hafniu th n intac a n di f mo t control rod. Water intrusion afted crackinro r s henci g e onlth e y possible sourc o hafniut e m hydride formation.

69 FIG 6 Hafnium plug with hydride formation Magnification7 FIG of hafnium hydrides

4. CONTROL ROD MANAGEMENT

Guidelines for control rod management must be based on experience data from inspections and PIEs. More than thousand control rods of ABB BWR design have been inspected after irradiation. Many have also been subjec r PIEfo t s suc s neutroa h n radiograph t celho l d an y examinations. Guidelines are formulated for each different design by frequent inspections of leading rod f thao s t particular design inspectionsy B . generae th , l appearancafted ro re operatioth f eo s na wel possibls a l e crack appearanc observeds ei .

4.1. Neutron radiography

case I nwateth f eo r intrusion int ocontroa l rod cracy b , k occurrence consequencee th , e th r sfo boron carbide absorber materia vitaf o s i l importanc thir Fo se reaso extensivn na e wor neutroy kb n radiograph s beeha y n performed especiall n Swedisi y h BWRs. Control rods from shut-down operation and control cell operation, with cracks, have been analysed in this way, see Table I.

TABL E. RESULTI NEUTROF SO N RADIOGRAPHY

Reactor Type Year No of Estimated Tip* 640 Leach-out Wings exposure (snvt) A CR70 1985 2 6,6-7,1 No A CR70 1986 2 6, 87,- 3 No B CR70 1986 13 6,8-7,7 limited loss in 2 blades C CR70 1986 10 7,6-7,7 limited losblade4 n si s B CR70 1987 13 7,2-8,1 limited bladelos2 n si s B CR85 1991 1 5,0 limited bladlos1 n si e D CR85 1992 2 4;4** limited bladelos2 n si s E CR82 1993 1 4,8 limited loss in 1 blade * Tip means the uppermost absorber hole ** Top quarter exposure

By exposin gcontroa wind neutroa ro l o gt n sourc detectind ean non-absorbee gth d neutronsa clear pictur boroe th f neo carbide contenobtainede b n ca t exampln A . f suceo picturha shows ei n ni Figure 8. The lighter areas in the upper absorber bores reveal some B4C loss.

70 FIG. 8. Neutron Radiography picture of an ABB BWR control rod wing tip

In the cases where losses of BC has been detected it has been very limited, which means that

ther bees influenceo ha n n shut-dowe th n eo n margin conclusione Th . s from4 these studie thae sar a t operatee b n controca dd witro l h cracks that appear durin goperatina g cycle recommendatioe Th . f no AB neverthelesBs i starreactoo a t p t tu sno r with know wingsd n ro crack e .th n si

4.2. Tritium

Tritium (3H s producei ) e B,)th Cn i d absorbe y severab r l reactions with d thermafasan t l neutrons. Other source tritiur sfo m production are:

• neutron activatio deuteriuf no reactoe th mn i r water; • neutron activation of lithium contamination in the reactor water; • ternary fissiofuele th .f no

The reactor water conten f tritiuo t m will reac hstabla e level during operation e leveTh s .i l determine e tritiuth y mb d productio e reactoth n i n e reactor th wate d ran r water exchange under normal conditions s beeha nt I . shown tha controa t witd ro lh crack contributy sma tritiu e th o emt content of the reactor water in case of water intrusion. This intrusion leads to formation of THO. The tritiated water may then be mixed with the reactor water. For a plant where the basic tritium level under normal operating condition knowns si , increased tritium levels probably indicate control rods with cracking. This means tha plannine tth controf go inspectiond lro made b n es ca mor e efficieny b t followin tritiue gth m conten reactoe th n i t r water methoA . usinr dfo g tritium analysi controe th n si l rod management work is now under development by ABB.

5 DEVELOPMENT

Experiences conclusioe frocontroR th mo t operatioBW d lB le rod ns thaAB s ha f nto avoiding stresse stainlese th n si s steel clad materiaexteno t bes e y mechanicae th d th wa t s i l l service life time of control rods.

Normally boron carbid uses powdeei a s da r withi nstainlesa s steel cladding avoio T . d stresses in the steel from swelling boron carbide, the void volume of the powder must accommodate the increasing volume of the grains. If we assume that this would be the case we receive a decreasing voi boron i d n carbide powder with irradiation r experienceOu . s indicate tha a largt e variation prevails between control rods regardin behavioue gth powdere th f ro . Modellin behavioue gth e th f ro swelling powder is hence a difficult task.

71 By using boron carbide without internal void, several advantages are obtained:

• The behaviour of a theoretical dense body is much easier to model; allocatiny B voie g• th d outsid voie accommodatior dense th defo f th o ee bodyus e th , f no swelling is ensured; • Designin controg a avoi o t d dro l stresse mors si e readily done.

A concep almose us o t t theoretically (>99%) dense boron carbide pins instea powdef do s rha been developed by ABB. The concept, that includes boron carbide compositions in the range of B4C to Bg5C, has been studied by irradiation of pins in the Forsmark 1 BWR and a PIE of the pins at PSI. From this study swelling has been determined as a function of 10B depletion in the pins. The accuracy in the modelling of the swelling behaviour of the pin can be improved by determining the 10B depletion profile throug e bodyhth . Suc a hwor s usiny performeb e SIMkwa I gth PS S t a d technique.

5.1. Isotopic analysis of a theoretically dense boron carbide pins with SIMS technique

Due to the large thermal neutron capture cross section of 10B, its burnup and helium

production within a boron carbide absorber body is inhomogeneous. The developed SIMS technique offerw no direca s t measur radiae th f elo burnuBgsa n i C 10f B po neutro n absorber test sample with almost theoretical density e sampleTh . , previously mounted lik a ceramographie c specimes i n bombarded within an ultra-high vacuum chamber with a focused 133Cs primary ion beam and the secondary ions emitted fro sample mth e surfac detectee ear separated dan d accordin theio gt n io r mass technique Th . e allow measuro st e quantitativel radiae yth l boron burnup.

5.7.7. Experimental

tese Th t sample selecte r radiadfo l boron burnup analysi cylindricaa s swa l B samplgC 5 f eo 1.7diametem 5m (initiallyd an r ) natural isotopic content, brough estimaten a o t t d. % burnu 5 5 f po sample Th mountes ewa aralditen di , (wet) ground with diamon jj,md0 d 1 disc )an d san (1250 2 , ,40 polished with 3 (im diamond paste, see Figure 9. The formation of cracks at the specimen periphery is clearly visible.

After an ultrasonic cleaning in alcohol, the sample was introduced into an ATOMICA-4000

SIM sputtered San d wit 133A ha n C 0 diametesenerg V m primar12 k p, 1 d 0 1 y 3 an bea n t ra y= io f mo current. For each analysis a shallow 50x50 (J.m crate2 r was eroded away from the sample surface before performing the isotopic measurements in order to get stable secondary ion count rates. The secondary ion optics was (individually) tuned for 3H, 7Li 10B and nB analysis (the tuning was set the lattee samnuclides)o th r rtw efo .

FIG. Macrographyofthe9. hipped, irradiated $Cg B specimen after preparation

72 A "checker board" area scan of 50 by 50 measurement points was then performed across the sample surface with steps of 50 |J.m. A direct (non-corrected non-normalized) area scan of the 10B ancounB dH t rate presentes si Figurn di . Thie10 s direct informatio correctee b o t d dnha becausf eo problems with the sample holder movement (which simulates an oval sample shape) and normalized (presently with HB) because a direct individual secondary ion count rate conversion to an isotopic surface concentration requires a careful calibrations. The isotopic distribution of the two boron isotopes after irradiation can be extracted from the SIMS measurement and is presented on Figure 11. The UB count rate was used to quantify the variation of the IOB content after the irradiation with the assumption that the UB content has not changed during the irradiation. The IOB contene specimeth n i t n afte e initiar th irradiatio f o l conten% e burnun i nth s wel a te s a lar p presente Figurn i d . Fro12 e m this curv n area e a average burnuB d10 p valu s calculatewa e d an d compared with the value measured by (classical) inductively coupled plasma mass spectrometry (ICP-MS) fro mneighbourina g sample after dissolution Table se , . eII

TABLE II. 1U10 B BURNUP DETERMINATION BY SIMS AND ICP-MS

Isotopic concentration 10B burnup (atom%) (%) 10g "B SIMS based 10B burnup analysis - - 47 ±2 ICP-MS burnup analysin so 9.77 90.23 51+0.5 neighbouring sample

5.2. Discussion

e applieTh d SIMS analysis technique requests sample holder movemente fronth n i s f o t (stationary) secondar opticaln yio , electrostatical energy filter movemene Th . t lead variationo st s electricae inth l field strengt secondard han extraction yio n conditions, resultin (witn gi h respect to the area) tilted count rate distributions. The 1!B normalization removes this problem and shows a symmetrical (cylindrical) 10B distribution. Further secondary ion peaks of 3H, 7Li, 12C and possibly also 4He can be identified but are not discussed in this presentation. date Th a presente Tabld Figurn di an 2 esho2 1 e w thatechnique th t e lead quantitativeo st , local (and averaged) boron burnup results whic indeen hca usee db validato dt e burnup modelse Th . e techniquvaluth t limitef o eno s i eo borot d n analysi s beeha ns a sdemonstrate n i I PS y b d hafnium burnup and aptinide distribution analyses on burnable poison and UO2 and MOX fuel rods respectivel . Obviousl7] yd an [Re 6 yf such analyses coul e introduceb d w d ne alsr fo o absorber isotopes in discussion. Presently the error assessment is only qualitative and needs to be improved (e.g compariny .b e futurB normalizatioth n n a i ge n normalization)C wit12 a h s A . indicated above e distributioth , f lithiuo n e assessetritiud b m an n m ca s well da . This analysis however makes sense only after improved (dry) sample preparation which wil testee b l d nexn o t full size dense boron carbide absorber samples.

6. SUMMARY

Several studie f failureo s n controi s l rods have shown thae mechanisth t f crackmo n i s stainless stee IASCCs i l avoio T . d IASC moste Cth efficien removo t seem y e b wa t o est causer sfo stresses in the component. If the absorber material can be designed not to strain the steel cladding, cracks will not appear in control rods under the nuclear life. Such an concept is to create a boron carbide pin of almost theoretical density where boron carbide densification is accomplished prior to operation. Modelling of the operational behaviour of pins is much easier then modelling of powder. By post irradiation isotope analysis in a cross section of the pins with techniques like SIMS, the modellin improveds gi usiny B . g boron carbide pins without internal voiensures i t di d tha voide th t ,

73 now allocated outsid bodye eth ensures i , accommodatior dfo f swellingno . Designin gcontroa d ro l avoio t d stresse mors si e readily done this way.

Managemen f controo t l rods mus e baseb t n experienceo d s from inspection programs. Examinations suc neutros ha n radiography have bee nvaluebla e hel reveao pt consequencee th l f so absorber material behaviour in case of water intrusion after cracks in the stainless steel cladding. Data shows that the BC losses are neglible in rods with cracks. Cracks can be allowed to appear

under operation. Knowledg4 e of the base-line tritium levels in the reactor under operation makes it possible to interpret peaks in the tritium levels, which can be used in the planning of control rod handling. A concept is studied at ABB Atom.

250 a)

Count rate

Hi 8000

000 0 5 2 5 2 2 0 0 2 5 7 1 0 5 1 5 2 1 0 0 1 5 7 0 0 5 0 5 2 0 0 0 0 X-Axis [mm]

b)

Count rate

llii 120000

IHI 60000

ooo 0 00 0 25 0 50 0 75 1 00 1 25 1 50 1 75 2 00 2 25 2 50 X-Axis [mm]

FIG. SIMS. 10 area scan 5e S5oth f C absorber specimen (Raw data, not corrected, not normalised), countB a)10 rate distribution, countB l! b) rate distribution

74 Boron isotopic distribution along X-axis

100-

I.P so-

c _o 60- "<5 .Q w 1 .75 mm ^ 40- o 'E. 2 o 20- (A

-1.0 -0.8 -0.4 60. -0. 2 40. -0. 0 20. 0 1. 8 0. 6 0 Radius along X-axis [mm]

FIG. 11. Boron isotopic distribution along the X-axis of the specimen

10B and Burn out distribution along X-axis

, Specimen averag% 7 e4 Bur: t nou 90- 90

80- ^-A.A.A. 80 i * A ^ ^ A -Jkl 70- \ x"- - - "x / 70 A * / 60r -* \ - 60 o o v o 50- V* - 50 CO 40- '/*/ \'% ' ' i7/^ 'V - 40 m 30- / *.. - - - A«-» \ k 30

20- *^^»^».*.^^" _ 20 - 1.75 mm • 10- 10 • n- n -1.0 -0.8 -0.4 60. -0. 2 40. -0. 0 20. 0.6 08 1.0 Radius along X-axis [mm]

FIG. 12. B content and burnup distribution along the X-axis of the specimen

(calculation :

75 REFERENCES

[1] JENSSEN, A., LJUNGBERG, L., Sixth Int. Symp. on Environmental Degradation of Materials in Nuclear Power Systems - Water Reactors, Breckenridge, CO, 7-10 August, 1995,p.l043. [2] Non published data. [3] STRASSER YARIO, ,A. , Contro W. ,Material d Ro l Burnabld san e Poisons- , 79 EPR S IIP 708(1980). [4] BART, G., et al, Hf Control Rod Analysis for ABB - Final Report, PSI report TM 43 90-41 (Sept 1990). [5] MASSIH, A.,"Hydrogen Uptak n B AtoHafniui e AB m n i mContro l Rods: Model Calculations" AtoB mAB , Report 89-441K U , , [6] ZWICKY, H.U., et al., "Evaluation of the Radial Distribution of Gadolinium Isotopes in Nuclear Fuel Pin Secondary sb Masn yIo s Spectrometry (SIMS)", Radiochimica Act, a47 9-12(1989). [7] Radial Plutonium and Fission Product Isotope Profiles in Mixed Oxide Fuel Pins Evaluated by Secondary Ion Mass Spectrometry (SIMS), JNM 202 (1993) 65-69.

76 EXPERIENCE OF CR AND RCCA OPERATION IN UKRAINIAN WWER-1000: ASPECTS OF RELIABILITY, SAFETY AND ECONOMIC EFFICIENCY

A. AFANASYEV Ministr Energf yo Ukrainen yo , Ukrainian State Department of Nuclear Power, XA0053632 Kiev, Ukraine

Abstract

The next topics are represented in the paper • A brief history of WWER-1000 control rod (CR) and WWER-1000 rod cluster control assembly (RCCA) design. • Evolutio WWER-100f no manufacturinR 0C g technolog designd yan , • Experience of RCCA operation, • Lifetime extension of WWER-1000 boron carbide CR, • WWER-1000 reactor core operation problem partiao t e sdu l RCCA insertion, • Designin licensind gan g procedure firsd san t operational experienc f WWER-100eo 0 RCCA (CR) wit combineha d absorber "boron carbide-hafnium chromium-nickea d "an l alloy cladding maie Th n conclusione sar • Fuel assembly (FA) bow is the main reason of partial RCCA insertion during reactor core operation However, the s driveRCCe it th r witus d f ba rAeo an h increased dead load, alongside with other measures, allo reduco wt e th e probability of incomplete RCCA insertion, • The materials used in CRs of RCCA in existing reactor operating modes have been working reliably, • The use of hafnium under an appropriate price policy can give certain economic advantages for the Ukrainian NPPs, however, additional researc needes hi orden di confiro rt specifie mth physicaR cC l characteristic reliabilitd san y

1. INTRODUCTION

The RCCAs of a reactor control protection system (CPS) are the most important elements of maintaining the safe reactor operation, ensuring control and distribution of reactor core power level and fast reactor core transfer from initial to subcritical condition during an accident. The share of RCCAs in fuel reloading cost is 1-1.2%. However, for CR materials choice it is necessary to take into accoun possibilite tth f inexpensivyo e RCCA recover disposalr yo .

SELECTIOE TH . 2 MATERIALF NO RCCD SAN A DESIGN, DEVELOPMEN RCCF TO A

formee Inth r USSR Moscoe th , w manufacturing plan f polymetalo t s (MPPdesignee th s )wa r manufactured an r almosfo l reactoR al tC f o r r types boroe Th . n carbide (E^C, wit hnaturaa l mixture f isotopeso s selecte absorbe)wa n a s da r material becaus f it'eo s accessible, chea quitd pan e effective material Stainless steel 06 X18 N10T was selected as CR clad material. The design of the WWER- 1000 RCCA is shown in a Figure 1.

Up to 1985, CRs were produced by a joint drawing method. An initial pipe with a diameter of j s fillewald wa an d, l thicknesm withavin1C mm m 14 hB 7 0. g densita f so f 1.6yo 5 g/cm s Thewa t ni extended through some dies. Drawing was performed till the moment when the sizes of the pipe became equa 8.2x0. thao o t averagle s , th t 6mm e densit absorbee th f yo r became equa o 1.9t l 5 g/cmj 3 densitC 4 centrae B th e n yi pipe ls Th 1.6-1.6 parth laye em wa f ri to re 2-2.th 5 n g/cmo 1d g/cnf".an

From late 1985 until now e RCCA,th e producear s a metho y db f vibrdo o compaction a f no initial pipe 8.2 x 0.6 mm with boron carbide of 1.75 g/cm3 density. The initial pipe is not subject to deformation [1]. The advantages of this technology are the following' • Less absorber materia neededs i l ; • Some technological operations are eliminated from the process, • The smaller the density of the boron carbide (1.75 g/cm3) in the rim (external) layer, the smalle swelline rparticleth C layee 4 th B f n rgi o undes si same rth e neutron flux.

77 spnng

til42"* rod cladding 8.2 x 0.6 stainless steel

FIG. 1. WWER-1000 Rod cluster control assembly (RCCA) design

78 3 3 The change of B4C density from 1.95 g/cm up to 1.75 g/cm has not reduced the efficiency of e RCCAth RCCe Th . A efficiency measurements, with respec variouo t t densitiesC 4 B s , have been made by using WWER-1000 critical assemblies. In the first case, the critical assembly was the fuel assembly (FA) fro firse mth t loadin f unit-Soute go th f h1o Ukrainia RCCAs)9 (4 secone P th nn NP I . d experiment, the critical assembly was the FA from the first loading of unit-1 of the Kalinin NPP (61 RCCAs). The results of the measurements [3] are shown in Table I.

TABLE I. RCCA EFFICIENCY MEASUREMENTS

Number B4C density Efficiency of RCCA relative Comments inCR RCCA efficiency g/cm3 % (%) 1 1.80 9.8x1 Q-2 100% 2 1.67 9.8xlO"2 100% 3 1.14 9.4x1 0'2 96% there corresponds to burnup 10B 33°/c 3 for initial density B4C - 1 .75 g/cm 4 0.95 9.2x2 10' 93,7% there correspond % o burnut s45 B 10 p 3 for initial densit - 1.75_j/cm C 4 yB 5 0.69 8.6xlO'2 87.6% there correspond % o burnut s60 B 10 p j for initial density B4C - 1 .75 g/cm

The initially RCCA lifetime was determined to be 1 year for control (regulating) groups (CGs) year5 shur d sfo an t down (scram) groups (SGs).

. EXTENDE3 D RCCA OPERATIO POST-IRRADIATIOD NAN N EXAMINATION

The results of high flux tests of an RCCA model in a material test reactor and commercial operation experience have created premises for RCCA lifetime extension. In order to provide lifetime extension, the RCCAs of unit-5 of the Novovoronezh NPP had been operated for longer time comparing with the period given by the project. After irradiation, these RCCAs have been examined e Researcth t a h Institut f Atomio e c Reactors e Novovoronez(RIARth d t cellsan )e ho Th P . NP h research was performed and paid by the Russian and Ukrainian utilities (ROZENERGOATOM and GOSKOMATOM) [1-3].

The value of the neutron flux was checked up by means of 10B burnup and accumulation of activation products (60 Co)maie Th . n result e givesar Tabln ni I [1-3]eI . Completed tests furthed an r research on CRs from CG have confirmed the expected outcome. Vibre compacted CRs have the best operating performances in comparison with joint drawn CRs.

e divergencTh n maximui e m thermal neutron flu xs betwee(FCR n i )n calculated data max (2.1 xlO21 by the Kurchatov Institute and OKB GP) based on 10B burnup (4.5x1021) and data based on 60Co accumulatio e steeth l(6.0xl0n R i ncladdinC e th s bee 21f gha o ) n determined througR C G S h research calculatee Th . d results appeare underestimatede b o dt .

analysie l researchTh al f o s , resulte establishinn di e valuegth r severafo s l parameters, i.ee .th maximum allowable 10B burnup for vibro compacted CRs was 57.5%, the CR swelling (A®) was less (implyin% tha2 1. n g that there were safety plasticitd an y margin claddine th f o s g materiale th d an ) relative thermal neutron flu3.6xl0s xwa 21.

79 TABLE II. RESULTS OF RCCA RESEARCH

Characteristic, vibratinR C g filling mode joinR C t drawing parameter (Absorber of vibrating packed type) mode Tim operationf eo , CG2y eff1 .49 , days, CG 3 y, 8 19 eff. days, SG 7 y, 2028 eff. days, maximal neutron neutron energy: neutron energy: neutron energy: fluence,Fmaxxl021 E>08MeV -3.3 E>08MeV -7.6 E>08MeV -4.5 E>01MeV~5.4 E>01MeV -12.5 EXHMeV -6.1 therma5 . 1 ~ l thermal -3.2 thermal -4.5 General state of All of CR have saved the form and tightness. Longitudinal cracks theCR Ther blacs ei k dense oxide fil thicknese mth f so (L=10mm detectee )ar d which 4-7 microns on a surface at the bottom CR swellinR C e th n gi 0.0 0.02- 6 0.11 -0.16 areaofFmax 0 0.24 - 0.72 1.35-2.0 (A0)[mm] % ,/ Burnup of 10B, % max. ~32.5 max. -53.2 max. ~72 - 77 Helium release 70-130 [cmj] 260-350 [cm3] ~100[cmJ], part of gas under cladding lefs ha t through cracks Pressurs ga f eo at 20°C 2-4 bar at 20°C 15-25 bar 2-9 bar, ,Part of gas under cladding at 333°C 4-8 bar at 333°C 35-56 bar lefs ha t through cracks

Absorbing Powder B4C freely At the CR bottom therR eC e swellinth f o g material B4C state hailed down under was a sintering of a bottoe leath C 4 mo dB t cutting of cladding on powder, which densely open cracks samples adjoins to the cladding. Change of cladding diamete 0.7- 2 %0. r+

Mechanical ab at 20°C - 820 abat20°C-893 abat20°C-830 properties of abat350°C-520 ab at 350°C - 630 0bat350°C-580 cladding in the t 20°0 a 50 2 C 0 - a 20°t 0 a 67 2 C 0 - a a02at 20°C - 640 areaofFmax cr t 0350°C2a 0 -29 350°t a 0 2 0 58 C a - o-02at350°C-480 a [MPa] b 50at20°C-14.2 50at20°C- 18.6 So at 20°C - 26,4 a [MPa] 02 50at350°C-7.1 50at350°C- 1.1 80at350°C-3.9 50 [%] General conditioR C s ni Cclose Rar extremo et e Actual duration of conclusion on satisfactory. It is allowable condition. CR operation has exceeded spent CR state probabl continuo et e swelling makes 0.24- allowable time for this their operation within 0.72 % in the area of CR type, that has more on e year (fuel pmax cause excessive dth e cycle) swelling of B4C and cladding cracksf o , , washingC awa4 B f yo

4. EXPERIENCE OF RCCA LIFETIME EXTENSION

4.1. Operation time

The operation time of RCCAs for CG was enlarged from 1 to 2 fuel cycles [3].

4.2. Life extension

unconditionae Th l specificatio RCCe th f Ano lifetime extensioyear6 froo G t sS m 5 r (an nfo d t possible eveyearsno 7 s o nt wa ) , e resultresearche base th th e calculate n f th do o s s a , d thermal neutron flux density above the reactor core was lower than the actual value.

80 In order to be able to specify the SG RCCA lifetime extension, the exact position of the RCCA lower flat end above the reactor core for each unit in each fuel cycle is needed. The lower flat end of RCCA would be within the 0.9 - 14.9 cm above the core in serial WWERs-1000 (e.g. for unit-5 of Novovoronez 7.7-8.- P hNP 7 cm) thermae .Th l neutron flux density withi same nth e range shoult dno b constanea t value dependence Th . allowable th f eo eRCC G timS f Aeo operation accordine th o gt positio lowee th gives f i nro d flaTabln ni en t [2]I eII .

TABLE III. ALLOWABLE OPERATING TIM RCCAF EO s

Positio lowee th f nro flad en t 10B burnu yea1 r f rpo fo allowable time of SG 640 above the core operation (300 eff. days) RCCA operation h (cm) (%) (eff. days/years) 0.9-8 11.4 1530\5 9 11.1 1570/5 10 10.7 1630/5 11 10.0 1735/5 12 9.3 1870/6 13 8.7 2000/6 14 7.85 2210/7 14.9 7.23 2400/8

Approximatel Ukrainiaf o % 5 1 - n 0 WWER-100y1 RCCAG 0S s have been operated durine gth 6 fuel cycles e totaTh . l duratio f operatio o nt excee no d ddi n 1700 -1790 effective dayse Th . experienc RCCG S f Aeo lifetime extensio mord an t receivee6 yearno o n t s sha d broad attentior nfo the reason of difficulties related to monitoring of the RCCA lower flat end position. During the whole perio f operatindo g WWERs-100 e Ukraineth n 0i , there occurre RCCo dn A damage e materialTh . s used in the RCCAs performed reliably.

5. RESULTS OF IMPROVED RCCAs BASED ON A COMBINATION OF Hf AND B4C

The cost, allowable tim f operationo e possibilitd an , f inexpensivyo e disposae maith ne ar l consumer feature f RCCAso s orden I . o increast r e RCCeth A lifetime requires i t i , replaco dt e eth

bottom (300-500 mm) n/oc absorber B4C by an n/y absorber (Hf, Dy2Ti05or In-Ag-Cd, which doenot claddine us swello t d g materia)an l that wil more b l e stabl radiatioo et n embrittlement.

12 RCCAs with combined Hf-B4C absorber were designed, manufacture accepted an dn a y db interdepartmental commission. In 1997, they were loaded in the WWER-1000 of the Rovno NPP for the operation accordin joina o gt t decisio f GOSKOMATOo n (witP MP h d participatioMan f o n Kurchatov institute, RIAR, OKB GP). The CR cladding was made from Cr-Ni alloy EP-630Y. The design of the combined absorber CR is shown in Figure 2. It was decided to use the new designed RCCA in SG.

e ingotTh s e Ukrainiamadth n i e n zirconium plant ,e calcium-thermabaseth n o d l recovery technolog d furthean y r double electronic radial remelting , f sourcwerH ew era use materials a d . Further processing and manufacturing of Hf rods for CRs were carried out in Russia. Samples of Hf pipes , wal rodd wit (9.6xan s lm mm diameteha ) , thickneslengtwitm 6 0 9. 2mm hmm ) h10 2 r(0 f so a diameter of 8.2 mm, length 10 mm were stage by stage irradiated and investigated in RIAR for the period 1993- 1997.

Irradiatio f sampleno s locate ampoulen di s were materiamade th n ei l test reactors (BOR-60n i , a sodium environment at a temperature of 340 - 360°C, and in CM-2, in water at 60 - 80°C). At the first stage irradiatioe th , s performenwa BOR-6n di fasa y t0 b neutro MeV1 > n 10 x (E flu 22) 1 .f xo

81 n \ 7 rod cladding 8,2 x 0,5 Cr - Ni alloy

V

FIG. Combined2. absorber designrod (Hf+ B4C)

The diameter of the pipe decreased by 0.3% and that of the rod by 0.8%. The lengths of the pipe and rod were increased by 0.3% and 1%, respectively after the irradiation. The densities of the pipe and rod have decreased by 0.42% and 0.55%, respectively. The pipe has plasticity at the level of damageo 1-2 n therd s %longitudinao an ewa Tw . l radial cracks, which could have d occurrero e th t da manufacturing stage, were detected. The Hf rod was damaged (an = 4 kgm/cm2) [4] by shock tests. For that reason the Hf rod has been located inside the cladding. As cladding material, the Cr-Ni alloy (Cr-42%, Ni-56%, Mo-1%) was used. This material named as EP-630Y has a high plasticity after long irradiation and a high corrosion stability (resistance) in a water environment. At the second stage, the irradiation was performed in BOR-60 and CM-2 by a fast neutron flux of (E>1 MeV) 3.4x10". The average radiation . growtAfte sample firse 3% th th rs tf ho stag ewa f irradiationeo , cracksame th f eso for sizd man e were detecte rode th sn di [5] mechanicae .Th l propertie EP-630f so givee Yar Tabln ni e IV [6].

TABLE IV. MECHANICAL PROPERTIES OF EP-630Y

22 Temperature Neutron fluencex 1 0 <702 c?b SP [°C] E>8MeV [cm'2] [MPa] [MPa] [%] 0 360 770 53 20 0.7 640 840 50 2.7 750 870 44 6.8 820 910 22 0 250 640 52 300 0.7 480 800 40 2.7 540 660 39 6.8 660 680 16

82 The behaviour of modelled CR was investigated under accident conditions:

• B4C, Hf, Dy2TiO5 absorbers do not affect the EP-630Y cladding at temperatures between 350 and 500°C during 1000 hours; • The CR research under simulated cladding leakage conditions (temperature of overheated

steam of 1150°C) has shown, that there is reliable compatibility of Hf and Dy2TiO5. The f modelleH surfacR dC e damag s revealeewa levea t 400)amdf a o l meltine Th . g occurred in the Hf cladding contact area under a vapour temperature of 1250°C [10].

6. SOME LICENSING REQUESTS

In orde o obtait r e licence nucleath nth f o e r regulatory administration (NRAtesR C t r )fo performance t shouli , e showb d e RCC ne tim th that th f Aeo t a , insertion reactivite th , y derivative (dp/dr) wil negative b l e (the accident Chernobyfactoe th f o r l NPP). This conditio always ni s required if the CR lower part has the lowest differential efficiency.

7. EXPERIMENTAL RESULTS OF THE COMBINED Hf-B4C ABSORBER EFFICIENCY. f mor o efficiency C e 4 f efficienceB us H e e e th s Th Th beef . o ha y n % calculate80 s i d dan durable cladding material allows to decrease the wall thickness from 0.6 mm down to 0.5 mm and consequently to increase the absorber diameter from 7 mm to 7.2 mm. The calculated combined Hf-B4C RCCA efficiency is equal to 1.025 of the ordinary RCCA [6, 7].

Efficiency measurements have been made in the beginning of cycle (BOC) at a level of 1-2% of Nnommaij with a temperature of 280°C and a pressure of 160 bar. The location of the RCCA groups, includin combinee gth d #7)d group an s show i , 2 s(# RovnFigure n no Th . oe3 WWER-1000 reactor core parameters in the BOC (for the 11th fuel cycle) are presented in Table V.

TABLE V. ROVNO REACTOR CORE PARAMETERS

Dat experimenf eo t Effective days CG position H3BO3 concentration Xe % g/kg 6 Nov. 97 0 0-90 8.5-9.3 no 17 Nov. 97 7 55-57 7.7 yes 18 Nov. 97 7 36-50 7.7-8.0 yes

On 6 November 1997, differential and integral RCCA efficiencies were measured at the time of RCCA insertion by means of a reactivity meter. The inserted reactivity was compensated by CG (#10) movement. Separated RCCA movement, combined groups (#2 & #7) movement, and groups (#1 and #8) movement have been made step by step (in turn) from upper position (H=100%) down to lower position (H=0%) by steps of 5%. The reactor criticality was maintained constant, as well as the

coolant pressure, temperature and H3BO3 concentration. The measurement results are presented in Figures 4 and 5.

On 18 November 1997, differential and integral RCCA efficiencies were measured at the time of RCCA insertion by means of a reactivity meter. Inserted reactivity was compensated by H3BO3 concentration change. Separated RCCA movement, combine ) movementd#7 group& d 2 (# an s, movemen8 # groupd an 1 # st have been made ste stey turnpb n p(i ) from upper position (H=100%) dow o lowent r positio ne reacto (H=0%Th . r ste5% y criticalit b )f po s maintainewa y d constants a , coolane welth s a l t pressur levetemperaturee d positioth f eH=55% G an t o l C a s e th n wa d . an ,

83 16 18 20 22 24 26 28 30 32 34 36 38 40 42

loop #4 0 Io°P * 3

01

02 © o 03 04

05

06

07

08

09

10

11

12

13

14

15

loop # 1 ^ loop # 2

1 4 9 3 7 3 5 3 C 3 1 3 9 2 7 2 5 2 3 2 1 2 9 1 17

\la xi "FA Number in 60° symmetry sector - Numbe ^ RCCf ro 2 i/ A group (X-Rf RCCA)

FIG Reactor3 core Rovno ofthe umt-3, lllh fuel cycle

In addition, the measurements of combined group (#2 & #7) 10% insertion efficiency (that correspond f e absorberheighH th f o o t ts ) alongside with insertio 8 wer# f d egroupo nan 1 # s completed The same was done for 20% of separate RCCA insertion The results are shown on Figures 6 and 7.

84 0.014

0.012

0.01

0.008

0.006 •Calculation data 0.004 .'A ————Approximation (group #1) Approximatio— —— n (grou) p#2 V 0.002

0 0 10 5 9 0 9 5 8 0 8 5 7 0 7 5 6 0 6 5 5 0 5 5 4 0 4 5 3 0 3 5 2 0 2 5 1 0 1 5 FIG. 4. Differential characteristics of RCCA group (#1 and #2 (Hf)), The reactivity was compensated by control group (CG)

0.8

0.7 * experimental data (group #1) 0.6 D experimental data (grou) p#2 •calculation data 0.5 -Approximation (grou) p#1

0 10 5 9 0 9 5 8 0 8 5 7 0 7 5 6 0 6 5 5 0 5 5 4 0 4 5 3 0 3 5 2 0 2 5 1 0 1 5 FIG. Integrated5. characteristic of RCCA (Hf)),#2 groupand (#1 The reactivity was compensated by control group (CG)

0.020

0.018 •

0.016' ExperimenS Nov detecton 1 o t , 6 . r • Exoenment on Nov. 17,2 detectors 0.014

0.012 S? ex 0010 C.008

o 0.006 CO 0.004

0002

0.000 group #1 grou (Hf2 p# ) group #7 (Hf) grou8 p# FIG. 6. Integrated efficiency of 10 % of groups RCCA (groups #/, #2(Hf), #7(Hj). #S, the RCCA lower flat end position in core (H%) was changed from 100% up to 90%

85 0014

0000 group #7 (Hf) group #8 FIG. 7. Integrated efficiency of 20% separate (one) ofRCCA in group (group #7(Hf), #8), the RCCA lower position flatend corein changed(H%)was from80% to 100% up (experimental data)

By the 13th of September 1998 at the end of cycle (EOC), the measurements have been made at the powe N% rnomma leve50 f |o l. Accordin preliminare th o gt y data e resultth , s were simila thoso t r e obtained earlier resulte Th . s allo concludewo t : • integral characteristics of RCCA with and without Hf are almost the same (Figure 8); momene insertion% th t 20 a f integrae o t th , • l efficienc lowee th f RCCe ro y th par f Ao t with Hf was 24% less then the ordinary RCCA; momene insertion% th t 10 a f efficience o t th , RCCe • th f equas Ao y witwa 0.53o t lf hH - 0.58 against the ordinary RCCA efficiency (the calculated value was 0.8); « at the moment of reactor scram, the summary efficiency of the RCCA was equal to 8.7% and exceede calculatee dth d value (7.5%).

In orde clarif o f t refficienc H e reasonw yth lo existin n f i yso g RCCAs usefus i t i ,perforo t l m calculation of: • burnup effect (neutron spectrum influence); • research on geometry effect on the RCCA lower part efficiency; • safety analysis ofRCCA with Hf as CG.

When RCCA with Hf are used as SG all safety requirements are met.

:00-* group #1 group #2 (Hf) grou (Hf7 p# ) grou8 p# Integrated8. FIG efficiency of separate (one) RCCA groupsin (groups(Hf),2 ) (Hf),(experimental# 7 # , #8 #1 data)

86 8. SAFETY PROBLEMS OF REACTOR OPERATION WITH INCOMPLETE RCCA INSERTION

During 1992 and 1993, almost at all WWER-1000 in Ukraine, Russia and Bulgaria incomplete RCCA insertion occurred, i.e. an RCCA stuck in an intermediate position and RCCA drop time exceeding 4 s (design time). A programme of additional quarterly measurements of RCCA drop time was developed and implemented. In the case of RCCA operation violation (and when there was no chance to cease one) the unit had been transferred to the operational mode with three loop coolant circulation and with preliminary power reduction to 67% of Nnomina|.

The following Ukrainian units were transferred to the 3 loop operational mode: • Zaporozy uni P ; fueGeneratio; 8 1 te NP l d cyclan 7 e# n loss xlO1 -2, 9 kWh; • Zaporozye NPP unit 3; fuel cycle # 7 , 8 and 9; Generation loss-3,3x 109 kWh; • South Ukrain uniP fue; 2 te NP l cycl Generatio; 7 e# n loss-1,0 kWh9 x10 .

In 1993 chiee ,th f reactor designe d calculateha r e neutrodth n physica thermad an l l hydraulic characteristics of the reactor core for the RCCA drop time up to 10 s, keeping in mind the case when all RCCA are inserted except the most effective RCCA. It was shown that the design safety criteria were met. However resulte th , s obtained wer t use eno canceo dt l operational restrictions accordino gt conservative th e principle. Fuel assembl maie th y s n i (FA reasow ) bo f incompletno e RCCA insertion during reactor core operation. Complicated form of FA bow results in: • appearance of additional friction force between CR and guide thimble clad; • either RCCA drop time increase in scram mode or RCCA jam in intermediate position;

In order to reduce the probability of incomplete RCCA insertion, to decrease the size of FA ensuro t d e reactor'th ean w bo s safe operation e followinth , g compensatory measures have been performed durin schedulee gth d repair 1993-1997n si : designew Ne 1)d heavier RCCAs with gafniu r titanamo f dysprosiuo t m weree useth n i d Rovno and Zaporozye NPP; deae Th dRCCe loath f ) dAo 2 increaseds drivewa r rba ; orden I decreaso rt ) 3 e losseRCCe th f so A kinetic energy during RCCA insertion (the pump effect), holes in the RCCA driver bar have been drilled; e positioe bundlTh th f o ne) 4 safety tube (BST s updatewa ) d wite purposth h f o e maintainin sprine g, withouth FA g e blocth t exceedinf ko comprese gth s maximum (axial compression of FA head).

Figur showe9 curvee sth f RCCAso s (bot originaf ho advanced an l d design) drop speed [9].

RCCA driver wrth designed bars (tubes) - Vt RCCA driver with drilled barc V s- RCC— —— A driver with drilled barincreased san d dead loay V d-

0.00 0.50 1.00 1.50 2.00 2.50 3.003.50 H core, m FIG. 9. Curve of average speed of RCCA drop

87 In curvee FigurRCCe th th , f Aso 10 e drop time changperioe th r defo 199 3- firs t quartef o r 1998 are represented as an example for Zaporozhe NPP-2. Before the implementation of compensatory measures for Ukrainian WWERs-1000, the average RCCA drop time (T average) was RCCA0 dro20 a , d s p s ha 3 tim 3. e largeRCCA8 2 d r an tha ss wern4 e jammed After implementation of measures, the average RCCA drop time was 25s, whilst 11 RCCAs had a drop time larger than 4 s onld an RCCy1 jammeds Awa .

t sec

Measurement10 FIG of RCCA drop time ZNPP-2on

Only afte beginnine th r implementatioe th f go compensatorf no y measure perioe th r dsfo 1995- first quarte f 1998 ro RCC 0 14 , A drop tests have been performed, from whic hav8 1 h e been performed

t 40-50a %i powe N r level, some measurements have been don t poweea r leve f 20-30o l %i N nomma

and the resnomma t was performed under "hot shutdown" conditions The results of "under power" tests have complied wit desige hth n requirement average Th s e resultRCCe th f sAo drop time measurementn so l Ukrainiaal n WWER-1000 units befor afted ean r compensatory measure presentee sar Tabln di . eVI

TABLE VI SUMMARIZED RESULTS OF RCCA DROP TEST

i averag )/Numbes e( RCCAf ro s wit 4s/Numbe> hT f ro jammed RCCAs NPP/UNIT perioe th r d Fo before For the period after compensatory measures were compensatory measures were applied (1993-1995) applied (1995 - first quarter of year 1998) Zaporozhe NPP 1 322/34/8 1 /O/23 O Zaporozhe NPP 2 3 18/30/0 2 18/0/0 Zaporozh3 P eNP 3 17/29/6 /O/1 26 O Zaporozhe NPP 4 321/11/0 2.52/0/0 Zaporozh5 P eNP 328/9/8 285/2/1 Zaporozh6 P eNP 3 15/1/0 2.68/2/0 South Ukrain1 P eNP 35/0/0 2.51/0/0 South Ukraine NPP 2 1 / 2 3 / 2 7 3 248/0/0 South Ukraine NPP 3 355/ 19/0 235/3/0 Rovn3 P oNP 3465/35/5 2 74/ 3 / 0 Khmelmtsk1 P NP i 3 72/12/0 253/1/0 REFERENCES

fll PONOMARENKO, V, RISOVANY, V., SHMELEV, V, KONDRATYEV, V., et al., "Control burnabld roan d e absorbe f reactoo pathd d ro an rrf thei o s r perfecting", Problem f Atoo s m Scienc d Engineeringan e , Series: Radiation Damaging Physic d Radiat.oan s n Materials Science, issues 2/63 and 3/63, Kharkov, (1994). ] [2 CHERNYSHO V VRISOVANY , KOSOUROVV. , , VASILCHENKOK. , t al.e e , Th ,I , substantiatio f WWER-100no RCCG 0S A lifetime extension", Moscow, (1994). ] [3 PONOMARENK , CHERNYSHOVV. , O , V RISOVANY, , V OSADCHIJ, , A , VASILCHENKO t al.e , , I. "Th, e substantiatio f WWER-100no RCCG 0C A lifetime extens.on up to 2 years" , Moscow, MPP, (1992). . F41 CHERNYSHOV V RISOVANY, V., PONOMARENKO, V., et al., "Research of hafn.um contro modeld ro l s irradiate reactoe th n di r BOR-6 fluenca o t p f fas0u eo t neutron MeV1 > s(E ) equal 1x1 022 cm'2", Dimitrovgrad, (1994). T51 HUDJAKO , KLOCHKOVA. V , RISOVANYA , t al.e , , "DimensionaV. , d structuraan l l stability of a' hafnium sample irradiated up to high dose", Problems of Atom Science and Engineering, Series: Radiation Damaging Physic Radiatiod san n Materials Science, issue 2/68,

F61 AHOVSKIH V, SHMELEV, S, et al, "State of WWER-1000 RCCA development and production, Problems of Atom Science and Engineering, Series: Radiation Damaging Physics

and Radiation Material, : ' . s Science, issue2/66, Kharkov 997)1 ( , . m Report #32/1 -89-97, Kurchatov Institute. ,Dr^A 8 SOKOLOV , BORISD , GRABKO, V. , , KOLISNICHENKOV , l a Resultt e f RCCM ,s o A neutron physical characteristic measurement WWER-100n i s(fueC lBO cycle th t e0a #11), Report, Rovno, (1998). F91 AFANASYE A V"Compariso f WWER-100o n coreR PW s d wit0an h fuel assemblw bo y operational experience and problems", Problems of a Atom Science and Engineering Series: Radiation Damaging Physics and Radiation Materials Science, issue 2/68, Kharkov (1998). flOl CHERNYSHO V VTROYANOV , "LessonV. , s learned from contro d irrad.at.oro l n experience- developmen f advanceo t d absorber theid san r refractory propert.es under accident condition",' (Proc. Int. Top. Meeting on Light Water Reactors, Portland, Oregon, 1997).

NEXT PAGE(S) left BLANK

89 DYSPROSIUM AND HAFNIUM BASE ABSORBERS FOR ADVANCED WWER CONTROL RODS XA0053633

V.D. RISOVANIY . ZAHAROVV A , , E.P. KLOCHKOV, E.E. VARLASHOVA, D.N. SUSLOV State Scientific Cente Russiae th f ro n Federation "Research Institut Atomif eo c Reactors" (SS RIAR)F CR , Dimitrovgrad

A.B. PONOMARENKO, A.V. SCHEGLOV Moscow Polimetal Plant (MPP), Moscow

Russian Federation

Abstract

Disprosium titanate is an attractive control rod material for thermal neutron nuclear reactors such as WWER and RBMK Its main advantages are almost non-swelling, no out-gassing under neutron irradiation, quit high neutron efficiency, a high melting point (~1870°C), non-interaction with the cladding at temperatures above 1000°C, simple fabrication, non- radioactive waste and easy to reprocess The disprosium titanate control rods have worked without operating problems in the reacto durinR WWER-100 n MI ryeari 7 g1 d san year04 s After post-irradiation examinations, this long-life contro typd ro le was recommended for using in the nuclear reactors Dysprosium hamate is a promising absorber ceramic material The research results confirmed that it has a large radiation damage resistance The examination results of hafnium dummies (GFE-1) irradiate BOR-6n di presentee 0ar maximue Th d m accumulated neutron fluenc 4xl03 s 22ewa cm"2 (E> MeV01 d )an the temperatur hig o 360°t o t e h0 eradiatio CDu rang34 s ewa n growtabsence th axia n betweea d p hf an ga elo (3- ) e 4% nth dummy and the upper capsule tip the dummies were bent The irradiated dummies have high mechanical properties Other aspects of the expected hafnium irradiation behaviour and the use of hafnium in control rods are discussed This report

presents some experimental dat Dyn ao 0 TiO, DyOHf , possibilitieHfOd an WWEn i theif so e us rR control rods 2 3 2 2 3 2

1. INTRODUCTION

At present, boron carbide (WWER-1000) and boron steel (WWER-1, -2, -3M, WWER-440) are the most widely used in control rods of Russian water reactors. These absorber materials have large radiation damages owing to (n,cc)-reaction on 10B isotopes, helium formation and swelling. The first

damag f WWER-100eo Clusted 0Ro r Control Assembl swellinC y4 (RCCAB claddino d t gan e du ) g cracking was noticed in the Novovoronezh Nuclear Power Plant.

Long before this accident, the State Scientific Center of the Russian Federation "Research Institut f Atomio e c Reactor" (SS RIARF CR ) (together wit e Moscohth w Polimetal Plant (MPP)) carrie t researcdou (n,yn ho ) absorbers, t swellwhicno o h.d More tha cerami0 n6 c absorber materials Hf-alloyd , puran f Cd eH , sHf wer, Gd e, investigatedSm , Eu , baseDy .n do

Post-irradiation examinations have shown that lanthanide oxides (Ln2O3-MO2) with fluorite structure have the largest radiation damage resistance among the ceramic absorber materials. Dysprosium titanate (Dy2-TiO5) was selected for WWER-1000 RCCA. It is used in commercial reactors since 1995. Dysprosium hafnate (Dy2O3-HfO2) has attractive properties too. It is characterized by a more stable fluorite structure, larger efficiency and less swelling in comparison

with Dy2O3-TiO dysprosiuw 2No . m hafnate researc continueds i h . Hafniu s attractivmha e properties too, especially when it is used simultaneously as absorber and cladding material [1].

2. DYSPROSIUM TITANATE

Natural dysprosiu stabl5 s mha e isotopes which have comparatively high absorption cross- sections for thermal neutrons (Fig. 1). Decay products are Ho and Er. All produced radionuclides have

small y-activit smald yan l half-life time. Som eactivity resulte th f irradiatef yo o s d Dy2O3-TiOe ar 2 show Tabln ni . eI

91 162Dy / 163Dy / 164Dy 240 220 2780 barn >— +n bam +n — > barn % 5 5 2 25% 28 2%

+n

FIG 1 Transformation of irradiated disprosmm

TABLE I y ACTIVITY OF Dy2O3 TiO2

Sample Fluence, xIO2 m 2c 2 Time after irradiation Distance y activity (E> MeV1 0 ) day m uR/s Pellets of 0 200±20

Dy2O3 TiO5 09 380 035 100±10 m=25g 080 15115

According to the Dy2O3 TiO2 phase diagram (Fig 2), an interaction between Dy2O3 and TiO2 can result in formation of two compounds They are dysprosium dititanate (Dy2 Ti2O7) and dysprosium tltanate (Dy2 TiO5), the last is polymorphous (at low temperature it has a the rhombic lattice, with rising temperatur t transformi e s int a ohexagona l lattice (~1350°C d thean )n inta o fluonde (~1680°C) meltine Th ) gTiO2 poin Dy 1870°s i 5f possibls o ti t CI obtaio et 2 TiOnDy 5 with different initial crystal structures verifying heatin coolind an g g regime e fluontTh s e structurf o e Dy2 TiO5 has less swelling under irradiation and is more promising to be used as a control rod absorber material During irradiatio pyrochlore nth e structur transmutn eca e intfluonte oth e (Fi) g3

92 2000-

1600

1200

Dy20 70 TiOo

a - rhombic L - liquid P - pyrochlore (cubic) P - hexagonal Ffluorit- e Rruti- l

FIG. Dy2. 2Oi'TiO2phase diagram

2,5-

1,5-

0,5-

a - rhombic Ffluorid- e P — hexagonal P - pyrochlore

FIG. Volume3. variation of phase components of dysprosium tltanate after irradiation (F^lxlO22n/cm2, T=250-450°C)

93 3 e initiaTh l physical efficienc f Dyo y 2-TiO5 (p~5 g/cm WWER-100n i ) 0 environmen% 15 s i t 3 smaller than that of natural B4C (p~ 1.7 g/cm ). On the other hand, Dy2-TiO5 has a higher life-time existenco t e (Figdu f daughte) o e4 . r isotopes characterize y higb d h absorption cross-sectionst I . makes Dy2-TiO5 very attractive for use in control rods. The physical efficiency of Dy2-TiO5 control rod decreased only by 5-6% within 17 years of operation in the research reactor MIR (RIAR) 2.2x10= F ( 22 cm'2 (E>0.1 MeV))

Physical efficiency% ,

. B4C0eff7mm 2. Dy2O3 TiO2 0eff 7.2 mm 3. Hf 0eff 7 mm . Ag-In-Cd 0eff 7.2 mm

0 10000 20000 30000 40000 time, eff.h. FIG. Dependence4. of efficiency of absorbing materials during irradiation WWER-1000in

Some thermo-physical properties of Dy2-TiO5 are given in Figs. 5-7 and the lattice volume increase upon radiatio a temperatur t Fign i nA . .8 f 350°Co e e heath , t capacity (Cp s )equai o t l 0.441 J/gK, the thermal conductivity (X) of powder (p = 4.8 g/cm3) is 0.33 W/mK, the thermal j 6 1 expansion coefficieng/cm2 6. f pellet o s equa= i ) ) p (a o ( (8.5st l 20.15)xlO"± K" . Dy2-TiOs i 5 characterize gooa y db chemical inertness. Dysprosium titanate doe t interacsno t wit austenitie hth c steel claddin t temperaturega t 1000°a s d betwee320°d an an C h 0 Cdurin0 n28 durin. 00 h 6 g1 g14

050

CP, J/gK 045 i 1

040

035

030

025

020 100 200 300 400 500 T,°C FIG Dependence5. ofDyiOs-TiO? heat capacity (Cp) upon temperature

94 0.7

W/mK 0.6

0.5

0.4

0.3

0.2

0.1

0.0 200 300 400 500 600 700 T,°C FIG. Dependence6. of Dy^O^-TiO^ thermal conductivity (A.) = upon temperature (ppowoER 4.8 g/cnf, PH^=0.1 MPa)

A - non-radiated sample 7.5 irradiate- • d sample

7.0

6.5 o 100 200 300 0 50 400 T, °C FIG.Temperature7. dependence of thermal coefficient ofDy^OfTiO^(a) (p=6.2 g/cm1)

Dy-TiO (D-TiO) could be used in control rods as powder or in pellet form. In the first case, 2 5 2 2 7

3 rulea s densits a it , secong/cme 5 equath - s y i n 4 i , j do t l case mors i t i , e tha g/cmn6 . Some resultf so _ •}•} -) Dy2'TiOs pellet powded an s r dimension stability after irradiatio neutroy b n n fluence 3.4x10" cm"" (E>0.1 MeV) are given in Table II. These results show that Dy2TiO5 is characterized by the low swelling rate. The powder does not sinter and freely extract from the cladding, the lattice volume, pellet diameters, length volumed an s s increased (du micro-crackino t e t g1.7%a (Fig) ,9) . 0.5+0.7%, 0.2+0. d 1.2+2.an , 7accordingly% 3% . Dysprosium titanate pellets were cracke fragmented dan n o d pieces of 1 - 5 mm at a temperature gradient of more than AT>60°C/mm. It is a ceramic material characterized by brittleness and low plasticity. But it is not a large problem because these absorber pellets do not interact with the cladding and do not deform it (as is with Dy2-TiO5 powder). The out- gassing absence and good corrosion properties in water at a temperature of 300°C are very important advantages of dysprosium titanate too.

95 A.V/V,,

1,6

1,2

0,8

0,4

0 0123 Fo.1, 10'W2 FIG. 8. Dependence of lattice volume (AV/V/) increase offluorite Dy^O^-TiO^ upon irradiation

TABL DIMENSIO. EII N STABILIT DyF YO 2O3-TiO2 2 (Fmax=3.4xlO" cm' , E>0.1 MeV)

Test Powder Pellets Visual observation freely emptied ,r=345-360°Ce th t a ) a , At=(3-5)°C/mm pellete sar sintering is absent not destroyed, pres claddinn so absents gi ; t=150-500°Ce th bt )a , At>60°C/mm pellete sar destroyed on fragments due to thermal stress, press of cladding is absent Size dimension temperaturt a e =345-360°C - Ad/d=(0.5-0.7)%, AH/H=(0.2-0.7)%, AV/V=( 1.2-2.3)%. X-ray picture a = 5.1692±0.003 A - AV/V=1.7%

. HAFNIU3 M

Hafnium is used in the world (e.g. in the USA) as absorber and cladding material from the end e seventiesth f o . Many attractive propertie f hafniuo s m were already discusse d publishedan d . Nevertheless, ther some ear e factors restrainin more gth e widelf hafniuo e y us contro n mi l rods. These are the relatively high cost, the absence of reliable experimental results on radiation behaviour and hydrogenation, sometimes low mechanical properties, the absence of industrial production in Russia, etc. Fro r poinm ou f view o t , hafniu mmose th remain f t o promisin e son g material r usinsfo waten gi r reactor control rods, especiall uses i absorbet ys i da claddind ran g simultaneously.

Hafnium investigations on various fabrication, technology and composition are carried out in Russia more than 20 years. The main results received up to 1993, were published in the book "Hafnium in Atomic Technics" [1]. Pure hafnium really possesses unique nuclear, physical, mechanical and corrosion properties. Most essential demands made to hafnium control rods without cladding are presented in Table III.

96 IV

2 2 /7G Structure. 9 . of Dy2TiO5 pellets after irradiation neutrony b fluence F-3.4xl(? cm' (E>0.1 MeV) at the temperature 340 - 460°C, x200

TABLE III. MAIN REQUIREMENTS PRESENTING TO HAFNIUM CONTROL RODS

Chemical contents O -0.03 Cr -0.01 M0.0- o 2 H - 0.003 Fe -0.03 Al -0.01 N -0.01 W -0.01 Co - 0.005 C -0.01 Ta - 0.02 Pb -0.005 Cu - 0.005 Nb -0.01 Zr - 1 .0 Ni -0.005 Ti -0.01 bas- Hfe V -0.005 235U -7xlQ-6 Hf+Zr > 99.8% Content additions ga f so s min Grain dimension lOmk< m Texture not

Mechanical properties HM<1900MPa

a aBMP =420 56 0-

a02=270-310MPa

8P>30% Metal Absence of internal cracks and blisters Surface of control rods Absence of micro-cracks after mechanical treatment and absencf eo Cu, N, Ni and W on the rod surface f hafniuo e us me controTh l rods without claddin mors gi e expedient. There exist many positive experimental operation result f suco s h water reactor control rods r instanceFo . , these control rods worked withou problemy an t f morof s e tha e Shippingporyear5 th 1 n n i s t reactor [2]. Since 1991,

97 control rods without cladding wer eresearce useth n di h reactor RBT-6 (RIAR). Their planned lifetime is more tha years0 n2 .

The physical efficiency of hafnium, as well as that of dysprosium, slowly decreases during operation (Fig. 4). Hafnium does not swell too. Its dimensional changes are connected with a radiation growth, as its crystal lattice is hexagonal close packed (hep). The radiation growth deformation can change in a large range, it depends on the fabrication method and the thermo-mechanical treatment (Fig. 10). The lengths of some hafnium samples increased by 3-4% after irradiation by a neutron fluenc f 3.4xl0o e 22cm"2 (E>0.1 MeV t )temperaturea s betwee - 360° 0 theid C32 nan r diameters decreased by 2-3%. Because of the absence of an axial gap in the radiation equipment some of them were curved (Fig. 11).

8,%

3-

2-

1-

22 2 F0,i, 10 cm-

Radiation10. FIG. growth of hafnium 1 - GPhA-1 (t=340 - 360°C); 2 - GPhI-1 (t=280 - 320°C)

FIG. 11. View ofHfrodlets (GPhA-1) after irradiation by neutron fluence F=3.4xl022 cm2 (E>0.1 MeV) at the temperature 340 - 460°C

98 Hafnium structural damage appear changen i s dislocatios it f o s n structure (mainly dislocation loops) due to an accumulation of nuclear reaction products (transmutants), e.g. tantalum (Ta) and lutetium (Lu), and hydrogenation. More than 2% of Ta is contained in hafnium after irradiation by a neutron fluenc f 2-3x10eo " cm"2 (E>0.1 MeV)necessars i t I . tako yt e into accoun developmene th t a t t of hafnium alloys. We suppose that only pure unalloyed hafnium must be used in control rods.

Ther a goo s i ed correlation betwee e dislocatioth n n loop concentratio e hafniuth d man n microhardness (Fig. 12). Dimensions of loops are equal to 25 - 50 A after irradiation by a neutron fluenc f 2x10eo 22 cm"2 (E>0.1 MeV t temperaturea ) s betwee 340°C- 0 n23 , densit 5xl0s yi 16 cm"3.

The hafnium mechanical properties depend widely on a hydrogen absorption and the hydride phase appearance. At the oxidation film thickness 3-20 mm, hydrogen contents in irradiated hafnium t morino s e than 2xlO"2%. Relative lengthenin f irradiatego d hafnium. samples4% - rulea 2 s s a i ,, rine viee th gf Th w sampleo f irradiateso d hafnium after mechanical test showe sar . Fign ni 13 .

, H MPa ,1016cm-3

400&

ssoe

sooe

250U

200&

150& 10 15 20 22 2 F01x10 crrT DependenceFIG.12. of microhardness dislocationand loops upon neutron fluence (E>0.1 MeV)temperaturesan between 340°C- 280

4. DYSPROSIUM HAFNATE

Dysprosium hafnate is considered as a possible material for water reactor control rods. It has a higher neutron absorbing ability and lifetime, in comparison with Dy2O3-TiO2, since Ti atoms are replaced by the second absorber Hf. Hafnate disprosium has also a stable fluorite structure and a higher radiation resistance.

3 Dy2O3-HfO2 pellets (06x1 Omm, p=6,2 g/cm ) were irradiated by a fast neutron fluence IxlO22 nm"2 (E>0.1 MeV heliun i BOR-6e ) th mn i 0 (RIAR t temperaturea ) s betwee 480°C- 0 l n36 Al . pellets saved thei r e initiashapeth d l an sstructur e after irradiation (Fig . e pelle14)Th . t diameter increased by 0.3%. The fluorite crystal structure did not change and the lattice volume increased by 0.6%. Ther s gooi e d agreement wit e resulte volumth hth f o s e measurement e Dyth 2 C>3-HfOf o s 2 pellets. Some dimension stability results of irradiated B4C, in comparison with (n,y)-absorber materials based on Dy and Hf, are given in Fig. 15.

99 20°C 150°C

250°C 300°C

FIG 13 View of irradiated hafnium samples after mechanical tests

100 FIG 14 Structure of irradiated pellets ofDy^Os (F=lxl022 cm \ E>0 MeV,1 (=360 480°Q,- x200

Widely used radiatiow materiaNe nl stability absorbers

15 AV/V, 22 2 F0i=lxl0 cm- t=340-600 °C

Dy2Ti0s B4C Hf

x x^ X X

15 Dimension stability of absorber materials after irradiation in WWER-1000 RCCA

(REC-Ln2O3 TiO2 In -Dy, Gd, Eu Sm)

ABSORBE) y , (n 5 R MATERIAL WWEN SI R CONTROL RODS

At present time the main reasons of WWER control rod degradation and their small life-time m reactors are cladding deformations due to absorber pin swelling and losses of cladding mechanical properties during irradiation Claddings, for instance, from X18H10T and SS304 steels, crack and distract, as a rule, at the deformation more than 0 5% (Fig 16) In using absorber pins which do not swell, the WWER-1000 control rod resource may be increased from 2-5 years, for B4C, to 15-20 years

101 Ad/d, A-B4C

O .-,-, •Dy- 2Ti05 Cladding distraction * - Dy2Ti05, MIR (06X18H10T) oDy- 2TiO5, SM

2 -

0 -

I I 4 8 T, year 12

22 2 0123 F01,10 cm-

Dependence6 7 G F/ of diameter cladding change (Ad/d) with Dy2Os TiO2 uponfluence operating(F(nand ) time (T)

From 1995, RCCAs containing Dy2'TiO5 powder (the botto powdeC m4 B part d r an )(th e upper part e use ar )Russian di n WWERs-1000 and, from 1997, RCCA usee ar n si d C containin4 B d an f gH Ukrainian reactors.

. CONCLUSIO6 N

New (n,y) absorber ceramic materials of Dy2O3-TiO2, Dy2OvHfO2 and Hf for WWER control rods were investigatedf theso e eus materiale Th . s allo o wraist e reliabilit th e safet th f o d y an y

operation and, also, to increase the lifetime of control rods in comparison with rods containing B4C f RCCAo e us se containin WWER-100n Th i f H TiO2 d gDy an 5 s take0ha n place. Test researcd san h on Dy2OrHfO continuede ar 2 .

REFERENCES

[1] RISOVANY, V.D., KLOCHKOV, E.P , PONOMARENKO, V B., Hafnium in Nuclear Technique, Dimitrovgrad, SSC RF RIAR, (1993) 143 pp (In Russian) [2] BECTON, I.M., Post irradiation examination and performance of hafnium as control rod material, Trans American Nuclear Soc , V 39, (1981) 399 pp.

102 BEHAVIOUR OF DIFFERENT BORON RICH SOLIDS XA0053634 AS PROMISING ABSORBERS FOR PWR

D. SIMEONE, X. DESCHANELS, P. CHEMINANT, P. HERTER Centre d'Etude Saclaye sd , Laboratoire d'Etude Materiaus sde x Absorbants, Gif-sur-Yvette, France

Abstract

Absorbing materials are used to control the reactivity of nuclear reactors by taking advantage of nuclear reactions (e g IOB(n,a)7Li) where neutrons are absorbed During such reactions, energetic recoils are produced As a result, radiation damag absorbinn ei g materials originates both from these nuclear reactions frod an ,m elastic collisions between neutrond san atoms This damage eventually leads to a partial destruction of the materials, and this is the main limitation on their lifetime in nuclear reactors Using different techniques, elementary mechanisms responsible of these ceramic destruction are studied The knowledge of such mechanism will help to improve the lifetime of these materials

1. INTRODUCTION

Among pressurized water reactors (PWR) control rods, many of them are made of two kinds of 10 absorbers, an alloy of silver, indium and cadmium, and boron carbide B4C. The isotope B can efficiently absorb low and fast neutrons (Fig. 1). This is why boron carbide is also used in Fast Breeder Reactors (FBR). With the future generation of PWR, absorber pins will need to be able to capture more neutrons. Hafnium diboride HfB2 coul dattractivn thea e nb e materia thesr fo l e reactors, becaus hige th hf eneutroo n absorption cross section botf so h hafniu 10-borond man additionn I . , their low cost of fabrication, their high melting point (Tm=2400°C for B4C, and Tm=3380°C for HfB2), and their low neutron activity after irradiation, makes them attractive materials for the nuclear industry. The evolution of such materials under irradiation is however not enough understood. It is observed that afte typicaa r pelletC lB , burnus5% falf po l apart [1]e mos.Th t likely explanation proposeo ds

fas thai r t this destructio4 purels ni y causee higth hy db stresse s buil arounp u t d micrometric penny- shape bubblee dH s [1]comine H , g fro 10e B(n,cc)mth 7Li capture reaction. However, radiation damages alsy oma pla yrola thin ei s degradation calculationprocesw fe t sbu f radiatioso n damages have been done for materials undergoing nuclear reactions [2]. In addition to the damage produced by elastic collisions between neutrons and target atoms, energetic recoils are a second source of damage, the 10B(n,a)7Li reaction generates heliu lithiud man m atoms wit averagn ha e kinetic energ f 1.4yo V 8Me and 0.83 MeV, respectively.

Two different techniques were used to study the evolution of B4C and HfB2 under a neutron irradiation: the Raman spectroscopy permitted to study the B4C [3] and the "Transmittion Electronic Microscopy uses analys"wa o d t HfBe eth 2 [4].

2. DAMAGE STUDY OF B4C AND HfB2

2.1 Raman study of irradiated B4C

To better understand the degradation mechanisms of irradiated B4C, some samples have been submitted to beams like of 1.2 MeV electrons or 180 keV He+. In these experiments, the only damages in the B4C structure are due to the recoil of the boron and carbon atoms knocked out of their equilibrium positions presence th inside r H o , matrixe f eo eth . Other samples have been irradiatey db neutrons, and the damages do not only come from to the recoil of B and C, and the presence of He, t alsbu o fro disappearance mth f boroeo n isotopes fro e structuree B(n,a)th mth o t e Ldu , i capture 10 7 reaction.

Arguments for using Raman are provided by comparing peaks between B4C and a-boron. The o structuretw e identicaar s l (the same 12-atoms icosahedron structures a C-B- t bu )C median chain

103 absens i d a-boron an i t C exist4 B nn si (Figur e Rama2a)e Th . n spectrum differences betweed an C 4 nB a-boro thif presence preciselo s n i th st o chaint no e r eo y du . This associated with some modelization allowe identifo dt y precisel peakso ytw cm"0 , locate53 cm"0 1,d associate48 an 't da d wit C-B-e hth C chain.

Neutrons

Thermal neutrons E E

dpa 10B(n,a)7Li reaction

V V Li,Te H , Heat

FIG. 1. Neutron absorption by 10 boron

Comparison of the C-B-C Raman peaks between B4C samples irradiated only with electrons or + He (no disappearance of B), and B4C irradiated by neutrons (disappearance of B) shows the material's capacity to restore its structure (C-B-C peaks) when damaged by electrons or He*. This is extraordinare th du o et y self-healing abiliticosaedrone th f yo s whic able har "pumpo et " atoms from the less stable C-B-C chain o complett , e their structures. Thus Rama nirradiateC 4 peakB f o sy b d electrons or He" are not degraded or promptly restored.

Neutron irradiatio s accompaniei n boroy db n isotopes disappearance (Figurcompleto T . 3) e e their structure, icosaedrons "pump" boron and carbon atoms from the C-B-C chain whose structure disappears (and the Raman peaks too) for lack of atoms. During a neutron irradiation, B4C move progressivel a-boron a o yt n structure without C-B-C chain. Onl presence yth f thio e s chain could enable formation and migration of point defects through the structure, thus its absence would block

any defect transfer and consequently the possibility for penny-shaped bubbles in irradiated B4C to relax their surrounding stress defecfiela y db t inflow.

o concludeT , arguments have been brough degradatio C explaio 4 t t B e th n n durin ga neutro n irradiation initiae th e buildin: th l o damagt f penny-shapee go du s ei d helium bubbles, associated with

104 high local stress fields. At the same time an aggravation is occurring inside the matrix, the materials becoming progressively unabl releaso et e thes disappearance th stresses o t e du , f o epoin t defects and/or their mobility.

FIG. 2a. Crystallographic structure ofB4C

FIG. 2b. Crystallographic structure ofHfB2

Figure 2a shows a schematic description of the Figure 2b presents a schematic representation of the HfB, rhomboidal unit cel Bf o l4site CC . s refe carboo t r n atoms unit cell. Hafnium atom whitee sar s dots situate t eacda h of the linear C-B-C chain, B site refers to the inversion vertex of the hexagon, and three boron atoms over six are center occupie boroa y db n atom i refeh , polao t r r sites drawhexagoe th n i interstitian i l positions. occupied by boron atoms and h2 sites refer to boron and carbon atoms occupyin equatoriae gth l sites.

FIG. Crystalographic2. structures ofB HfB4Cand 2

2.2. MET study of irradiated HfB2

e symmetrTh y grou f HfBpo P6/mms i 2 m (Figure 2b) e uni Th .thexagon n cela s i l e possessing hafnium atoms at each vertex of the structure. Boron atoms occupy vacancies positions inside the hexagon. TEM photographies of HfB2 irradiated under 1073°K at low 10B burnup (about 1020 captures/cm3) show clearly dislocation e basaloopth n li s plans [4] r hig.Fo h temperature values (above 1073°K) and similar IOB burnup, helium bubbles appear in the grain boundaries of the material. Therefore, helium bubblet penny-shapeno e ar s t sphericabu C d4 B lik l n i elik UC>2n i e o N .

105 strain fields are associated to these bubbles in HfB2. The destruction of the material is due to the agglomeration of basal dislocation loops (vacancies loops) which meet each other near the grain boundaries forming micro cracking.

As a conclusion for HfB2, in this material helium bubbles are not damaging thanks to the defects mobilitye anisotropth t Bu .f relocatioo y f theso n e defects induces anisotrope th f o y elementary grain's strain thu threasa material'e th o dt s integrity.

ODPA

0.4 DPA

0.7 DPA

C o o

A 15DP

0 80 120 0 40 0 1600 Raman shift (crrrl)

FIG. 3. Evolution of Raman spectra ofirradiated B4C samples at different dpa values

CONCLUSIO. 3 N

differene Th t P6/mmunid an t differene celC 4 HfBr th m B l fo d r symmetr2an )fo t m natur 3 R y( e of the chemical bonds leads to opposite behaviour under neutron irradiation of these two materials : I n2 icosaedri BBi 4e Cth , c cluster t permino appearine o th sd t f vacanciego diminiso st h strain fields around helium bubbles, which remain damagin materiale th r gfo . In HfB2, many vacancie presene sar allod an t w helium bubble materiale relao st th n xi . irradiatiow lo t A n temperatures (under 1073°K) e higth , h numbe f vacancieo r s loops (basal loops responsible )ar microcrackinf eo g apparitio materiale th n ni .

REFERENCES

. STOTOT ] . ZUPPIROLI[1 L , . PELLISSIERJ , , Radiation Effects , (198590 , ) 160-170. [2] A. ALBERMANN, D. LESUEUR, Am. Soc. for Test, and Mat, PA 19103, (1989). . SIMEONED [3] . MALLETC , , DESCHANELSX , . BALDINOZZIG , . KAITASOVO , , Studf yo 0 uni84 t cell evolution under neutron irradiatio Ramay nb n spectroscopy, submitte Physn di . Rev. B,( 1998). . CHEMINANTP [4] ThesisD Ph , , Orsay, (1997).

106 REINFORCEMENT AGAINST CRACK PROPAGATION OF PWR ABSORBERS BY DEVELOPMENT OF BORON-CARBON-HAFNIUM COMPOSITES

. PROVOTB . HERTEP , R Laboratoire d'Etude Materiaus sde x Absorbants, XA0053635 CEA, Saclay, Gif-sur-Yvette, France

Abstract

In orde improvo t r mechanicae eth l behaviou f materialo r s use s neutroda n absorber nuclean i s r reactors have w , e developed CERCER or CERMET composites with boron and hafnium. Thus a new composite B4C/HfB2 has been especially studied have W . e identified three kind f degradatioso n under irradiation (thermal gradient, swellin fissioo t e gndu products and accidental corrosion) that induce imposed deformations cracking phenomena. Mechanical behaviour and crack propagation resistance have been studied by ball-on-three-balls and double torsion tests. A special device was developed to enable crack propagatio associated nan d stress intensity factor measurements. Effect structurf secona o s f o d d ean phas e ear underline. First results show that these materials present crack initiation and propagation resistance much higher than pure boron carbid hafniur eo m diboride observe W . e R-Curves effects, crack bridgin r branchinggo , crack arrests toughnesd an , s increases that we can relate respectively to the composite structures.

1. INTRODUCTION: DAMAGES IN NEUTRON ABSORBERS DURING IRRADIATION

Neutron absorbers in nuclear control rods consist in boron compounds in metallic cladding. The neutron capture by boron atoms leads to the creation of helium atoms which cause an important 20 3 3 3 swellin materiale th f go s (~0,1 10r 5 pe vol captures.cm"% . correspondin He/cm"cm 0 8 Bo gt 4 C) [1] according to the reaction:

10B + n -» 4He + 7Li + 2,6 MeV.

This reaction causes material damagatoo t me edu recoil , swellin thermad gan l stressesn I . pressurized water reactors (PWRs), a high flux depression involves a differential irradiation induced swelling along the pellets radius and thus induces high stresses at the pellet's periphery, but no thermal stresses. Associated with neutron damage, this leads to fragmentation starting from the surface of the material.

In Fast Breeder Reactors (FBRs), contrar PWRso yt , neutrons penetrat pelletse bule th eth f ko . This leads to an homogeneous volume heat release, a temperature gradient of about 1000°C between pellet's centre and surface and tangential stresses, which largely overcome material's strength, causing premature ruptur absorbere th f eo s (radial cracks). Associated with bulk neutron damage, this lead completa o st e fragmentation [2].

In both typ reactorsf eo , absorber fragment segregaty ma s gravit y bottome b rode th th o yst f o or stick between pelle cladd an t . Associated with irradiation induce swellinge dH , thiinducn sca ea significant damag absorber'e th o et s cladding PWRsn I . seaa f i cla,e l losth d f so occurs , pellety sma be corrode primare th y db y coolant. Corrosion pits induce then crack initiations leadin prematuro gt e failure [3].

All these absorber phenomena involve imposed deformation cracking mechanisms: thermal dilatation (FBR), oxide phase dilatation (PWR corrosion) and irradiation induced swelling (creation pelletsR FB d ) an (se heliuR ef o FigurPW mn . i e1)

This work displays the mechanical study of several composites based on boron carbide or hafnium diboride matrixes. The purpose is to characterize the materials (strength, rupture modes and crack propagation resistance identifo t d an )y reinforcement modes towards crackin comparison gi n with associated single phase materials.

107 FBR Thermal stress and damage PWR Peripheral swelling and damage Corrosio __coolany nb _ t water

FIG Damage1. modes of pellets during irradiation

2. DEVELOPMENT OF MATERIALS

The aim is to give to our materials sufficient reinforcement to stop or minimize crack propagation and to enable them to conserve a sufficient cohesion though a high mechanical damage. The reinforcemen a cerami f o tn tak ca ce different routes. Most frequently a secon, d phass i e introduced (meta r ceramico l differend an ) t mechanism takee sar n into accoun tcermet a [4] n I . e th , metallic component (continuou disperser so matrixa n di ) confers globally some ductility case th en I . of a reinforcement by ceramic inclusions, we try to create a network in which cracks will propagate with difficulty A process zone can develop in the vicinity of the crack tip leading to a significant decreas crace th f keo driving force e morTh . e frequent mechanisms actin procesa n gi s zone ar e residual stresse differentiao t e sdu l thermal expansion coefficien r phaso t e transformation r microo , - crackin differentiao t e gdu l elastic constants.

Several composite materials have been fabricate t uniaxiaho y b d l pressin f mixego d ceramic

and/or metallic powders. Fine micrometric HfB or BC powders have been mixed with fine powders 4 like metallic hafnium [4] or ceramic hafnium oxide2 , or particles of hafnium diboride [5, 6]or metallic hafnium. Tabl eI show differene sth t studied materials obtainee W . maio dtw n grade f materialso s : fine grained homogeneous composites named "microdispersed" and heterogeneous composites in which reinforcement is constituted by large secondary phase aggregates or inclusions named, "macrodispersed" (by adjusting the hot pressing parameters, macrodispersed materials are obtained by coalescence from microdispersed powders).

TABL DIFFERENE. I T TYPE MATERIALF SO COMPOSITED SAN S STUDIED

Matrix Symbol Additive Second phase after sintering

HfB2 A

Al Hf micrometric powder microdispersed HfB, HfBC 2Hf ,

A2 HfO2 micrometric powder microdispersed HfO2

B4C B

Bl micrometric HfB2 powder macrodispersed HfB2.d

B2 HfB2 submillimetric agglomerates HfB2 submillimetric agglomerates macrodispersed B3 Hf submillimetric agglomerates Hf submillimetric agglomerates macrodispersed

108 . EXPERIMENTA3 L PROCEDURE

Ball-on-three-ball1 3. s biaxial bendin] g 8] tes , t[7

The load is applied thanks to a central ball on the upper side of a thin disc (Figure 2). The disc is supporte lowes it n dro side peripher threy yb e balls regularly distribute concentria n do c ringn I . this case, radial and tangential stresses are equal and maximal at the centre of the sample, and are expressed by an analytical simple plate theory formula of the form:

geometry), v ( GF ™ a"x = eP "=

Poisson'Pe th bein v loae d gth d an s ratio estimation .A Youne th f no g modulu alss si o obtained from the central deflection w. Thus measurement of the critical load at the sample's collapsing enables to calculate material strength.

This test is particularly well adapted to characterization of very hard and brittle ceramic material s boroa s n carbid r hafniuo e m diboride gived an , s better results than 3-point r 4-pointo s s bending tests (These tests are however kept for samples with strain gauges for measurement of Poisson's ratio).

KNOOP indentation

FIG. 2 Ball-on-three-balls bending test FIG. 3. Double torsion test

3.2. Double torsion test [9]

Double torsion test allows to measure ceramic toughness Ktc. Samples are thin rectangular plates in which a notch is prepared (Figure 3). The notch is submitted to bending and the crack propagate e notch th e sampl f o Th . p ssupporten i eti fro e mth y foudb e loar th s alsballi d d oan s applied witbeginnine balth h a notchn e lo stresth e f go .Th s intensity facto proportionas i r loae th d o t l P and depends on geometry and elastic constants. It is expressed by analytical formula:

The critical value KIC, where the crack propagates, is then obtained by a formula like:

G(elastic_COnstantS,x P = C I K c geometry), criticae wherth s i c l eP loa d (Figur. e4)

determinn Thuca e s w evolutio e eth f stresno s intensity facto functioa s a r f cracno k lengta hvi the load. In the case of a brittle homogeneous material, crack propagates for a constant value of critical load (Figur f cracI . k5) e shielding occurs (deflection, bridging, arrest) e deflection-loath , d curve is perturbed, with (Figure 6) or without (Figure 7) global reinforcement. In this second case, material demonstrat R-Curvea e effect i.e increasinn a . g toughnes cracs sa k propagate materiale th n si .

109 R-Curve (reinforcement)

\, 36 - X R-Curvo N e -5 A -/I ;racktem m nh jt (no reinforcement) ^ 9

5 10 15 20 25 30 0 2 5 1 10

R-Cvrve4. effect FIG Brittle5 homogeneous material

200 120 - - AW. | ® /f^^^l 100 - OA *L ^ ! ! 60 - / ! 40 - / i 20 -i/ ? ——————— del tection um -t E I • —-—---- —- 0- . 0 3 0 2 0 1 0 25 50 75 !

FIG 6 Crack shielding without reinforcement FIG. Crack7. shielding with reinforcement

3.3 Thermal gradient crack test

This test consis applyinn i t ceramia o gt c thin dis thermaca l gradien focusiny b t g wito htw elliptic reflectors high power halogen bulb light at the centre of the sides of the sample which is cooled at its periphery. The parabolic like temperature of the pellet's surface is measured by thermocouples and the light power is measured with a power meter.

Testing no-notched discs lead thermao st l strength determination materiale Th . submittee sar d to increasing thermal gradient, so far as they crack and/or break. Measurement of thermal gradient lead thermao st l gradient strength accordin classicao gt radiaD 1- l l analytical formula.

Testing notched discs lead thermao t s l gradient toughness analytican A . l method basen o d superposition principle in mechanics leads to toughness calculation via a Gauss-Chebychev resolution. Thu t cracsa k start, toughnes obtaines i K sK d throug analytican ha l formula dependinn go crack length, geometry, material's thermo-mechanical properties lamd an , p power e toughnesTh . s calculation confirmes i s finita y db e element calculation (CASTEM 2000). Finite element latee sar r use cracr dfo k propagation calculation.

This test offers the possibility to proceed to a step by step propagation experiment, where the sampl s removei e d examinean d y scanninb d g electron microscopy (SEM t )eaca h step. Items obtained are: crack length, reinforcement mechanisms (crack arrest before, inside or after an inclusion interfaco t n a ot ra e inclusion-matrix), calculatio toughnesf no t cracsa k starcracd an t k arres orden i t r to quantify reinforcement mechanism like:

composit matrie , xT ^ inclusion C 1 ~ ~ Ai\K.1C v A

110 At last, we can obtain a curve describing a global reinforcement as a function direct measuremen cracf o t k length identifying each respective reinforcement mechanisms (see Figur. e8)

Composite HfJEWHfO; crack branching Composite H£B2/Hf crack smcicung (micromdentauon) (thermal gradient test)

3S1881 ISKV X4.88K 7.Sue

Composite B^C/macodispersea HfB2 crack deflection [ syadient test)

FIG. Different8. crack morphologies obtained

4 RESULTS AND DISCUSSION

Strengths and toughness have been respectively measured by ball-on-three-balls [10] and double-torsion [11] tests. Result thesf so e experiment reportee sar strengte e Tabln db Th i . n I ehI ca strongly affected, especially in the case of "macrodispersed" composites because of the change in critical flaw size. However, strength is not a critical material parameter versus the loading conditions during o avoit pellet t dno s crac s i irradiation km ai initiatio e o avoit Th t . dbu n fragmentation. Toughness or crack arrest are certainly more relevant parameters for the considered application.

Optical and SEM observations, on polished or fractured surfaces reported in Figure 8, allow the mechanisms of crack propagation and of crack arrest to be understood. In all composites, Al excepted, the crack propagation is disturbed by the second phase. Deflection by the inclusions, shielding by the particles and branching of the crack can be observed. Inclusions, hafnium-rich phases, residual stressed areas and microcracked particles [12] play a role. The composite inclusions delay or sometimes stop simple intragranular brittle cracking of HfB2 or B4C matrix. The efficiency of these reinforcement phenomena is confirmed by the measurements of the mechanical properties Compared with single-phased materials e fracturth , e toughnesn A . s % increase 2 6 o t sp u fro 1 m3 R-curve is also sometimes observed, probably because of crack bridging by particles. Only A2 composite does not present a R-curve effect because of the small size of inclusions. B2 composite exhibits moreover a dissipative behaviour during strength measurements probably due to the presence of microcracked aggregates.

l casesInal , these results confir possibilite mth delao yt y complet prematurd ean e collapsf eo ceramics

111 TABLE H. TEST RESULTS OBTAINED BY BALL-ON-THREE-BALLS BIAXIAL BENDING, DOUBLE TORSIO THERMAD NAN L GRADIENT TESTS

Symbol Brittlr o ) e(B R-Curve Young Strength Weibull K,c Dissipative effect modulus (MPa) coefficient (MPa.m'1/2) (D) rupture (MPa) m A B no 418 398 10 3,1 Al B yes 416 688 7 4,8 A2 B no 405 527 16 5,1 B B no 310 497 7 3,6 Bl D yes 242 264 17 5 B2 D yes 207 176 10 5 B3 D yes 147 120 5

REFERENCES

[I] ZUPPIROLLI al.t e , , PhilL. , . Mag, 60-A 5. (1989) 539-551. [2] STOTO, T., et al., J. Appl. Phys., 68-2 (1990) 3198-3206. [3] DESCHANELS, X., Key Eng. Mat, Vol. 13 (1996) 189-198. ] [4 EVANS Ceram. , Am A.G. . J . ,Soc. , 73-2 (1990) 187-206. [5] GOSSE al.t e , , JJAPTD. , Serie (19940 s1 ) 216-219. [6] CHEMINANT, P., et al., Key Eng. Mat., Vols 132-136 (1997) pp.643-646. [7] KIRSTEIN, A.F., J. of research of the Nat. Bur. of Stand. - C Eng. and Instrumentation, Vol. 70C, N°4, (Oct-Dec. 1966). [8] DE WITH, G., J. Am. Ceram. Soc., 72 [8] (1989) 1538-41. [9] BOUSSUGE, M., PhD Thesis ENSMP, Paris, (1984). [10] DE WITH, G., et al., J. Am. Ceram. Soc., 72-8 (1989) 1538-1541. [II] FOURNEL PhesisD Ph , , B. Ecol, e Nationale Superieur Mines ede Parise sd , Paris, (1991). [12] FABER, K.T.al.t e , Proc. Int. Conf. Fract, 4th, (1977 )529-56I .

112 NSC KIPT'S EXPERIENCE IN PRODUCTION OF ABSORBER MATERIALS, COMPOSITE PRODUCTD SAN CONTROR SFO L MECHANISM VARIOUF SO S NUCLEAR REACTOR TYPES

N.P. ODEYCHUK, V.F. ZELENSKY, V.A. GURIN, YU.F. KONOTOP National Science Center, XA0053636 Kharkov Institut Physicf eo Technologyd san , Kharkov, Ukraine

Abstract

Data on NSC KIPT developments of absorber composites B4C-PyC and B4C-SiC are reported. Results of prereactor studie reactod san r test f absorbeso r composites develope givene dar shows i t I . n thaBe C-Pyth t C composites hav higea h

radiation resistance at temperatures up to 1250°C. 4

1. INTRODUCTION

The NSC KIPT has been engaged in the development of absorber composites and products base then do r reactivit mfo y compensation system element f high-temperaturso e gas-cooled reactors and fast reactors since the sixties. The processes developed to produce absorbing composites and item originae s relth n yo l technolog methodye makinth f o volumetrif e so g us phass cga e densification of porous media with pyrocarbon [1-6]. This paper reports the results from studies into basic characteristics of B4C-PyC and B4C-SiC produced by the gas phase technology, and also, their behaviour under neutron irradiation conditions.

2. B4C-SiC COMPOSITE

Potential corrosion-resistant materials, capabl f operatineo oxidizinn a n gi g environment under reactor radiation condition t temperaturesa s betweed 1300°Ce compositean th 0 e ar n80 , s basen do silicon carbide (fast reactor design wit hdissociatina g coolant N2O4) [4].

KIPC NS T e grou Th developes pha d absorber composites comprisin gdispersioa f borono n particles in a carbide-silicon matrix (B4C-SiC composite) for nuclear reactor control rods. The absorber materials of this type are produced in the process of binding a mixture of silicon and boron carbide powders with pyrocarbon. After bindin d mechanicaan g l e requiretreatmenth o t dp u t parameters, the products obtained undergo heat treatment in vacuum at a temperature of about 1500°C. In the process, silicon interacts with a pyrocarbon binder to form a carbide-silicon matrix which comprises boron carbide particle inclusionss sa .

The corrosion resistance of B4C-SiC composites was studied for several thousand hours at temperature 1300°o t p su airCn i , water vapour environment othed san r reactiv medias ega . With time, the oxidation rate decrease weighe th d f specimeno tsan s becomes invariant carry-oveo N . borof o r n and failure of specimens were observed in the course of corrosion tests. Table I gives the reactor test result absorber sfo r composites B4C-SiC resulte [4]Th . s bear witnes higa o st h radiation resistancf eo the B4C-SiC composite.

methoe Th d devise manufacturo dt e B4C-SiC products make t possibli s mako et e intricately- shaped contro elementsd lro , including units with threaded joints. Figur showe1 s some products made basie on th Bf so 4 C-SiC composite [4].

113 TABL . REACTOEI R TEST RESULT BR 4C-SISFO C ABSORBER COMPOSITE] S[4

Description of specimens Irradiation conditions Variations in size and volume after irradiation Variations

B4C B4C Total in Dimensions content particle density Temperature Fluence 10B burnup Diameter Length Volume strength (natural) size

% mm g/cm3 °C n/cm2 % % % % % 1 2 3 4 5 6 7 8 9 10 11 4x4x40 mm 0 80 1.65 270-300 7.1020 - - +0.360 - - 9 1.72 +0.317 16 1.82 +0.242 32 1.80 +0.242 47 1.87 +0.262 58 1.84 -0.015 2x6x24 mm 46 80-120 1.75 550-650 1.2.102' 50 ~5 1.75 (thermal) 20 020mm 45 80-120 1.95 760-1070 3.6.10 6.5 +0.22 -0.28 -0.15 -10(

Be 4C-PyTh C absorber composit producte th basis it d en an s so were develope KIPC NS T t da for application in highly stressed reactivity compensation systems, mainly, in high temperature gas- cooled reactors (HTGR) [2, 4-6]. In particular, the VGM facility, designed in the former SU with the pebble-bed HTGe basisth s R,a provide r threfo d e independent reactivity compensation systems (RCS): a) control and shield rods (CSR) that can shut down the reactor and maintain it in a subcritical stat worse casn th ei f eto possible acciden emergencr t fo [6] ) b ; y shutdow additionan na l independent designeds wa bases S bal wa fillin n t do i RC l, g special channel laterae th n si l reflector with absorber elements, < 10 mm in diameter; c) at the initial phase and when the reactor is brought to the equilibrium operating condition e excessivth s e reactivit s compensatei y y sphericab d l absorber elements (SAE) wit outeplacen e ha ar thes; rE ddiamete edirectlmm SA 0 core 6 th f ero n y i togethe r with spherical fuel elements and empty (fuel free) elements.

As an absorbing material, the absorber elements of all three systems make use of B4C with natural compositio specifie boronn ni th t ye ,c conten f absorbeo t r being different. Thus exampler fo , , 3 3 the B4C content in the SAE matrix graphite is about 4.10" g/cm , and in small diameter SAE and in 3 CSR rings is > 1.3 g/cm . The NSC KIPT technology of manufacturing the B4C-PyC absorber composite and its products, relies on the methods of volumetric gas phase saturation of porous media with pyrocarbon [1-5], namely, powdeC bindin4 B e rgth with pyrocarbo r mixinn(o e graphitgth e powder with the required content of 640). During the process development, both pilot specimens of material productd san researcr sfo h purpose batchepiloE d san SA t s (2500 piece criticar sfo l stand "Astra" wels (200) a S bal s a l lRC 0 piece OKBr sfo M stands) were manufactured.

metallographie Th c section f materialso s with increased absorber contents clearly exhibie th t B4C particles bound with pyrocarbon into a monolith (B4C-PyC composite, see Fig. 2) [6]. The pyrocarbon binde s higha rh strength characteristics [6], that eventually ensure higa s h mechanical strength of the composite. The main characteristics of B4C-PyC absorber composites are presented in 3 Tabl (1.I eI 6 g/cm contentC 4 B ) [2].

FIG, Microstructure2. ofB^C-PyC absorber composite 3 with a specific B4C content > 1.5 g/cm , x 75 [6].

116 TABLE II. MAIN CHARACTERISTICS OF B4C-PyC ABSORBER COMPOSITES 3 (1.6G/CM B4C)

Characteristic Y end X a g/cm3 MPa MPa W/m7 10-67-l Value 2.1-2.2 300-330 80-100 10-17 4.8-5.3

s seeIi t n fro e Tablmth e thae strengt th e B4C-Pyt th f ho C composit s ratheei re highth s A . annealing temperature rises thermae th , l coefficien f lineao t r expansio electricae th d nan l resistance values decrease, and the heat conduction of B4C-PyC is nearly doubled (Figs. 3 and 4).

!

1000 JMC

FIG. 3. Effect of annealing temperature FIG. 4. Thermal coefficient of linear expansion 3 on thermal conduction (I) and electrical (a) ofB4C-PyC with B4C content of 1.6 g/cm resistance (r) ofB^C-PyC composite [6] versus measurement temperature [6]: 1 - initial sample; annealing- 2 1400°C;at 3 - annealing at 1300°C; annealing- 4 2000°Cat

specimene Th f productso s were irradiate higa n di h flux reacto modee E rTh SM-. SA l 6] , 2[4 diamete loadine th d theif rgan o varie m r0 40m m d9 m4 betwee d kernelan 5 n4 s with boron carbide was 3.9xlO"3, 3.6xlO' 1.9x10"d 2an ' g/cm3, i.e variet .i factoa y db abouf ro . 50 t

3 The specimens with an increased B4C content (> 1.3 g/cm ) were of three types: cylindrical mm0 4 ;x 6 :0 bushings: 0 = 45 mm, 0 = 25 mm, H = 40 - 48 mm; sectors cut from the bushings at an angle of 120°.

Cylindrical specimen f matrio s f industriallo xd graphitan P y GS e made hot-sintereC 4 B d (produced by pressing the powder followed by sintering) were used as reference specimens.

To study the effect of heat treatment on the radiation deformation of B4C absorbers, part of the specimens were annealed at temperatures between 1600 and 2000°C. The post-reactor examination reveale underwenE dSA thae th t t isotropic shrinkage e valuth , f whiceo s independenhwa C 4 B f o t content (ranging from 3.9xlO" o 1.9X10"t 3 1 g/cm3), irradiation temperatur d burnuan e f 10o pl B .Al

117 experimental points showing relative diameter variations (AD/D e withili ) e rangth n f scatteo e r presente Fign di .goon [6] 5 i e variation e d.Th ar agreemen E f AD/o sSA r f(F= Dfo )t wite hth variation Al/f so matrif(F1r = fo graphit)P xGS e specimens curv(see th e Fig. n ei 5) .

FIG. 5. Radiation deformation (AD/D) of spherical absorber elements with B4C content of 0.004 g/cm3 (O, f, 9); 0.036 g/cm3 (A) and 0.19 g/cm3 (Q versus fastneutron fluence [6] O, A, D- initial spheres; ™- annealing 1600°C,at hours;6 annealing- • 2000°C,at hours.6 (Figures near encircledthe regions denote irradiationthe temperatures; curvethe shows deformation of matrix graphite GSP specimens irradiated at 1100 - 1200°C)

t shoulI notee db relatinburnuE B d5 IO 5 SA tha s foune e region o pgi b t tth Fig wa n i o d t 5 . n C 3 3 witcontenE C 4 hB SA r f 3.9.10'o tfo 0.1d % an9an d21 g/cm , respectively irradiateE SA n a I . o dt high fluence of neutrons (see region D in Fig. 5), practically a complete burnup of 10B nuclei took place. The annealing of SAE at 1600 and 2000°C for 6 hours has no appreciable effect on the AD/D value, and the scatter of experimental data can be due to both the structural peculiarities (density, pyrocarbon content) of individual SAE and their irradiation conditions.

The SAE, absorber composites based on B4C dispersions in GSP with B4C (natural) content up g/cm6 to1. 3 were also teste widn di e range temperaturef so s (30 1250°Co 0t . 6] , fluenced 4 , )an 2 , [1 s All the materials have exhibited an extremely high radiation resistance. Even for the materials with a 3 10 B4C content of 1.6 g/cm , irradiated at 1200 to 1250°C to 90% burnup in B, the dimensional changes were no more than 1%. After irradiation, the B4C-PyC specimens remained intact, no spalling, cracks or other defects were observed e samth t e A .time t similaa , r irradiation condition hot-sinteree th s d

B4C specimens showed swelling: AV/V - 0.75 - 1.5% [6].

s knowIi t ] thaspecimene [6 nth t f hot-presseo s r hot-sinteredo exhibiC 4 dB t swelling under irradiation, and the swelling value drastically increases with increasing irradiation temperature and burnup depth particularn I . , burnuB wit10 h a t irradiatiof abou a p o % 5 t n temperatur f 900°eo e Cth AV/V ratio reaches 30%. Therefore resulte th , f investigationso s into radiation deformatio f typno e B4C-PyC absorber composites ,y uniquereporte wa e near a Th . n -i dn [6] i zere herd ,ar oan e deformation of the composites under study can be attributed to a competing action of two factors: swellin particleC 4 B shrinkagf d go san pyrocarboe th f eo n matrix [6].

The present results suggest that further studies are needed to clarify the applicability of B4C- PyC absorber composite insteae vibrth a f f o do o R compacte CS pard f en o tpowde C e 4 dB th n i r WWER-1000 reactor (Fig. 6).

118 upper- part; 1 end upper- 2 weld; 3 - compensation volume; 4 -plug -fixing catch of absorber material; B- 4C-PyC5 absorber composite; 6absorber- element cladding: 08.2 x0.6; Q8X18H10T; 7lower- weld; 8 - lower end part

FIG. Design6. ofWWER-1000 absorber element proposed furtherfor studies

119 4. CONCLUSION

So, the bench and reactor tests undertaken here allow us to draw the following conclusions:

Be 4C-PyTh B1d .4C C-Sian C absorber composites have high strength characteristics; . 2 Radiation deformatio f Bno 4 C-PyC absorber composites wit increasen ha contenC 4 dB t (> 1.3 g/cm3), irradiated at temperatures between 1200 and 1250°C to a 90% burnup in moro n s e 10tha; wa B, n1% . 3 Irrespectiv higa f eho damagin matrixe th f go , estimate somn di e experiment aboue b o st t 30 dpa [6], the absorber composites under study show an extremely high radiation resistanc thid esan makes them promisin t onl HTGr gno reactore yfo th alst r Rbu o fo f so other types; recommendes i t I . 4 investigato dt possibilite eth usinf yo Be g4th C-Py C absorber composite instead of the vibration-densified B4C powder in the end part of the CSR of the WWER- 1000 reactor.

REFERENCES

[1]. GURIN, V.A., ZELENSKY, V.F., "Gas-phase methods of producing carbon and carbon-carbon materials", Voprosy At. Nauki i Tekhniki, Ser. Fiz. Rad. Povrezhden. i Rad. Materialoved., Kharkov KEPC ,NS T publ. 3(69). is , , 4(70), (1998) 83-8 Russian)n 5(i . [2]. ZELENSKY, V.F., GURIN, V.A, KONOTOP, YU.F., ODEYCHUK, N.P., et al., "Spherical fueabsorbed an l r element HTGR"r sfo , Vopr . NaukAt . i Tekhni . Ser., Fiz. Rad. Povrezhdeni . Rad. Materialoved., Kharkov, NNTs KhFTI publ., is. 5(71), (1998) p.40 (in Russian). [3]. ZELENSKY, V.F., GURIN, V.A., KONOTOP, YU.F., ODEYCHUK, N.P., et al., "GSP graphite", Vopros . NaukyAt i Tekhnikii . Ser. Fiz. Rad. Povrezhden i Rad. . Materialoved., Kharkov, NSC KIPT publ., is. 3(69), 4(70), (1998) 86-87 (in Russian). [4]. IVANOV, V.E., ZELENSKY, V.F., TSYKANOV, V.A., et al., "Dispersive fuel and absorber elements based on pyrocarbon-bound graphite for high temperature gas-cooled reactors", book: Reaktorn. Materialoved. (Proc Reactof o . r Materials Science Conf. June1 ,- Alushta,y Ma 9 2 , 1978), V.6, Moscow, TsNIIAtominform publ., (1978) 308-325 (in Russian). [5]. GURIN, V.A., ZELENSKY, V.F., EVSEEV, V.M t al.e . , "Developmen f monolithic-typo t e pyrocarbon-bound absorbefued an l r element r HTGR"sfo , book: Nuclear hydrogen energetics and technology, Moscow, Ehnergoatomizdat publ., (19835 . is , ) 213-22 Russian)n 5(i . [6]. ZELENSKY, V.F., GURIN, V.A., KONOTOP, YU.F. t al.e ,, "Radiation resistanc absorbef eo r composites with a pyrocarbon binder", (Proc. of the Int. Conf. on Radiation Materials Science, Alushta, 22-25 May, 1990), Kharkov, KIPT publ., v.8, (1991) 68-76, (in Russian).

120 DAMAGE ANALYSIS OF CERAMIC BORON ABSORBER MATERIALS IN BOILING WATER REACTORS AND INITIAL MODEL FOR AN OPTIMUM CONTROL ROD MANAGEMENT 1 W. SCHULZ XA0053637 Preussen Elektra Aktiengesellschaft, Hannover, Germany

Abstract

Operating experience has proved so far that BWR control rods cannot be used for the total reactor life time as originally presumed, but instead has to be considered as a consumable article After only few operating cycles, the mechanism of absorber failure has been shown to be neutron induced boron carbide swelling and stress cracking of the absorber tubes, followed by erosion of the absorber material In the case that operation of such a control rod is continued in control cells, this can lead to an increase of the local power density distribution in the core and, under certain conditions, can even cause fuel rod damage A non destructive testing method has been developed called "UNDERWATER NEUTRON RADIOGRAPHY controR " applicabld "Lead-controBW ro ly an r efo l rods" being radiographe e use dar evaluato dt e their actual nuclear worth by the help of a special analytical procedure developed and verified by the author Nuclear worth data plotted against bur historp nu y data will allo creatwo t "EMPIRIn ea C MODEL" This model include e basith s c idef ao operating control rods ofcertaina design controla first in positiontargeta to fluenceup limited amountan to just below the appearance of control rod washout Afterwards they have to be moved in a shut down position to work therefor the total remaining holding period The initial model is applicable to any CR-design as long as sufficient measuring-data and thus data abou e nucleath t r wort available har resulte Th e f thesso e experience extrapolatee ar s whole th o dt e reactor holding period After modellin furtheo gn r measurement thif so s particular contro typd necessare ro le ar reacto> secone an n yTh i r d focal poinprovido t s i t APPROXIMATIOn ea N EQUATIO knowiny B N absorbee gth r radius densitC 4 B .absorbed yan r enclosure data an engineer will calculate reliably the working life of any control rod design on control position, indicated as maximum allowable neutron fluence margin until absorber wash-out starts This concep calculato t - t e controeth l rod's working life both in the control position and the shut-down position - will automatical!) lead to an optimization of the control rod strategy Control rod optimisation is demonstrated by accumulating the total amount of control rods required in a medium-sized BWR up to the total reactor holding period At least 60% of the first core inventory - for this control rod type an existing EMPIRICAL MODEL is already available - may be used up to the total operating period without any safety loss Lookin presene th o gt t disposal situation this concept represent practicasa reduco t y l higeal wa l h level wast addition I e n benefit of utilizing this concept is that it minimizes tritium emission Control-rods utilized within Boiling Water Reactors (BWR e designe ar e purpos) th r fo do controt e d shap an e neutrol eth n flux profile reactorth n i eo adjus t , e rangth tf o e regulation referring to the weight rate of the reactor coolant and thirdly by- shutting down the reactor at any time and under any conditions with regard to nuclear aspects, mechanical integrity and control rod history The designation control- or shut down rod characterize the particular field of activity for a given control rod The focal point of my work had shown to be a calculation of the nuclear working life of any control rod design as well as an optimisation method with reference to the holding perio a give r fo dn contro d inventorro l resula s f a ymeasurino t ga theoretica datd an a l analysis describine th g parameters in a general validity form

1. DESIGN VARIATIONS

e controTh l rods typically use boilinn di g water type reactors (see Figurprovidee ar ) 1 e d witha cruciform shape and pass in the interstices between fuel channels. The typical basic product line of the GE control rod (standard control rod) is shown in the Figure 1 Four absorber blades are arranged in order to form the required cruciform shape. Every blade is filled with up to 21 small absorber tubes placed in side by side relation along the axis of the cruciform shaped control rod containing B4C powder inside, compressed to a theoretical density of 70 % by vibro compacting. B4C is preferably used in most of all reactor types because of the high reactivity worth reduction capability and its low price compared to other absorber materials

The mechanical integrity of those control-rods is safely fixed by a sheath metal jacket surrounding the absorber tubes and welded to the upper and lower end fittings of the control rod and a connectin o t d arrange ro e gabsorbee centr th th f t o a ed r cross Design variations have been performed in order to extend the service life limit by using high purity steel, different absorber

1 Summary report of thesis including some additional aspects to the actual problem by using B4C and Hf within a control-rod in general

121 volumef locatee H highe th d t a an dsr bur p zonesu ne lates controE Th . G t d versioro l n called marathon shows a new basic control blade structure consisting of vertically extending square tubes welded togethe lona n i rg line (about 450m foro )t m solid stainless steel blades fille witp du h absorber segments.

In another prio t configurationar r bees ha nt i ,know provido nt esolia d stainless steel member with drilled holes, the ABB basic design (CR 70) [1, 2]. This solid steel member has the same length and width of the prior sheath metal jacket. Every solid blade has a plurality of accurately drilled holes extending horizontall d towardan e yrequiree centro t th th fros d e sidf o emth ro e d e edgeth f o s cruciform shape solie Th .d stainless steel sheets designed wit thicknesha weldee ar abouf m so dm 8 t togethe centree th t a r . Depending from different design variation powdeC B4 eace n i rsth h wins gi filled by vibro compacting in horizontally drilled holes of about 6mm diameter and spaced at a pitch of about 7.5 mm. Regarding to newer design variations the holes are additionally filled with hafnium pinr additionallo s) (th82 win e eR th uppermos gf (C y o fillep to dholes9 e 1 togethetth t a ) r witf hH end plugs (CR85) at the outer section of the blade wing. To balance the gas pressure between different absorber hole e borde sth absorbe e fronth n i rf o t r win equippes gi d wit hcommunicatioa n channel to balance the gas pressure between different exposed absorber sections. The nuclear worth is adjustable by varying the depth of the holes, the absorber cross section and the space between two holes.

At the moment the latest ABB control rod versions CR82m and CR85 m are tested all over the worl exteno dt nucleae dth r data base. This particular contro desigd ro l nsignifican4 showo t p u s t variations [2] by using:

pelletC 4 B - s instea f B

Lilting handle

'—— ————Cooling hole

Uncoupling handle

SSJARE Ties

Standard Hybrid/Duralife XY Marathon ABBCR XY

Different1. FIG. basic design variations ofBWR control rods

122 2. OPERATING EXPERIENCE

Operating experience has proved so far that this reactor component cannot be used for the total reactor life time as originally presumed, but instead has to be considered as a consumable article [3-11]. After only few operating cycles, the mechanism of absorber failure has been shown to be neutron induced boron carbid comparee% swellin16 o t unirradiateo dt p gu 0 powderd84 , density reduction of the absorber material at the same magnitude and stress cracking of the absorber tubes (enclosure) followed by erosion of the absorber material. The occurrence of absorber gaps will start at a local-burn up between a range of 50 to 55% boron 10 depletion in the case by exposing the GE standard contro d desigro l n [12, 13]. Eddy current testinrode th sf go showe d defect signals, i.e. longitudinal cracks, corresponding to a slight diameter increase up to 0.2% (Stainless steel 304).

The mechanism during absorber exposure with neutrons has to be pictured as follows. During irradiatio nsolia d rin formes gi d 10 reactioresule a (FigurB th s a f o t) ne2a with neutrons, starting from the outer surface absorbeth f o e raxiae areth lo groo at t centr p wu absorbee e th lin f o e r tuby b e following an exponential burnup expression. This means the density reduction will start directly at the inner wall of the tube to indicate massive circumferential stress. Those forces are necessarily needed to move small particles of absorber product forcibly to the gaps (free volume) available. The original free volume of 30% (B4C 70% at theoretical density) is not completely available for swelling accommodation, sinc e forceeth s neede movo dt e small particle f absorbeo s r product will pase th s mostly intergranular (see Figures 2b and 2c) and longitudinal in view of the tube axis (see Figure 2d) alloo t wwasa t mechanishou m throug crace hth k (see Figur durin) e3 g exposure ultimate strengtf ho the absorber enclosure which is actually used. From this it may be concluded that the cracks could only aris connection ei n with mechanical steady state innestrese th ro st surface

a) B4C cake [4] c) crack (more older] [3 )

b) Intergranular young crack [31 d) View through the coolant hoie to absorber tutje (segment) showin glongitudinaa l crack '.4j

cake C cracking4 FIG.B and 2. behaviour within absorberthe tube

123 gf "8

FIG 3 Radiography of the upper section of an absorber blade showing absorber loss (marked)

In the case that operation of a control rod showing massive absorber gaps is continued in control cells, this can lead to an increase of the local power density distribution in the core and under certain conditions even ca ,n cause damagfued ro l e Figur se (PCI , g wher e4 e ( ) elosa f abouo s t regardin% 73 originae th conteno gB t o burnut 10 l e d wash-oudu tpan t occurred [15] thermae ,th l accumulated neutron e fluencabsorbeth s m

Figur illustratee5 core sth e loading schemWurgasseP NP f eo n cycl showine4 g locationf o s defect fuel elements (fuel array d defecan ) t fuel rods (marked dots) includin l controgal l positions (cruciforms) being inserted during the fourth cycle

The nuclear behaviour of ABB control rods is basically comparable to those of the GE-design Destroying investigations performed at the first core control rod CR70 design verified that the damage mechanism has to be exactly the same neutron induced intergranular stress cracking of the absorber blade resultin absorben gi r los t higsa h bur zonep nu s Figur showe6 radiograpsa uppermose th f ho t part of two CR70 control rod wings [16] The gaps located at the outer ends of the horizontal fillings indicate, tha cracke tth s must hav elongitudinaa l shape (horizontally cracks everd e )an b yo t hol s eha calculated separately in view of "critical" neutron fluence (defined as beginning of absorber loss)

As part of a research program 8 ABB lead control rods were tested One of 4 CR70 type control

rods showed cracks after 3x18 months cycles (Ocm = 2-23 x snvt) [17] The accumulated operating experience with all B4C control rods (CR70) showed, that more than 90 % of the replaced rods were found to be cracked in the uppermost top section [18] The reason is explained by G Vesterlund vera e (ABByb strono t ) g axial gradien f absorptioo t n Same result bees sha n observe authoe th y db r regarding to GE type control rods

124 FIG 4 Radiograph of the upper quarter of a high burnup control rod (GE standard design) showing massive absorber gaps

A Johnso ] report[2 n s abou o CR8tw t 5 control rods operatin n deepli g y inserted control positions durin l cyclegal s fro m rode s founcrackee th b 198 swa f o o t d o 1995e t d durinOn 0 e gth inspection in 1990 The crack was positioned above hole number 20 from the top, i e the uppermost hole containing boron carbide

A Huttmann [19] reports about 2 "Ultimate" control rods (type CR85) operating in deeply inserted control position neutroa o t p su n fluenc snv3 3 tf (quarteeo r segment) They were also found to be cracked The author came to the result, that CRs of this particular type will not loose their shutdow d scranan m capabilities f defecti e describe, th f o s d natur presene ar e I giv( t e e prooth f o f incorrectnes f thiso s conclusio [15]n ni )

125 1.43 1J7 0.84 0/,., U4 1,48 0,04 j,2 U3 0.73 U5 O 0.72 0.69 0.54 0.01 0 0.03 0

2,25 1,51 2.10 1.69 2.25 1.51 1 |9 1.76 2.23 1.66 2.23 1.66 I'.03 ».S 2.23 1.77 2.23 1.77 0 0.67 0.57 0,0' 0.67 0.57

l->5 1.23 1.35 1J2 0.71 0.57 0.750 - 0 J O o.o r ; 2.09 1.67 2.09 1.67 2.17 2.17 1.74 1.74 1.01 0.79 1.01 0.79 0,09 0.09 0.01 0.01

1.74 1J7 0.82 »."' 0.09 O.IO 0 0

2,09 1.6" 2.17 1.74 1.01 0.79 1109 0.01

1 I i

1,69 1.42 1.26 U6 1 23 1.76 •* >3 1.66 1.46 ]J4 1-35 1.22 0.811 2^23 1.77 0.76 0.75 o."l o.«" 0.01 0.6" 0.<7 0.01 V 00/• 0,03 0

< J

1.2 •> IJ6 IJ6 i_r 0.74 07; 0 69 0,112 (1.01 "03

fluence in snvt

EOC 4 BOC A upper part

EOC 4Endofcycle= 4 BOC4 = Begin of cycle 4

lower part

FIG. 5. Core loading scheme ofNPP Wurgassen cycle 4 showing locationsall of defect fuel elements (yellow) defectand fuel rods (marked red) including controlall positions (blue) being moved (inserted) during fourththe cycle

Introducing those high worth control rods (110% relative worth) gadoliniue ,th m concentration can be reduced significantly up to 3% by using a control rod management to exchange all control rods up to the outermost located core row against CR85 CRs (type 110%) within 3 cycles (ABB patent). This strategy must include tha l controal t l rods hav holo et relative dth e effectivenes t 110a s y %an time and during each cycle! Own investigations for a given example showed, that the number of

126 FIG 6 Radiograph uppermost ofthe part ofCRTO control Kingsrod [16]

simulatio damagf no e e fronth f to absorber blade horizontally crack

- vertically crack

FIG 7 Schematic representation of an ABB control rod segment and mechanism of damage randomly designed digitalby image processing visuala get impressionto (any resemblance to an existing blade section is purely coincidental)

127 control rodr cyclpe s e will o 200raist p %reducinu ey b e gadoliniugth m concentration onla t a y quarter fuel loading. This fact should explain enough! According to the latest reports [20], it may be concluded that control rods, additionally filled with hafnium pins at the top of the wing, will show horizontall wels ya verticalls a l y cracks (Figur. e7)

combininy B g both absorber materials withi e absorbeon n r enclosure o differentw , t damage mechanism will work together e crackTh . s havin a glongitudina l shape will only arisn i e connection with mechanical steady state stress as a function of the accumulated neutron fluence. The vertically shaped cracks must have another cause.

At ABB control rods (CR82, CR85) first cracks were preferably found at the upper part of the control rod e hafnium/boro,th n carbide area r thiFo s. reaso e boroth n n burnup reaction shoule b d investigated with respect to the reaction B4C with Hf during exposure to neutrons. Control rods f materiacontaininH d an l C whic4 g B directl e har y located close pic 2 togethep H t kneeu y no dan o d r mechanism through stainless steel clad. The nuclear reactions building tritium within the control rods containin followss a e ar B :g10

(thermal) (]) n H+ ) (2 Q,23MeV+ H 3 + n->9) Be+(3 3H (fast) (4)

They might act excellent as supplier for the hydrating process to form Hf HI 7 bulges. The kinetic energy (see eq. (1) ) resulting from different reactions during neutron capture probably works as a moderator with respect to the tritium to react with hafnium to form a metallic gas mixture. Hf hydride precipitates in producing Hf hydride bulges. The following formula describes the chemical mechanism

10Hf+ 17H^ 10HfH,7 (5) (6)

n(H)xM(H)——— — = ^x- SQ0095 K9500ppm KF n(Hf H 17) x M (Hf H 17) 1x180.314 Figure 8 represents the H Hf phase diagram which is based on the isothermal equilibrium hydrogen vapour pressure measurements of R.K. Edwards and E. Veleckis [18]. For a different pressure (the pressure adjusting during exposure principle )th e phas eexpectee linb o et markes di n di red. Figure 8 shows that at locations of lower temperatures Hf hydride precipitates due to thermal diffusion. Local volume extension by 14.7 % (Hf H! 7) obviously reacts in removing the metallic border froabsorbee frone th m th f o t r wing (vertically crack).

The two different mechanism of damage might be the reason to the longitudinal and vertically shaped cracks found in the ABB control rods.

Additionally swelling deformation of several absorber holes side by side (ballooning), as it has been observed recently in different BWRs in Japan [20], is probably only possible by using the latest ABB control rod design CR82m/CR85m (Figure 9). The following reasons should be responsibl thio et s event:

• reduction of the absorber diameter in comparison to the CR70/82/85/ control rod design reductiothiy n i swa d nuclea e th an f no r life limit (described later); • distance "b" is reduced (closer hole to hole separation) in comparison with distance "a" to claim the way the cracks should follow; • new steel with higher strain capability (creeping) does not bring any advantage in view of the nuclear life limit (described later).

128 Weight Percent Hydrogen 6 0 5 0 4 0 3 0 00 2 0 01 08 10 900 J_____I

D G.

200-

100-

10 20 30 40 - 50 (60 Hf Atomic Percent Hydrogen

FIG. H-Hf8. phase diagram

HfHw

a

FIG. 9. Schematic representation of the new ABB control rod concept and mechanism of ballooning (which -will lead to "stuck rod")

3. MANAGEMENT FOR AN OPTIMUM CONTROL ROD STRATEGY

e limiteTh d service lif f thio e s important core component requires carefully planned strategies in order to minimize the number of control rods to be exchanged. For this reason the author develope dnon-destructiva e testing metho systematicallo dt y stud operatine yth g behaviou f controro l rods since 1983 and to optimize the control rod management with regard to their life expectancy.

129 The present work demonstrates the measuring results of different BWR control rod design variations evaluated wit hele non-destructiva hth f po e testing method utilizing neutron referre- s o dt s underwatera neutron radiography. This particular testing metho s comparabli d e X-rath o yt e examination s developemethowa authoe d th d an y db r specificall thir yfo s application evaluatioe Th . n of all neutron radiography results including a re-analysis of results from destructive tests of B4C absorber samples, led to two basic findings.

First: An approximation equation to calculate reliably the control rod's working life on control position, indicate s neutroda n fluenc functioa s a e f differenno t construction features like absorber radius, theoretical density and absorber tube parameters until absorber washout starts.

secone Th d focal poin f thio t s worprovido t s ki empirican ea l model applicatio usee b do nt for control rods in the shutdown positions for the total remaining holding period corresponding to the safety-related evaluation criteria required. By using both instruments to calculate the control rod's working life, users can optimize their control rod strategy to save 60% of the control rod inventory durin reactoe gth r period.

Figur illustrate 0 methoe 1 th w dsho generally work achievo st optimizen ea d controd ro l management resultw ne l s e authorAl .base th n do s thesis wor surveillanc a shadede t kar ge o T . f eo the control rods working life, generally 4 different methods are available:

• Monitoring of the tritium activity in the reactor coolant; • Monitorin neutroe th f go n fluenc substitution ei reactivite th r nfo y control; • Optical inspection method; • The submarine neutron radiography.

3.1. Monitoring of the tritium activity in the reactor coolant

Tritium is produced in a remarkable amount during reactor operation. It shows a long half life and is difficult to detect, because of its low radiation energy level. It is often to be found as HTO or steamn i O r 2 liquiyo T d form. Becaus f thieo s great expenditure woul necessare db separato yt e this element from natural water. Storage of 3H for a longer time period in the reactor coolant is unsuccessful, because of the long half-life of 12.3 years [21]. disposee b waste o t th s o det ha wateH 3 e r th treatmen o S additionalld an t waste th o eyt airn I . contrast to German regulations, where the tritium release [22] is fixed to different amounts depending reacton o r typfederad ean l authorit e respectivth f yo etritiue federath A m US releasl e stateth s n ei i , generally fixed to a maximum limit of 10 curies at any one time or to 100 curies in any one calendar year [23].

nucleae Th r reactions building tritium withi controe nth ldescribes a rod e sar equation di o t ) n(1 (4). So monitoring the tritium activity in the reactor coolant is generally a good tool to get information about contro behaviourd ro l e methoTh . d work compariny b s e measuringth g resulttritiue th f o ms transmission rate into the reactor coolant of cycle A and B being existent as an equilibrium concentration. The balance equation referring to a BWR working with leaking control rods (see Figur followine bases i th ) n de11 o g expression:

130 EMPIRICAL MODEL: CR' SHUSN I T DOWN POSITIONS BOUNDERY CONDITIONS CORRELATED MICROSKOP. BURN UP MODEL: CR' CONTRON SO L POSITIONS TO THE CONTROL ROD MANAGEMENT Mechanical crlterions

CONTROL ROD MANAGEMENT TARGET MINIMU COSTF MO S

f permlsrtbltI EVALUATION OF CR'S SHOWING A HIGHER BURN UP LEVEL EXCHANGE

INSPECTION METHODS NEUTRON FLUENCE REACTOR OPERATION MONITORING OF OPTICAL NEUTRON RADIOGRAPHY a THE 3H-ACTIVITY WITHIN INSPECTION REACTOE TH R COOLANT a W

ACCORDINO GT THE SUPPLIER

(j) a lt - NEUTRON FLUENCE MIDDLED OVER A CR-QUARTER SEOMENT IN -fO " « / CM *

- REL: "B-BURNU " P MIDDLED OVE CR-QUARTERA R SEOMEN% TN I M 5f (4 « = PROPORTIONALITY FACTOR IN % REL: -B-BURNUP /10 " H /C/tt '

W A - LOS RELF SO : EFFECTIVENESS MIDDLED OVE CR-QUART6RA R SEOMEN% TN I SB 1

FIG 10 Strategy for an optimized control rod management river

FIG.Schematic11. representation reactor ofthe circulation including leakage of coolant

tritiue Th m releas absorbere th n ei , describe functios d a inventorye th f no , wil: be l

fT=Pr-NT (8)

tritiu= r / mwherea releases inl/s

NT= number of tritium atoms

proportionall/3= 7 y factos 1/ n i r r-j = tritium production rate in s"1.

e tritiuTh m transmission rat finallT eR y wil derivee b l : das Rr = A '^' m + > ' (9) 3 whereas A m fQ_ specific tritium activity per m coolant (as to be measured) T'~"

Ar m $• / n? m ~ reactor coolant loss rate (out of the circulation) reacto= r coolant inventory constan= radioactivf o t e transformation with respeco t tritium

This analytical study of a balance equation, shown in Figure 12, generally only permits to get integran a l answer about leaking contro lcore e rodth n .si Figur show2 1 e s that comparing d cyclan eA tritiue Bth , m transmission rate inte reactooth r coolant wil reducee b l d significan shuffliny b t g (48) high burnup f cyclcontroo d froeA en m le rodcontroth t sa l position shuo t s t down positions during e followinth g cycl ("effecB e t 48U")- This fac s importani t t regardin e "Empiricath o gt l Model" applicable for control rods on a shut down position showing low burnup until end of reactor life (this will be described later). Same effect occurs if (8) high burnup control rods on control positions are going to be replaced at the end of any cycle ("effect 8E").

As a second important fact regarding to Figure 12, it may be concluded that the tritium release will only be a function of the control rod leakage itself. Defect fuel elements may not show

132 0,08 A:8E] n Ci/m)i m ( A" 0,07 §aee s T feffect "8E"] [EOC B:48U] 0,06 —*- KWW A B —KR 0,05 [effect "48 U77) — cycle KWW 0,04

0,03

0,02 time in month- 0,01-

116 cycle KWW-* 3 IT T5T6"

relative number of defect fuel elements in NPW K P tim monthn i e - 6 9 2 7 6 8 4 21 2 19 8 16 4 1214 0

caused by pcrwer caused by chang o t 100%O (B f eo ) defect control —preconditioned fuel rods showing elements to 80 % pcrwer massive rate absorber gaps

j ChangeFIG.12. specific ofthe tritium activity AT(nin) perreactorm coolant shufflingby or unloading of high burnup control rods in NPP KRBA [4] and in NPP Wiirgassen

remarkable tritium release by comparing the time periods from cycle 3 to 10 and cycle 11 to 16 of NPP Wiirgassen: after cycle 10 in principle no fuel defect occurred. The tritium activity measured after cycle 10 had been verified by the help of neutron radiographic investigations only to be a result of leaking control rods (3 of ,,high purity steel" version and one of a prototype version).

3.2. Monitorin neutroe th f go n fluenc substitution ei reactivite th r nfo y control

procese Th s computer accumulate fuee sth l exposure adjacen eaco t t h quarter axial segmenf o t 10 every control rod and converts the exposure to a neutron fluence <$CRy4 respectively to a B burnup . Generall a reactivite yth y wort givea f ho n controdetermines i d e initiaro l th y ldb amound an t mi.Tt t individual type of control rod, his irradiation depletion and last not least the total reactivity inserted

133 into the core. The accumulated neutron fluence OcR/4 basically being proportional to an equivalent 10B burnup a is convertible to a burnup by the help of a proportionality factor a.

(10) whereas

2 2 =far,z,t)dt snvtn i 10= 'ncm / (ID

averag= fa4 ^ e neutron fluence ove controra quarted ro l r segmen 10n i t 21 n/cm2

IO ara 4= relative average B burnup over a control rod quarter segment in %

a= relative 10B burnup in % = _A_ A,^ in (n,a)/ar? ^imu proportionalit= a y factorcn i t _ 10:1n/cm2

So, if the relativBe burnup a has not been indicated as a correct neutron fluence value

paste proportionalite th th ,n i O modifie e controy b yo an desig t d r factos ro ldfo wels ha na ra l IO relative th s a e worth realistiA . c forecas service th f o te life only base quarten do r segment fluences "^CR/4 currently indicate designee th y db possiblee rb wilt no t lwil bu , e mentionel leath o dt d effects described before. Figur show3 1 e e basith s c relation betwee e locanth l fluence/burnup (O/a) with corresponding averaged (over a quarter segment) values OcR/4/acR/4 to be explained by the help of local factor , fm radsf .

f =1 58 for rad -

rad(core|

BasicFIG.13. connection between localthe fluenceburnupand with0 a corresponding

averaged values CP(-R 4 respectively OCR 4

134 3.3. Optical inspection method

e opticaTh l inspection method (see Figur ) performe14 e d durin n outaga g y usinb e n a g underwater camer n a visuaa t permitlge impressioo t s n about metallography (corrosion characteristics), mechanical properties and in case of a prototypical control rod version information about dimensional changes (dimensional measurements) But this method is not suitable to give any answer about absorber behaviour (burnup condition, boron loss) described before

Monitor GE-Design ABB-Design

Camera

absorbertube

manipulator

GE-Design control blade section ABB-Design

hafnium

coolant hole in the cover sheet blade segment wit hAbsorberhole2 s and visible parts of the absorbertubes filled up with B4C and hafnium (longitudina axie f th s o o lt (transvers to the axis of the control controe th l rod) rod)

FIG 14 Optical inspection of control rods in the fuel storage pool

135 3.4. The neutron radiography

neutroe Th n radiographic inspection system show Figurn ni consis5 e1 inspection a f o t n station, which safety retain inspectede b contro e o st th d inspection ro la , n hood with collimato referencd ran e 9 2 rods, film cassettesg Cf-25n 0 250 neutroa , n source (Q(t=0 = 1.2xl0) n/cm ), nuclear tracer films (approximately 5.9 x 9.84 in) and a film development and evaluation device.

JNSPEKT1ON HOOD

NR-FfLSVJ

REFERENCE

Cf-252 NEUTRON SOURCE

FIG. 15. Neutron radiographic inspection system

3.4.1. Measuring principle

controe Th l blade segment energyexposee w sar lo a ,o d t paralle l neutro nfuee fluth l n xstoragi e poole Th . measuring principle is based on the absorption of these neutrons by the B present in the absorber rods and the irradiation of a nuclear tracer film by the neutrons which were not absorbed IO . To calibrate the measurement IO segmented reference rods filled with B4C of different theoretical densities simulating different B bum-up as well as water and air. The reference rods are exposed together with every radiograph of a particular control blade section.

Those neutrons not being absorbed will react with the 10B converter located behind the film foil to indicate invisible small nuclear traces in the foil as a result of recoil protons and a particles (see Figure 16).

136 fikn foil converter J__/ |it secondar. y emnrission neutrons neutron conversion He-particlei L B(n) —• * , s | •': ','.'.

FIG 16. Illustration of the film/foil arrangement showing nuclear traces to form a picture

The invisible traces have to be etched in an alkine solution (10% NaOH) at about 60°C for at least 1.5 to 3 hours (see Figure 17).

CELLULOSEN TRAT

FIG. 17 Film developing mechanism and part of a track-etch foil after exposure

The weak contrast is amplified by means of diffused light to convert the trace density to different points of brightnes follows sa s (Fig. 18): higa h trace density result brighn si t field illumination; a weak trace density will form a weak illuminated picture.

black background track etch foil

observation point

FIG. 18. Principle to convert the trace density on the track-etch foil to an image

137 4.2.3 Method evaluateto "loss relativein control worthrod particular a - WA" for control quarterrod segment

evaluato T e los relativn si e contro wortd quartea ro lr hfo r segment firsneutroe th t n radiography resultf so 16 track etch films have to be digitized (see Figure 19) and converted to the actually burnup by the help of an image analyzing system. Usin higga h level CCD-camera reaa , l time digital image processing system videe th , o input signals (image) are digitized and converted (A/D-converter) into 1024 x 1024 x 8 bit pixel values with a range of 256 grey levels or RGB false colours.

control blade segment uppee ofth r part

track etch foil

FIG 19 Digitisation of track-etch foils

realizo T automatin ea c analysi contenB storerelatio6 10 n i s1 e f th o dt o nimaget eacn si h image firse th t centr eabsorbee lineth f so r tubes detectede havb o et . After drawin (seF frod o et line o man E gtw sD froo t mC Figure 20) marker flags are automatically generated from a line luminance profile. The straight line connecting "valleys"o tw e th - each automatically created fro e sammth e absorber tube t frobu ,m different axial positions- represent unknowe sth n centre lin. eB froo t mA

After digitizing the boron content of a control rod quarter segment this part has to be divided virtually in 92 horizontally cross shaped slices of 1 cm length (see Figure 21) or 92 B4C arrangements (equivalent to a control rod quarter segment length). Those arrangements are used as an input for k-infinity calculations (KMC). geometre Th y model includes heterogeneous core geometric, fuel rods, absorbe dimensionsd ro r , boron fillings and fuel channel particulaa f o s r core cell differeno S . t radial boron fillings detecte neutroy db n radiography inspection will be taken into account.

reduco T time expensivr eth efo e k-infinity calculation correlatiosa bees nha n develope calculato dt e loss of relative control rod worth as a function of different empty absorber tubes. This linear function allows to calculate loss of relative control rod worth "WAKMc" f°r every of the 92 "axially homogenized" control rod slices in a simplified way. However, to evaluate a realistic loss of relative control rod worth WA in a particular core each of the 92 WA^c results have to be modified by using correcting factors.

138 255 255 A SS vaneM H v II ? | grey value \ grey value

A B lin> = e C line =>

255 i valley" G tracking method: "vailev" grey value

marker flag F lin> e=

the lowest grey value correspond locae th lo st position regardin axiae th lo g t centr absorbee th f eo r

PrincipleFIG.20. of auto-tracing

139 NEUTRON RADIOGRAPmC MEASUREMENTS STUCK ROD MEASUREMENTS BCD BWR-CORE: BOC

COLD CRITICAL CONDITION

| RESULTS

CONTRO SLICED LRO S WA(Z) |A z = 1 cm \ AXIAL DISTRIBUTIO INTEGRAE TH F NO L CONTROL ROD WORTH SHOWING THE HIGHEST REACTIVITY WORTH

A 2 I IBS- WAin%

88

68 DIFFERENT B

8 15 8 18 SS

K-INFINITY CALCULATIONS

MONTC KM E CARLO METHOD

CORRECTING FUNCTION

CORRELATION RESULTING

FROM K-INFINITY CALCULATIONS

APPROXIMATED LOSS OF RELATIVE CONTRO WORTD LRO H (KONSERVATIVE) WA

FIG.Method21. evaluateto realistica loss of relative control worthrod Waappr(). ipprox

140 Those correcting factors wil ease b l y derivable from reactivity worth analyses (cold worth condition) generally cycl y performebegie an finth o controe t f t no dth a wit d R e higheseacn ro dli hth hBW t reactivity particulaworte th n hi r core. axiaBasee th ln do distributio integrae th f no l (total) contro wortd ro l h here defined versu% corn e i sth A e heighW s a correctine tth g factors wil easile b l y derive differentiatiny db e th g f eaco m hc mentioned control rod worth function along the quarter rod length to provide 92 correcting factors. Each of those correcting factors will calibrate the corresponding (same core position) 92 k -infinity results described in flow char WAs a 2 t approx simulatiny B . g conservative condition realistisa c los relativf so e contro wortd ro l h WAapprox is available.

EMPIRICAE 4TH . L MODEL

The empirical model applicable to used control rods in the shut down positions includes the basic idea of operating control rods of a certain design first in a control position up to a target fluence (Figure 22, horizontal front axis) hi order to move them afterwards in a shut down position (Figure 22, left horizontal axis), where they may work for the total remaining holding period corresponding to the safety-related evaluation criteria required (diagra m, verticall3 y axis) controe .Th l rod stile sar l only require start-ur dfo shut-dowd pan n proceduree ar d san no longer exposed to steady neutron irradiation. The initial model - in this case primarily used for BWR control rods of the type "first core load" - hi principal is applicable to any design as long as sufficient measuring data is available abou behavioue th t correspondine th f o r g contro typfunctioa d s ro lea f theio n r burn-up historye Th . present model, calibrated to neutron radiographic measuring data and gathered by the author during the past 12 operating years, includes a total operating experience of 16 reactor cycles. The results of these measurements are transforme dato dt a representing los relativf so e contro wortd ro l h WA extrapolatee ar approd an wholxe th o det reactor holding period (see Figure 22).

MICROSCOPIE TH . 5 C BURNUP MODEL

In the microscopic burn-up model basic influences of the single design related parameters are described which are important to the nuclear bum-up behaviour of cylindrical IOB absorbers in general. An approximation equation has been developed to calculate reliably, the working life of any control rod design, indicated as neutron fluence crit in snvt until absorber wash out starts - and as a function of different construction features such as absorber radius R, influence of the absorber-tube F(s) (-enclosure) as well as the absorber density pa, in g/cm3. Measurements, destructive investigations of absorber tubes and results from irradiation tests conducted in the Hanford thermal reactor Washington i s n (USA e basicall)ar y use creato dt microscopie eth c burn-up model. Different burn-up distribution absorbeC 4 B f so r pellet wil) % ls 0 demonstrat (pa10 = , maie eth n differencee th f o s shell model depicte [24n di comparison ]i microscopie th o nt c burn-up model develope authore th y botdn b I . h diagrams exemplary 4 different local burn-up profiles a(<&) as well as the corresponding specimen averaged l°B burnup a^O) are depicted as a function of the applied neutron fluence (see Figures 23 and 24).

relationshie Th p between irradiation exposur averaged ean bum-uB d10 p a^® beed ha )n establishen di [24 masy ]b s spectroscopy analysi 10e B/th f nBso rati representd oan basie sth c expressioa whicaj = s hi ) n(O already describe equation di n (10). However re-analysia , thosf so e investigations showed tharelationshie th t p between specimen averaged bum-up am (4>locad )an l bumup a(), calculate shele hele th th f l p o y mode db s ha l to be modified. This modification has been successfully verified by using the theory of functions.

By comparin e locagth l burnup profile botn i s h diagrams each correspondin e samth eo t g specimen averaged burn-up (same colour obviouss it ) , thae locath t l profiles show divergence preferably withie nth outer absorber zones, where the burnup shows a steep angle of gradient. Therefor it may be assumed, that mass spectroscopic analysis conducte thost da e locations generally will contai nlarga e margi errorf no .

Baselocae th ln d o burn-u p equation

wher absorberradiu= er applicate= ) (] d sdan neutron fluence follows specimee th , , tha . n3 t averaged burn-u functioa s pi f no 03) maximue th locae = wherth lf m o burnu. 3 e p profile max

141 to LOSS OF REL.WORTH FOR A QUARTER SEGMENT W A(CR$) IN % mam HI * s " £ a 3 = s SS ? s §•• § §

. S > n o 2 a- 3 § ? s s ? » Q n> 2 o S S 3 a- cat*1' » a t a SI - II w&* -at>

2 s'

***3• ox TARGET FLUENZ REFERRING TO A CONTROL POSITION AFTERWARDS: MOVING TO A SHUT-DOWN POSITION TO WORK THERE FOR THE TOTAL REMAINING HOLDING PERIOD

RESULT: OPTIMIZED WORKING LIFE AR resultig fro boudare mth y conditio criticae th f no l fractional radius change ___ts absorbee th f o i r tube R indicated as to be time proportional to absorber tube cracking follows, that

a = (14)

T?

That means, 3, will only be dependent as a function of the material properties of the particular mknt absorber tube and not as a function of the tube radius itself.

absorber center [4,5 * 10-20) |om=3,3%i lime

FIG.Local23. a(r)averageand burnup a,,, distribution 0.25in inch cylinders 100%of TD as functiona applied ofthe neutron fluence based measurementson calculationsand

shellbythe model [24];in logically false relation betweenm a a(r)and

absorber center line i

borde: __ QiiniiiitiiuniLiciMnaictffiiHamf- r 0,5 1 0

FIG.Local24. a(r)averageand burn-up distributionm a 0.25in inch cylinders of 100%BjC TD as a function of the applied neutron fluence based on the microscopic burnup model in [15]; logically corrected function of a,,, -f(a(r))

143 shoFigure7 2 - w 5 exemplars2 ,,criticale yth ' local burn-up distribution a(Oe wels critth a s )) a l (Fig25 . critical density distribution (Fig. 26) and in addition the distribution of the critical free volume (Fig. 27) as a function of the absorber radius referring to the standard control rod design basically calculated by the help of the microscopic burnup theory.

t center of absorber axis a in x 108 -«- a at (j)(krit) in %

88 —— a at

68

48

28

8 8, 6 8, 4 8, 2 8, 0 6 1, 4 1, 2 1, borde f absorbero r n n n i r

CriticalFIG.25. local burn-up distribution averaged a(0and % mlin ) relative burnupa aas m (0% min j function of the absorber radius r in mm (standard control rod design) calculated by the help of the microscopic burn-up theory

center 1,4 1,6 r in mm—

as FIG. 26. Critical local density distribution p(&crit) a function of the absorber radius r (standard control rod design) calculated helpthe microscopic by of the burn-up theory

define beginnins da washouf go t mechanism

144 After passin e "criticalgth " fluenc a solie d rin f absorbeo g r materia s formei l a soli o t d ceramic cake(100% burnup zone). At this area the original free volume of 30% existing at the beginning of the exposure even completels i t illustrationr y zerFo o use t . op % du , Figur show8 e2 crose sth s section (polished sectionn a f )o exposed original B4C absorber tube just after passing the critical burnup margin of about 50% as measured by the helsecondara f po masn yio s spectrometer (SIMS GKSe th n )Si Forschungszentrum Geesthacht.

center

3BT '/n .i _ u „ 25 F

r n mm

Critical27. FIG. local free volume distribution Vp functiona mas , absorber ofthe radius (standardr control rod design) calculated by the help of the microscopic burn-up theory

Absorber28. FIG. tube Nr.3, section from locateduppermm the 340 end; burnup am(3>cnl) &50% (averaged over the cross section) [15]

145 markee Th d line illustrate , whersr' fre lefe smoothin s y i raiseo th b et t % p volum eu t shape a th 0 o gf et e o exponentian a f o l function (se direction ei ) Figcentreline 27 th . absorbeo ne t th f eo r affected. Evaluatiol al f no neutron radiographic measurements result re-analysia d san s from destructiv sampleeC 4 testB f so s finally leao dt a master equation describing the nuclear working life of any particular BWR control rod as a neutron fluence margin 4>cnt until absorber washout starts:

' ' I krt(CRquartersegment It> h? (s )\ / T )

= absorberadiue whereath R f so rs tube

n = theoretical density (B4C) i th F(s) = factor representing the influence of absorber tube properties

„ R-R R+—— 2 where axia= l burnu m f p factor (available data accumulated with process computer (TCREX)) frad ~ radial burnup factor (available by the help of KMC Monte Carlo analysis) m m n i absorbe= rR radius

m m n i oute= r radiua absorbee R th f so r tube m m n i inne= r radiu: absorbee R th f so r tube s\ =0.635 wall thickness referring to the standard absorber tube in mm

% n i C 4 B = theoretica h l p l density B =101.28 constant in % % in A =-0.5902 constant R =3.175 absorber radius of the test specimen as reported in [24] in mm

0 C =17.152 constant in%/snvt C\ = - 0,1 constant in 1 / snvt

Equation (15) involves different construction features which are responsible to the nuclear working life in general. This will be shown in Figures 29 -31.

5.1. Influence of the absorber radius

nucleaThe r working particula a lif eof r contro willrod l increas absorberradiuthe eas s increasesby following a linear function (see Figure 29).

5.2. Effects with respect to the wall thickness of an absorber tube

t informatioge o T n about influence nucleae th o st r working lif absorbef eo r tubes showing different wall thicknesses, neutron radiography results have been compared to calculations based on the elasticity law as well as on the plasticity method, see Table I.

An influence referring to the wall thickness of an absorber tube - verified to be a calculation based on the elasticity method - showed not to be a predominant factor regarding to the nuclear life tune as presumed before (see Figure 30).

146 applicablo e1 {JKR krtt)CR/ n snui 4 t wall thickness 0,63m 5m 3,5 wall -thickness 0,8n 4mn 3 [T212/2] (R 2 kr,JSE/4=2,52 snvt

2,5 \. j. wall Ihlcknesm m 5 s0, i 2 (}(R 2 ^ )CR/4=2,52 snvt wall ihfckness=Q,63m 5m 1,5 (jt(R, )CR/4=1,6smrt 1

8,5 R in mm i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i i - B 1,75 2 2,25 2,5 2,75 3 3,25 3,5 3,75 4

FIG. 29. Critical fluences

TABLE I. FACTOR OF INFLUENCE REFERRING TO THE WALL THICKNESS OF AN ABSORBER TUBE F(S) RESULTING FROM MEASUREMENTS (NEUTRON RADIOGRAPHY) COMPARE CALCULATIONO DT FOLLOWINY SB ELASTICITY-PLASTICITE GTH Y METHOD

% F(s n i ) neutron radiographic calculations based calculations based on the plasticity method measurements elasticite th n o y method Variant A VariantB T212/1 =case 1 -5.6 -3.3 -17 -20 T2 12/2 = case 2 + 6.8 + 5.7 + 21 + 28

5.3. The steel selection

The steel selection of an absorber enclosure showing a large creeping capability used to obtain a large volume for swelling accommodation of the absorber material will not lead automatically to a longer working life, beed sincha nt e i demonstrated, tha radiae tth l creeping velocit absorbee th f yo r tube material defines da As / At shows small values in comparison to the absorber growing capability (velocity) defined as Ar / At. The swelling behaviour of the absorber described in principal by the help of the applied neutron fluence A — J(J>(Jt is basically controlle forcee th y dsb actinabsorbee ' th o gt r tube after exposure.

5.4. Variatio theoreticae th f no l density

The capability to use the original free volume within an absorber for swelling accommodation will basically grow up by reducing the theoretical density from 70% to a lower margin (Figure 31). The Figure shows tha usiny b t greducea d densit better yfo r swelling accommodation this optimizehave b o et relation di e th o nt numbe atomB 10 f sro require helo d t relative dth e reactivity worth durin whole gth e operation period.

The working life determined by using the master equation is in line with the operating experience an compares di valuee th o dt s currently indicate designee th y db r whic highefactoe e har th y abouf ro b t 2. The reason may be explained by comparing the local burnup profiles a(r) in Figures 23 and 24 showing divergences within the outer absorber zones.

147 literature Th e doe t offecomparably sno an r e studies concerning these problems. Accordin presene th go t t

research, it has to be concluded that the control rod design of all reactors using B4C as neutron poison have revisee b o t orden di predetermino rt reae eth l holding periods.

15 calculate usiny db g elasticity model X PCsn i ) IB [T212/2 as measured by using radiography]

IBWR/Hvbri (s-0,651 d 5 rnm]

[T212/TTs=0.5l

0,65 0,75 0,85 0,95 cladding wall thicknesn n n i ss

s rnaosurBo y usinb d g radiograph^

-18

FactorFIG.30. of influence referring wallthe to thickness absorber ofan tube F(s) (same steel selection) as functiona wall ofthe thickness normalizedmm in wall s the thicknessto standard ofthe absorber tubes = 0.635mm

insnvt

. insnv„, t

7 -

insnvt

insnvt

1 -

I I II I I I I I I I I I I I I I I II II I I I I I I I ( H I I I I I I |+| I I I I I I I II I - Pth 40 50 60 70 80 90 100

FIG. 31. (r,TD) cm ^n snvt as a function of the theoretical density p,h in % until absorber wash out starts depicted for absorber radius- Rt = 7.75 mm; R2 = 2.5 mm; R3 = 3 mm - normalized to the wall thickness of the standard absorber tube0,635= s mm

148 Coming bac Fig o kbot, t .10 h model EMPIRICAe th s- L MODEL applie usedo t d controle rodth n so shut-down positions and the approximation equation to calculate the control rod's working life on control position MICROSCOPIbasee th n do C BURNUP MODEL wil loptimize n presea w tho d controd lro management will be handled (see Figure 32).

differene Th t coloured square boxes show Figurn ni wil2 e3 l simulate control cellmid-sizea f so R dBW core. All yellow square boxes are shown as typical shut down positions whereas the other coloured square boxes will simulate the control positions. Based on the calculated nuclear working life (equation (16)) of any different contro versiod lro n being inserted "lone ,th g life" control rods wil positionee b l d firstl controa n yo l positiop nu to a maximum allowable neutron fluence margin (below the ..critical value" until absorber wash-out starts) to move them afterwards in a shut down position where they may work for the total remaining holding period. At firse th tf coro leas % etotae inventorth t60 luseo e t operatin b p dy u yma g period withou safety an t y loss.

shut-dow_ ._ n position

- high worth control positions

- -contro l positions showing control cell less burn-up increments

FIG. 32. Typical control rod loading pattern

149 existinn A g EMPIRICAL MODE alreads i L - y controavailablR BW r ltype rodefo th ef so "firs t core principan i load t bu "- l suc modeha applicabls i l designy an o et . Some measuring data (radiographic inspection) abou behavioue tth correspondine th f ro g contro typd lro e mus available b t functioa s ea theif no r burn-up history and working life in a shut down position. After modelling no further measurements of this particular control rod type are necessary in any reactor.

REFERENCES

] [I VESTERLUND , HALLSTADIUSG. , , HOFFMANNL. , , CORSETTIH. , , DevelopmenL. , f o t ABB control rods and operational experience, Keratechnik, Vol. 57 No. 2, (April 1992) 102- 106. [2] JONSSON, A., "Control rod performance at Oskarshamn" in Kerntechnik, Vol. 57 No. 2, (April 1992)113-115. ] [3 EICKELPASCH , MULLAUERN. , , SEEPOLTJ. , , SPALTHOFFR. , , "ErfahrungeW. , t nmi SWR-Steuerelementen", Tagungsbericht, Jahrestagung Kerntechnik, 1980. [4] EICKELPASCH, N., MULLAUER, J., SEEPOLT, R., SPALTHOFF, W., "Operational Experience with and Postirradiation Examinations on Boiling Water Reactor Control Rods", Nuclear Technology, Vol , (Marc60 . h 1983) 362-366. [5] SCHULZ, W., "In-Pool Neutron Radiography of BWR-Control Rods within the Scope of a Routine Refuelling Operation", Neutron Radiography , Secone Procth f o . d World Conference, Paris (1986) 359-366. [6] SCHULZ, W., "Neutron Radiography of BWR Control Rods during Outages", ANS Topical Meeting on LWR Fuel, Performances, Williamsburg, Virginia (1988) 364-369. [7] SCHULZ, W., "Betriebserfahrungen mit SWR Steuerelementen", Jahrestagung Kerntechnik (1990). [8] HOFFMANN, H., BARO, G., SCHULZ, W., RYTER, H., "Neutronenradiographie an SWR Steuerelementen", Jahrestagung Kerntechnik (1990). [9] SCHULZ, W., "In-Pool Neutron Radiography of BWR Control Rods", Eighth Int. Conf. on Nucleae th n i E r IndustryND , 17-20 November 1986, Orlando, Florida. [10] SCHUL , "NeutroZW. n Radiograph ControR BW f ly o Rod European si n NPPs" OwnersR BW , ' Group (BWROG), General Meeting Josen Sa , , Dec. 1990. ReaktoB [IIAB ] r GmbH, Nuclear Services "Neutron Radiograph ControR BW f l o yRod n i s European NPPs", Firmenbroschure. [12] GIDARAKOS , SPALTHOFFE. , , "TritiuW. , m Untersuchunge einen na m Hochabgebrannten SWR Steuerelement", GKSS Forschungszentrum Geesthacht GmbH, Jahrestagung Kerntechnik (1983)581-558. [13] BEAVEN, P.A., GIDARAKOS, E., GREIM, M., MULLAUER, J., NIELSEN L, SCHMELZER, F., SPALTHOFF, W., STAHLENBRECHER, L., WILHELM, H., WITTSTOCK, D., "Untersuchungen an einem hochabgebrannten Siedewasserreaktor- Steuerlement (AbschluBbericht)", GKSS 86/V/ t freundlichemi l r Genehmiguns de g Auftraggebers. [14] SE T212-Inspektion, Kopie der Videoaufzeichnungen von 1991, Siemens-KWU BA43 mit freundlicher Genehmigun . SiemensFa r gde . [15] SCHULZ, W., Schadensanalyse von keramischen Bor-Absorberstoffen in Siedewasser- reaktore Ansatzd nun e fiir eine optimale Steuerstabeinsatzstrategie, Dissertation , BerlinTU , , (1998). [16] Kopie neutronenradiographischen nvo Studsvikn i E S n E Untersuchunge2 S 7 , R C B AB n na Blatt 938 und 1029 obere SE-Spitze, mit freundlicher Genehmigung der ABB Atom AB. [17] ADIN , "SummarS. , Dresdef yo Contron3 Experiences"d lRo , Asea Brown Boveri 250/90B M , , (1990) (mit freundlicher Genehmigung der ABB Atom AB). [18] EDWARDS, R.K., VELECKIS , BinarE. , y Alloy Phase-diagrams, Second edition, Volum, e2 Journa Phys.f o l , Chem , (196266 . ) 1657-1661.

150 [19] Ht)TTMANN, A., KETTELER, M., "Operation of High Worth Control Rods in Boiling Water

Reactors", Kerntechnik 57, No. 2, (Apr. 1992) 116-117. AXTOT7^/fTTT [20] Press Release Information Nuclear Power Safety Administration Division, ANRE/MITI,

[21] SCHRUFER, E., "Strahlung und Strahlungsmefitechnik in Kernkraftwerken", Elitera-Verlag, Berlin 33 (1974)'. ^ A 1504[22KT ] , Sicherheitstechnische Rege s KTAde l , Fassung 6/78, Messung flussiger radioaktiverStoffezurUberwachungderradioaktivenAbleitungen. [23] Office of the Federal Register National Archives and Records Administration: Code ot tederal regulations, Par (Revise, Januar1 10 t f o s da y 1991) §30.55. [24] PITNE , RUSSCHERL RA , G.E, "IRRADIATIO BOROF NO N CARBIDE PELLETD SAN POWDERS IN HANFORD THERMAL REACTORS", Wadco Corporation, Richland, Washington, Dec. 1970, UC-25, Metals, Ceramic Materialsd san .

PAGE(ST EX ) left BLANK

151 FABRICATION AND METALLURGICAL PROPERTIES OF HAFNIUM ALLOYS FOR CONTROL RODS

J.L. BECHADE Servic Recherchee ed s Metallurgiques Appliquees, SRMA

P. PARMENTIER CE2M/LETRAM

CEA-CEREM, CEA-Saclay, Gif-sur-Yvette, France

Abstract

higs Duit ho et thermal neutron absorption cross sectio excellend nan t steam corrosion resistance, hafniu bees mha n nucleae useth n i d r reactor contros sa l rods (solid hafnium rods) sinc late eth e 1950's. However hige th , hmateriae costh f o t l and als e higoth hsoli n i densitshapd f dro H ef yo hav es wid it rule t e dou applicatio commercian ni le reactorsth t a t Bu . presen t s beginnini time a potentiaf e H ,b o gt ls neutroa candidat e us nr absorbinfo e g materia n controi l le th rod f o s pressurised water reactors of the next generation. An important research programme has been carried out at the CEA in order to propose a new type of control rod which would replace the steel cladded Ag-In-Cd, currently used by HfB pellets in a

small diameter thin wall Hf cladding tube. The purpose of this paper is to present the metallurgical studies performe2 d on Hf laboratorier alloyou n si obtaio st tubef nH s with optimised properties firse Th .t part relatemetallurgicae th o st l (grain size, chemical composition, texture, etc. mechanicad )an l (tensio creepd nan ) tests, which were performe differeno tw n do t tubes elaborated by Sumitomo Metal Industry (Ltd, Tokyo Japan) and by CEA (CE2M/LETRAM). The wide difference in the mechanical behaviou d metallurgicaan r l parameter f theso s e tube o acquirt s u hav a ebetted le e r knowledge th f o e metallurgical and functional properties of Hf. In particular, the contents in oxygen and other alloying elements, the grain size andeformatioe dth n level parametery appeareke e b o d t avoio st d cracking durin route gth e proces especialld san colde yth - rolling phases e seconTh . d par f thio t s paper deals wit e resulte ongointh hth f o s g programme pursuin e followingth g objectives to produce Hf alloys plates with different compositions in order to have a better transformability of Hf, to improve the proces e associatef rollinth o s d an g d heat treatment n ordei s o obtait r n thin plates with small o n grain d an s heterogeneities finalld an , optimiso yt materiae eth l (composition, coating, etc. orden )i obtaio rt goona d wear resistancf eo Hf to restrict attrition in operating conditions in reactor.

1. INTRODUCTION

An important research sincA programm e earlCE th e e y th 1990' s t carriei ea r t fo s ou d Pressurised Water Reactors (PWRs nexe th tf generation)o orden i , proposo rt typw controf ene o ea l rod to replace the steel cladded Ag-In-Cd (AIC), mainly to improve the neutron efficiency and also to increas longevite eth safetd yan f controy o f HfE$ o e 2l us rodspellet e Th .smal n si l diameter thin wall Hf claddin solutioe th s gi n proposed.

Hf is used because of its high thermal neutron absorption cross section and excellent water corrosion resistance. Solid hafnium rods have been used in US naval nuclear reactors (submarine 571 Nautilus) since the late 1950's and in American PWR commercial power reactors (Shippingport-1 Indian Point Unit-1 and Yankee Rowe) in the early 1960's as control rods [1]. Hf has also been used as control rod Russian si n reactor hige sth h[2]t cos.Bu f thio t s materia s bee ha obstaclln na s it o et wide application. At the present time, Hf seems to be a good solution for control rods as its availabilit s improveyha d wit increasee hth d productio , especiallZr f no y whe e drawbacnth s it f ko high density is balanced out by the use of small diameter thin wall tubes.

The purpose of this work made at CEREM/SRMA in collaboration with CEREM/CE2M, both from CEA studo t fabricatioe s ,i yth metallurgicad nan l propertie f alloyH contror f so sfo l rods. First, we have performe a comparisod n betwee o differentw n t tubes (CEREM/CE2 d industriaMan l processes) whic mako t hs u bibliographia ehav d le e o improvt c f stud H r knowledg f ou eyo n eo metallurgica functionald lan propertie Secondly. Hf f so have ,w e trie optimiso dt alloyf eH s especially to improve their fabrication process and their wear resistance.

153 2. COMPARISON BETWEEN TWO Hf TUBES

2.1. Chemical composition

e firsTh t tub s beeha e n elaborate CEREM/CE2y db M fro mCEZUa S ingot (referres a o t d LETRAM tube) The second one has been elaborated industrially by a tube manufacturer (henceforth referred to as INDUSTRIAL tube) from a WAH-CHANG ingot Both tubes have nearly the same external diameter (9.8 mm) and thickness (1 mm), but the chemical compositions are quite different. The INDUSTRIAL tube has a higher Zr and Fe content and a lower O content (~ 3 or 4 times for each element comparison i ) e LETRAth o nt M tube (Tabls wil seee A b l. nI) e later, these composition differences hav greaea t influenc metallurgicae th n eo l properties.

TABL E. CHEMICAI L COMPOSITIO LETRAE TH F NTUBf MO H E

Tube Ingot Position Zr Fe O H N % ppm ppm ppm ppm 1 1.06 31 310 11 14 LETRAM CEZUS 2 0.87 26 285 12.5 15

•1 J 097 <20 300 <3 <10

2.2. Metallographic analysis

The characterisation of the microstructures by optical metallography has shown a wide difference in grain size between the tubes (Fig. 1). Although both tubes are in a fully annealed condition with equiaxed grains, the LETRAM tube (grain size from 40 to 50 urn) exhibits larger grains than the INDUSTRIAL tube (gram size from 4 to 8 um). Similarly to Zr alloys [3], this fact coul explainee db fabricatioe th y db n compositionroute alsth d y eoan b .

INDUSTRIAL LETRA 200x ( M)

5m 0ji I—I

FIG 1 Microstructure of the INDUSTRIAL and LETRAM Hf tubes

154 2.3. Texture analysis

Texture analyses of the tubes have been performed by using X-Ray diffraction. The (10.0), (00.2), (10.1), (10.2) and (11.0) pole figures have been determined with a SIEMENS D500 Texture Goniometer and analysed with the Orientation Distribution Function. As can be seen from Figure 2, the axes (normal to (00.2) pole figure) are mainly orientated along the hoop direction for the LETRAM tube, where their main orientation is at 40° from the radial direction in the hoop-radial e INDUSTRIAth plan r fo e L tube. This strong differenc e s fabricatiolinkei eth o t d n routes a , previously observed in Zr alloys [4].

Radial INDUSTRIAL Tube Direction

LETRAM Tube

Hoop Direction

(00.2) D.A. (00.2) D.A.

INDUSTRIAL LETRAM

FIG. Texture2. analysis INDUSTRIAL ofthe LETRAMand Hf tubes

2.4. Mechanical behaviour

2.4.1. Tensile and burst test

The mechanical properties of the tubes have also been examined. Tensile tests at 20°C along the axial directio e tub bursd 320°d th an ef an o n tC 0 test 2 hav t a s e been performed. Froe mth stress/strain curves obtaine t 20°Cda notices i t i , d thaLETRAe tth M smal a tub s eha l total elongation compare e INDUSTRIAth o dt L tube, this difference disappearin t 320°ga o T C . (Figure4) d an 3 s

155 understand the early rupture of the LETRAM tube we have used Scanning Electron Microscopy (SEMrupture th n )o e surface afte tensioe th r seee nb tesn n t 20°C frota ca s mA . Figur rupture th , e5 e surface of the INDUSTRIAL tube is ductile with dimples, when the rupture surface of the LETRAM tube is nearly brittle with some intergranular cracks. This type of brittle fracture is certainly due to the chemical composition, perhaps to the high oxygen content. This demonstrates that the LETRAM tube can not accept a very high level of deformation and presents early cracking to accommodate the deformation.

20 30 40 50 Strai) n(%

FIG. 3. Tensile test at 20°C 2.4.2. Creep test

At last, we have performed internal pressure creep tests at 320 and 350°C for different stress levels (from 50 to 160 MPa) and compared the behaviour of Hf tubes with that of Zy-4 guide thimbles. We can see from Figure 6 that after 240 hours of deformation the INDUSTRIAL tube has a very low hoop strain, whereas the LETRAM tube has a higher hoop strain than the Zy-4 tube, which could be explained mainly by the microstructure (texture, grain size, etc.) and perhaps by the chemical composition of the LETRAM tube.

3. BIBLIOGRAPHIC STUDY

wide Th e difference thesf o s e tubes (metallurgical parameter mechanicad san l behaviour) have led us to acquire a better knowledge of the metallurgical and functional properties. We have mostly focused our study on the transformation of the product to obtain a thin material with fine grains and no cracks and also on the mechanical wear resistance. In this paper we will give only the main results r studyou f o . Using this bibliographic stud e havbettew yw eno r understandin e differenceth f go s between INDUSTRIAL and LETRAM tubes and we can now propose an optimised Hf alloy taking into accoun parametere th t f transformatioso controllind nan mechanicae gth wead an l r propertief so each Hf alloy elaborated.

3.1. Elaboratio routd nan e process

To avoid cracking during the transformation of the product (tube/plate), it is essential to have: oxygew lo a n content • ; • an addition of other alloying elements which help obtaining high deformation (higheb N r o ) r conten5% levelo t 3 t( s 35% r sucZ s ha ) [5];

156 • a small grain size which is in contradiction with a low oxygen content for Hf ; obtaineSi e b d addition n an alloya ca e y t dF b f sbu n o • a high level of deformation during the hot rolling process which is essential to obtain small grain especially whe oxygee nth n conten lows i t .

These condition parametery ke e th s e seehavb o st mo t egooa d product especially durine gth cold rolling phases. But compariso,in alloysZr nto alloy ,Hf mucsare h more difficul transformto t .

BURST TEST AT 20°C

600 --

- - 0 50

• - 0 40 2 s •INDUSTRIAL Tube

« 300 - - -LETRAMTube b. *J CO 200 -

100 --

0 -- -10 10 20 30 40 Strai) n(%

BURST TES 320°T TA C

400

350 --

300 -

- - 0 25 " a« . 7" 200 -- w •INDUSTRIAL Tube 0) ^ 150 - -LETRAMTube

100 --

50 --

0 0 2 - 55 1 0 1 5 25 30 Strain (%)

FIG. 4. Burst test at 20 and 320°C

157 INDUSTRIAL

LETRAM

FIG. 5. Fracture surface of the INDUSTRIAL and LETRAMHftubes after tensile test at 20°C

0 003 n Industrial 320°C

• Industrial 350°C

0.002 A Letram 320°C

A Letram 350°C w 0.001 oQ. O Zy-4X1-1 RX320°C o I * Zy-4X1-1 RX350°C 0 TT 40 100 160

-0 001 - Stress (WlPa)

FIG. Internal6 pressure 350°C creepand test320 at

158 3.2. Functional properties

3.2.1. Mechan ical properties

Ther vera eis y strong influenc smalof e l addition alloyinof s g elementand Fe) s (sucO, has impuritiemechanicae th n o ) sN , l(suc C propertie s ha f alloysH f so . Cold workin gstrona alss oha g hardenine effecth materialn e o tth f go comparison I . r alloysZ effeco e nt th ,f oxygeo t n contenr fo t exampl increases ei [5]facto5 a r .y o d b 4 r

3.2.2. Corrosion hydratingd an

Hf has a very good behaviour in water corrosion, much better than that of Zr alloys and HfO2 is a very good barrier against hydrogen diffusion. Hydrating can become a real problem if hydrogen penetrates insid materiae eth cracksy b l . Therefore materiae th , l mus crace b t k free.

3.2.3. Irradiation

Except for Russians studies [6], very few results are available about irradiation effects on Hf especially about creep under irradiation. As for Zr alloys, irradiation strengthens the material and decreases the elongation (total and uniform). Irradiation growth which is mainly linked to texture is quite diametelimiten i % d(1 r after 6xl025 n/m2). During irradiation, ther alss ei o appearancw ne f eo products linketransmutatioe th o dt , i.e.Hf formatiof no secona f no d phase ric tantalumn hi .

3.2.4. Wear resistance

f seemH havo st verea behavioud yba wean i r r resistanc comparison ei steelso nt . This result has been shown by tests performed at CEA in the 1980s at 320°C in PWR water.

4. OPTIMISATION OF AN Hf ALLOY

4.1. Provisionin fabricatiod gan n programme

An ingot of very good purity has been provisioned from Wah-Chang. The chemical compositio gives ni Tabln ni . ThereII especialls ei ppm) 3 yvera conten (6 w ylo O . f Metallurgicao t l analyses of the initial product have been performed: grain size hardness and tension test.

TABLE II. COMPOSITION OF THE INITIAL PRODUCT

Ingot Zr FE O H N % ppm ppm ppm ppm Wah-Chang 0.47 51 63 3 5

Starting from this product we have planned a programme of optimisation for Hf plates of differ- t compositioen n (additio principaf no l elements, Tabl wile ew lIII d dea)an l only with Hf-3.5%Zr.

TABLE III. OPTIMIZATION PROGRAMME FOR Hf PLATES

Composition Objective + 3% Zr to improve the fabrication process + 3% Zr (250-500 ppm Fe) to obtain a small grain size and to improve the wear resistance lighteo t improvo t tub e d nth eweai (3%an e T e th + r, resistanc25%) e improvo t fabricatio e (8%b eth N ,+ 50%n )weaproces e th rd resistancsan e

159 4.2. Elaboratio routd nan e fabricatio Hf-3.5%Zf no r plate

The final chemical composition of the alloy elaborated by CEREM/CE2M is given in Table IV. The main difference is the addition of 3% Zr and a slight oxygen pollution (O content increases from ppm0 alsn 14 observed6e )ca oo 3b t fabricatioe Th . n proces: sis

rollint Ho 930°Ct ga plat e • th , e thickness goes; frommm 8 10.1. 5o t • Annealin 930°t ga C ; durinmn 0 g3 • Cold rolling, the plate thickness goes from 1.8 to 1.2 mm (thickness reduction: 33%) • Determination of the optimised heat treatment after cold working, following the evolution of hardness test and grain size as a function of the temperature (Fig. 7). Finally, a good compromise is obtained after 2 hours at 800°C, which gives a 30 (am grain size; • Cold rolling plate Th .e thickness goe 0.8s o (thicknes t fro m 2 2m 1. s reduction: 32%); • Annealin t 800°ga C durin ghour2 vacuun si m whic graim h y. given 2 siz2 sa e (Fig. .8)

TABLE IV. COMPOSITION OF THE Hf ALLOY Hf-3.5%Zr

N H O e F r Z Plat e1 ______%_____ppm____ppm____ppm_____ppm____ Hf-3.S%Zr 3.5_____60_____140_____5______15

At this point of the fabrication process we have obtained a quality product with no cracks and wit ha fin e grai e materia nth s sampl a s sizen accep t A e i refe ca . lw :o A et ra highe t r levef o l deformation have w , e trie colo dt d wor produce kth t further:

• Cold rolling. The plate thickness goes from 0.82 to 0.33 mm (thickness reduction: 72%); • Annealin t 800°ga C durin hourg2 vacuun si m which give |u,6 m1 s a grai n size (Fig. 9) .

produce Th t obtaine t thida s pointcole b dt ,worke no sampl n ca , de B furthe r becaus have ew e reached the limit of the tools used for this fabrication.

4.3. Texture analysis

texture Th e analysis performe e samth en i d condition onee th ss a sgive Section i n n 2.3s ,ha shown that the sample B has a very marked texture with the axes mainly orientated along the normal direction of the plate (Fig. 10). This type of texture is close to the texture observed in the INDUSTRIAL tube. The prismatic planes (11.0) are orientated along the rolling direction which is characteristic of a fully recrystallized product.

4.4. Mechanical properties

4.4.1. Tensile test at 20°C

Tensile tests have been performed on samples A and B at 20°C with the tensile axis along the rolling direction of the samples. We found a high resistance for both samples with an ultimate tensile strengt whica nearlf hMP o h 0 decreasey40 s slightly with cold work (Fig mose . th 11) t t interestinbu , g feature is that the plastic elongation (uniform and total) increases with cold work (at 72% of cold work totae th , l elongation reaches 35%).

4.5. Wear resistance

This study has been performed in collaboration with M. Jandin from the Ecole Superieure des Mines de Paris, Centre des materiawc EVRY. The first part of this study was to carry out a wear test which could help us to rank the different Hf alloys.

160 A 0 19 w H v^(transverse)

180 iIm- - - - |8S|i t H. . rV- O(longitudinal) 170 4 ° Hv k J 1 (transverse1 . ) 1 60

150 '^^^ 140 ^^^* Hardness j^^J ^^t ^ B 120 110 100 700 800 900 1000

Tern perature (°C)

x 100

Annealin 800°t ga Chour2 % s 33 afte W rC

FIG. 7. Determination of the heat treatment conditions

testine Th g condition detailee sar d below:

pin-on-disk apparatus; alloyf disH n ko s (diamete mm)0 r3 ; alumina ball; normal load of 1 to 10 N; constant sliding velocit m/s1 0. ; f yo teslaps0 t00 ,wer0 1 eo t carriep u t dou room temperature; relative humidit (on% e y25 test with 50%).

161 froe se m n Figurca e tha2 1 eOn t onlmechanise yon f attritiomo observeds ni foune W . d also that, for this test, the dispersion is small The attrition coefficient ji, obtained after 10 000 laps and calculated fro dise mth k rotation couple bees ha , n measured This coefficient, whic independens hi f o t

loae th d (Fig increase) .13 sthes i rapidl firslape w d nth fe stconstantan r yfo specifie Th . c wear ratek

3 ~m/N2 ) has also been calculated (k = material loss (m)/normal load (N) x sliding distance (m)) This coefficien xlO" 3 equas i t5. 1o 3t l m2/N whic less hi s tha coefficiene nth t obtaine 7x10steel- r 1 dfo 6 s( " j ml 2/N) This resul contradictiotn i seem e b o st n wit bibliographie hth c differencee resultth t bu , n ca s be attribute face th t o tha dt weae tth r resistance tes vers i t y dependan compositioe th f o t alloye th f nso and the testing conditions.

x100

n ur 0 10 I—1

FIG 8 Microstructure of sample A

x400

25 urn I—I Microstructure9 FIG of sampleB

162 (00.2) (10.0) (11.0)

FIG. 10. Texture analysis of sample B (DL: rolling direction)

400

350 •

1? 30° ' O. 1 725 T £ • - 0 20 5 5

150 --

100 20 30 40 50 60 70 80 Cold Working '0/-

0 6 0 5 0 4 0 3 0 2 70 80 Cold Workin) g(%

FIG.Tensile11. test 20°Cat sampleson andBA

163 Io o o

1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 Number of laps

FIG. 12. Attrition coefficient as a function of number of laps (F = 1 N, v = 0.1 m/s, 25%R. H.)

1,00 T

o it 0,80 o Stee! o 50% relative humidity 060

040 - 33 TO 25% relative humidity 0> 020- 0)

000 1234567 10 Loa) d(N

FIG. 13. Specific wear rate as a function of the load

5. CONCLUSION

First, this metallurgical stud ys shown abouha f H t, frodiffereno mtw t fabrications (LETRAM and INDUSTRIAL tubes), the noticeable influence of the metallurgical parameters (grain size, alloying element contents, etc.) on the fabrication process and on the Hf tubes properties.

bibliographiA c stud confirmes yha d these preliminary experiment s helpe tubeha n so d s du s an to propose an optimised Hf alloy product for the fabrication process and for a specific use in reactor, oxygew lo i.e a . n content with iro essentialln na y zirconium addition (3-5% Zr), hardl t rolledyho , neede improvo dt weae eth r resistance.

164 Secondly firsa , t optimise f plat s beedH eha n elaborate characterisedd dan . This Hf-3.5%Zr plate allo s beeyha n colthe% dn72 rolleannealeo t p du t 800° da hours 2 r plate Cfo Th . e exhibito sn cracks and a fine grain size of 15 urn. This Hf alloy presents also a good mechanical behaviour with a marked cold rolling texture. Finally, pin-on-disk wear testing has been carried out and has shown a good behaviou thif ro s allo comparison yi steelo nt .

perspectivea s A , other alloys elaboratee havb o et characterised dan d and, concernin weae gth r resistance, surface treatments on Hf alloys have also to be developed (oxidation and nitrogen hardenin t last transpositioe A . g,th ) tube th e o ngeometrt y (thicshoul) mm kmade e d5 tubeb 1. f .so

REFERENCES

[1] HERBERT, KELLER, W., et al., Development of hafnium and comparison with other pressurised water reactor control rod materials, Nuclear Technology, Vol. 59, (1982) 476-482. [2] RISOVANIY, V.D., et al., Hafnium in nuclear engineering, Third Int. Conf. on Reactor Material Science, Dimitrovgrad, Russian Federation, 27-30 October 1992. [3] DOUGLASS, D.L., The metallurgy of zirconium, IAEA, Vienna, (1971). ] [4 TENCKHOFF , DeformatioE. , n mechanisms, textur d Zircaloyd anisotropan an e r Z ,n i y ASTM, STP 966, (1988). [5] THOMAS, D.E., HAYES, E.T., The metallurgy of hafnium, Naval Reactors Division U.S. AEC, (1960). ] [6 RISOVANIY, V.D. al.t ,e , Structural damage f metaso l hafnium during reactor irradiation, Int. Conf Radiation o . n Material Science, Alushta, Russian Federation, 22-2 1990y 5Ma .

NEXT PAGE(S) left BLANK

165 IRRADIATION BEHAVIOUR OF BORON CARBIDE IN WWER-1000

A.V. ZAKHAROV, V.D. RISOVANIY, S.R. FRIDMANN Research Institut Atomif eo c Reactors, Dimitrovgrad, Ulyanovsk Region

V.B. PONOMARENKO, A.O. POSLAVSKIY Moscow Pohmenal Plant, Moscow

Russian Federation

Abstract

Rod Cluster Control Assemblies (RCCAs) with rodlets made from vibropacked boron carbide powder are used in WWERs-1000 The e use yar s automati a d c control rods (CRs d safet)an y rods s hav(SRsSR e d bee) an RCC ns ACR examined after operation years7 period year2- 4 d 2- , f san respectivelso examinatioe Th y rodlete th f no s shows absencf eo the corrosion effects, external mechanical influences and cladding wear The 10B burnup was uneven along height and radius rodlete oth f s Thi typicaa s si l characteristi RCCAf co s Considerabls useSR s da e valueburnuB !0 f observee so par a t da length of 1000-1500 mm for CRs and, at 300-400 mm for SRs Not more than 5% of released gas penetrates to the rodlet gas plenu wors producee it parth inefficiens m d ki f mo tan ma de heliuTh t mlocates i bottoe th n di m rodlete parth fre a f o ts e sa gas or inside the boron carbide grains The gas pressure under the cladding was 10-17 MPa Under irradiation the plasticity of the cladding decreases according with fast neutron fluence along the rodlets At a burnup of 20-25%, the powder core forms coagulated sintered substance filling the cladding, but its deformation is not observed yet Appreciable values of rodlet cladding swellin observee gar t burnuda f aboupo t 45-50 bumuB rat10 e f %eRCCAo Th n p i s variedifferene th n si t reactors Zaporozhe th f o s TheSR Nuclear Powe Kalinie r th Plan d nan t Nuclear Power Plant hav averagn ea e yeaburnue on r r pfo afteaccordingl% 5 r d endinan 5 y2 g operatio similae Th n r distinction characteristie sar CarrieR C r t investigationcfo dou s allow to consider the mam criterion of resource opportunities of the RCCAs based on a boron carbide is maximal average on

section bumu f isotoppo Be, which shoul exceet dno d 45-50% Ther methode ear increaso st lifetime eth RCCAf eo s based on boron carbide up10 to 20-25 years

1 INTRODUCTION

Boron carbid s manha e y unique properties a hig, h neutron worth, chemical stability, high melting temperature, low density and low cost. It has caused extensive use of boron carbide in RCCAs f variouo s nuclear reactor types. Boron carbide powder containing natural isotope t presena s i s t applie s absorbea d r materia e RCCAth n i f WWERs-1000lo s s behaviouIt . r during irradiation determines in a decisive degree the capacity for work of the rodlets and the RCCA's lifetime

Research on WWER-1000 RCCAs with various operation times is presently carried out in the nuclear power plants (NPPs) at Novovoronezh, Zaporozhe and Kalinin. Based on the results of this research, it was recommended to increase the operation time from 1 year to 3 years for control rods (CRs) and to 5 years for safety rods (SRs).

o maintaiT e growinth n g requirement n lifetimo s e extensio d reliablan n e operatioe th f o n RCCAs, the Moscow Polimetal Plant (MPP), as the developer and the manufacturer of the rodlets, has accepted that not less than one WWER-1000 RCCA will be rexamined at the State Scientific Centre of the Russian Federation «Research Institute of Atomic Reactors» (SSC RF RIAR) annual research. Some result recommendationd an s e researcth f so h carries producei t dou d with theis ri held pan submitted in this report.

2. BORON CARBIDE RCCA THEID sAN R WORKING CONDITIO WWERs-100NN I 0

WWER-1000 RCCAs are unified and can be used as CRs or as SRs. During their operation a change in operation mode is possible The WWER-1000 has 61 RCCAs, of which 10 are used as CRs anothee dth r 5las sSR

167 WWER-1000 RCCAs (Fig. 1) have 18 rodlets with a length of 4 240 mm. The rodlets are made 06X18H10a f o T stee fillee ar l tube d wala an d , sl an witthicknes mm m hdiametea 6 m 0. 2 f 8. so f ro with vibropacked powde f boroo r n carbide with natural isotopic compositio densita 7 o 1. t f p yno u g/cm claddine ende th f jTh .so g tub sealee ear d wit detailp hti meany sb f argon-aro s c weldinge Th . abovd absorbee lengt th an absorbee f eth m ho m r 0 columr 71 column 3 s ni s plenu, ga ther a ms ei filled with heliu t matmospheria c pressure.

UPPER TiP

XGAS PLENUM

NICET LNE

8,2mm

CLADDING

^ABSORBER B,C

BOTTOM TIP

) b a)

FIG. Design1 WWER-1000of RCCA androdlet(a) (b)

168 The rodlets are moving inside guide tubes, made from 06X18H10T steel with a diameter of fuee th l wala n i assembly d , l an thicknes mm 12.internae m 8 6m Th .0. f o sl diamete e guidth f eo r . Coolane radia th witp mm e rodlet d ga 8 lhth an t 2. (water, s i s mm tube1 )1 flow s i s s througe hth channel wit temperaturhe speea m/sth 2 watee f d do ,th an ra pressur 300°Cs ei MP 5 1 s e.i

The bottom ends of the rodlets of RCCAs working as SR are approximately positioned 80-100 mm above of the upper edge of reactor core. At reactor shut off they fall down under their own weight with a speed of about 1 m/s. For RCCAs operating as CR, only the bottom ends of the rodlets are inserte reactoe th n di r cor depta t e a 300-70f ho . 0mm

. RESEARC3 REGULAN HO R RCCAs AFTER OPERATIO WWERs-100N I 0

WWER-100 s werSR e d 0examine an RCC e nucleas th ACR n i d r power plants (NPPst a ) Novovoronezh, Zaporozh d Kalininan e , afte n operatioa r 7 yearsn 2- 4 year period 2- , an f so d respectively. The assessment of the rodlets exterior condition indicates that, among the possible damaging phenomena ,corrosiono n ther e ear effects, external mechanical effecte wead th an f s o r cladding. These conclusions have been confirmed by eddy current non-destructive testing and metallographic research of cladding. Some diameter increase of rodlets cladding was observed in RCCAs fro Novovoroneze mth afteP hrNP bein gdurinR useC s dyearga 3 durinR sS (0.7%s a g d )an 7 years (2% with cracking of cladding).

burnuB 10 e vers pi Th y uneven amongs rodletse th t moss i t I .t e RCCAtypicath r fo l s uses da SRs considerablA . e burnu isotope observerodlets R th i f C p bottolengtB a o e f e10 t o th s a t d da hm en higs (FigA . hSR 2) t .300-40 r f unevea 1000-150o fo d m an n 0m m distributiom 0 e radiath f no l burnup is noted in the absorber. For example in RCCA rodlets of the Kalinin NPP, the 10B burnup on absorbee th f outee o ) (abouth m r rri corjus s mm 1 tetwa aboue t centr50%s e 3-8th th d wa % n t i ,an ei average radial burnu 25-30s pwa % (Fig . Thu .3) maximu e sth m average yearl burnuyB rat10 f d eo p di not exceed 5%.

B/C absorber core

0.0 0 80 120 0 40 0 1600 Distance from bottom endm m ,

10 FIG. 2. Relative B burnup distribution in the B4C core of WWER-1000 RCCAs after operation as SR (1) and as CR (2)

169 50

40 -

30 - a. 3 c CD 20 - Average 10 - bumup

5 3 0 3. 5 2. 0 2. 5 1. 0 1 5 0. 0 0 -0.5 4.0 Distance from rim of absorber core, mm

FIG. 3 Radial B burnup in B4C rodlet after 6 years operation as SR (Kahmn NPP)

e measurementTh maximue th f o s m radial average burnuB e absorbed10 th f po re corth f eo rodlets used as SR in different NPPs, have shown, that the rate of burnup essentially differs: from 2.5%/year (at Zaporozhe NPP, for 2 years operation) and 5%/year (at Kalinin NPP, during 6 years) up to 11%/year (at Novovoronezh NPP, during 7 years) (Table I).

TABLE I. SOME OPERATION CHARACTERISTICS OF EXAMINED BORON CARBIDE RCCAs

NPP Operation Calendar Effective Maximum radial Maximum increase Mode time of time of averaged 10B of rodlet cladding operation operation burnup of rodlet diameter Years Effective % % days CR 2 491 32.5 0 Novovoronezh CR "i 793 42.6 0 CR 3 819 53.2 0.7 SR 7 2030 77 , destructio2 f no cladding Zaporozhe SR 2 3.5 0 Kalinin SR 6 25-30 0

The absorber material changes with increasing 10B burnup. At a burnup of 20-25%, the absorber material remain powdesa t highea d ran r burnu t wilpi l for mcoagulatea d sintered substance. This swa observed at cuts on the bottom ends of RCCA rodlets of the Kalinin and Novovoronezh NPP after burnua t a morf R po S beind e thaan g R usenC 25%s da .

e heliuTh m produce o neutrot e du dn interaction, partl s releasei y d from grains B4C, filling empty spaces partld an , y remain poren si f grainso s B4C. Result pressurs ga f so e measurements under the cladding have shown that only up to 5% of the released helium moves to the gas plenum on top of the rodlets. The main share of a helium stays in the bottom part of the rodlets under the cladding. The

170 calculated maximum internal pressure in the gas plenum did not exceed 0.3 MPa under normal operation a t a a conditionn MP temperatur6 0. d san 350°f eo C wit outsidn ha e coolant pressur f 15.eo 3 MPa.

Properties of the cladding material (steel 06X18H10T) testifies the appreciable decrease of plasticity at the bottom ends of rodlets. However, plasticity of the cladding has not decreased to critical values minimue Th . m value f uniforso m relative lengthenin t temperaturega s 20° 350°d Can C were 17-20% and 5-6%, respectively. The maximum increase of microhardness in the bottom part of the rodlets was 30% (from 1800 MPa up to 2380 MPa). There is no interaction of the cladding with the boron carbide absorber.

Ther strona s ei g correlation between change claddine absorbee th th f so d gan r propertied san e rodlete lengtth th f o hs with distributio f induceo n d gamma-activity, thermad fasan t l neutrons burnufluenc B absorbine 10 th d n pi ean g core.

4. TESTS OF BORON CARBIDE RODLETS DUMMIES IN THE REACTOR SM

More detailed researc boron ho n carbide absorber material propertie rodlete th carrie s n i sswa d out through test f shorso t rodlets dummies, made with applicatio f nominano l material researce th n si h dumme reactoTh . y boro e SM desigr Th . nshows c ni carbid d dumme Fign an th i n use s n b i e.4 ywa d botpowdea pelles s a ha d tr an mad coly eb d compacting with subsequent sintering densite e th Th . f yo absorbing core was 1.7-1.8 g/cm3. The length of the rodlet dummies is about 100 mm. A demountable irradiation device was used for carrying out the tests (see Fig. 4a), allowing to periodically replace and inspect the dummies. Simultaneously, 24 rodlet dummies can be loaded in the irradiation device.

Irradiatio fasa carries n ti nwa t neutro dou n flu f 3.3x10xo MeV 1 thermaa 4 0, d cm"> ) an E 2( l neutron flux (without the account of depression) of 8.9xl014 cm"2 (E < 0.46 eV). The temperature of the coolant was 290 ± 20°C, the pressure 14 ± 0.5 MPa and the water flow 1.7 ±0.3 m/s. After each f irradiationo 20-3 y 0da , inspectio e dummieth f o n s conditio s carriewa n d out maximuA . B mIO burnup of about 40% was achieved. All dummies have kept their form and integrity. The maximum increas f claddineo g diameter t exceeno d ddi s 0.3%. With increas neutroe th f o en B fluencIO d an e burnup release H e eth , increased r pellets release (FigH Fo mors e . .th e5) wa , e than twice higher than that for powder, because of the higher temperature caused by the irradiation.

Irradiation result swellingo st , crackin destructiod gan pelletC 4 B 5 fragment n f i s2- no o t p u s IO burnuB ma f 10-15% po t m a lineae Th . r swellin pelletC g4 approximatelrats B f eswa o y 0.3% 1 % t a IO 10 powde C burnup4 B B e f ro Th .practicall y doe t changsno burnustructurs B eit o 20-25%f t p o p eu . Above this burnup sintering was observed, as well as in nominal rodlets. In its structure there are no porosite crackth d an sy decreases.

At average radial !0B bumup of 35-40%, there is no essential difference between pellet and powder with densities of 1.7-1.8 g/cm. 3 For a final conclusion about the advantage of various rodlet designs, reactor tests shoul continuee db higheo t d r burnup.

5. ANALYSI RESULTF SO S

In practice r safreliabld fo ,an e e operatio f RCCAsno s necessari t i , o superviso yt tw r o e eon operational parameter, which determin e conditioth ee rodlet th a maximu f o o t ns m degreee Th . researc n nominao h l RCC d boroAan n carbide rodlet dummies after irradiation unequivocally indicates thamaie th t n parameter determinin operatine gth g burnuplifetimB RCCAe 10 th e f th o e.s si Safe and risky limits of operation can be specified depending on the 10B burnup (Fig. 6).

171 V*

GAS PLENUM

POWDE4 RB o CD

PELLETS

40 a) FIG. 4. Design of sectional irradiation device (a), dummy with powderpelletsand of (c) (b) boron carbide

172 cm3/lgB C 20 4

18--

16 •• 14- 12"

10--

8- 6- 4..

2- 0-- 5 FO.I, 10,21__-21cm'2

FIG. Dependence5. releaseofHe thermalat fromB^C neutron fluence 0.465eV;< E (1) pellets; (2) powder

SAFE RISKY OPERATION OPERATION

cO) •(50 1

0 7 0 6 0 5 0 4 0 3 20 80 90 100 10B burnup, %

10 F/G Dependence. 6 . of diameter changes ofrodlet cladding with B4Cfrom burnupB t a operation WWER-1000in

173 At operation of RCCAs in WWERs-1000, different conditions of burnup rates are realized. Therefore, their lifetime canno e determineb t a calenda s a d r service s expedientli life t I . x fi o yt dynamics of burnup change for each RCCA by computational or experimental methods. Criterion of achievement of a limiting condition should be the average radial 10B burnup of about 50%. Above this burnup deformatioe th , claddine destructios th it f no d gan observes ni t lineada r swelling rate f 0.5so - 0.7casn i %e 06X18H10T stee uses i l s claddinda g material lowee Th . r burnup provides safd an e reliable operation of RCCAs.

From research results follows, that ther simple ear e solution r increasinsfo lifetime ge th th f o e boron carbide RCCAs, especially when use s SRsda r example.Fo , change positioth e f th o ef no bottom end f rodletso thao bottoe s th t m e absorbin edgth f eo g colum s placeni distanca t da f eo about 150 mm from the upper margin of the reactor core, will lower the average 10B burnup o ratt p eu 2%/yea sucn i d h moderan RCCe th , safe n Aca wor20-2o t p ku 5 years (Fig. 7b).

There is a Russian patent for rodlets with the gas plenum in the lower part, which provides not only a decrease of the 10B burnup rate, but also reduces the internal gas pressure under the cladding thud an s ) increase(Fig7c . RCCAe sth s lifetim 20-2o t p eu 5 year.

6. SUMMARY

e lifetimTh f boroo e n carbide RCCA sWWERs-1000e useth n di , directlyB 10 depend e th n so burnup, achieve t theida r operation averagn A . e radia burnuB 10 l e pth approximatels i % 50 o t p yu limiting factor for safe operation of such RCCAs. Lifetime possibilities of B4C RCCAs is not complete. Actions reducing the burnup can increase the lifetime of such RCCAs up to 20-25 years.

BOTTOM 'GAS PLENUM

UPPERMARGINE •OF REACTOR CORE

^GUIDETUBE

C)

FIG. 7. B burnup in rodlets (a) and change of their position relatively of reactor core (b) and design (c) to increase safe operation time

174 WWER-1000 CONTRO MATERIALSD LRO : XA0053640 EXPERIENC DEVELOPMEND EAN T

. PONOMARENKOV POSLAVSKIY. ,A . SCHEGLOA , V Moscow Polimenal Plant, Moscow

A. ZALJETNIY Novovoronezh NPP, Novovoronezh

. RISOVANIV Y State Scientific Centre "Research Institute Atomif o c Reactors", Dimitrovgrad, Ulyanovsk Region

Russian Federation

Abstract

e MoscoTh w Pohmetal Plant (MPP senioe th s )i r designe produced an r f nativo r e nuclear reactor control rodf o s various type purposed san s since 1966 year standare (TablTh ) eI d requirement Russiaf so n nuclear reactor control rodd san absorber material e givear s Tablnm I DurineI g 196 d 19982an Moscoe ,th w Pohmetal Plant mastere e productiodth f no various absorber materials meeting the different requirements (Table III) As a rule, the Moscow Pohmetal Plant uses the same construction material fuen i s l assemblies, fuel elements, fuel rods c Theset , e austemtiear d martenstatcan e stainless steel, alloys based on Zr, Al and Ti and heat-resistant and high-temperature alloys based on Ni and Cr The control rods are mad accordancn ei safeto et y guide Russiae th f so n Federatio g OPB-88 e n( , PBJ AS-89O aR meed )an t high standarde Th s Moscow Pohmetal Plant carries out its work together with other commercial enterprises and research institutes Some of them are presented in Table IV In this paper we consider different aspects of experience design, testing and operation of WWER-1000 RCCAs and, also, their development

1. DESIG SERVICD NAN E CONDITION WWER-100F SO 0 RCCA

Since 1971, the Moscow Polimetal Plant has produced RCCAs for WWERs-1000. The main operation characteristics of the WWER-1000 are given in Tables V and VI and Figs 1-5. The main absorber material requirements are-

• the physical effectivity of absorber materials is not less than 80% of B4C physical effectivity (natural contents of IOB is -18.9%), • the rodlet cladding must be produced from a stainless steel able to be in the water of the first reactor circuit during long time (P=160 to 180 atm., T < 350°C); • the work resort must be as much as possible (i > 5...7 years).

A WWER-1000 RCCA contains 18 rodlets A commercial WWER-1000 has 61 RCCAs. Six of them are used as regulation control rods (CRs), the rest are shut-down rods (SRs). RCCA characteristics are given in Table VII There exist four primary modifications of commercial WWER- 1000 RCCAs, i.e. CR-71, CR-85, CR-95 and CR-97, with different absorber and cladding materials (the CR's number corresponds to the first year of operation). Designs of these rodlets are shown in . 7 Figsd an 6 .

175 TABL E. CONTROI L RODS FABRICATE MOSCOY DB W POLIMETAL PLANT

Reactor Control rod WWER-l.WWER-2, • cast hexahedral shel ACf o l A from boron steel WWER-3M, WWER-440, VC-50 • burnable poison rods WWER-1000 • absorbers of RCCA • burnable poison rods RBMK-IOOO(ISOO) • automatic and manual control rods, scram rods • additional absorber rods AM-1, AMB-I, AMB-II, EGP-6 • automatic and manual control rods, scram rods BN-600, BN-350 • automatic control rods, scram rods, shim rods Research reactors e.g. SM-2, MIR, • automatic and manual control rods, scram rods, BOR-60, IRT-2000, IW-2R ,M burnable poison rods Military reactors • automatic control rods, scram rods, shim rods, burnable poison rods

TABLE II. MAIN STANDARD REQUIREMENTS OF CONTROL RODS

1. Large physical efficiency 2. Low decrease of burn-up rate of absorber 3. High resistanc radiatioo et n damage (undesirable change shapea n si , dimensions, a structure and properties of materials during irradiation) 4. Comparability absorber materials with cladding and cladding with coolant 5. Low depth of deterioration 6. Good movemen reactoe th n i t r core 7. Good behaviour at LOCA and other accident conditions 8. primw Lo e cost 9. Technological effectiveness of production 10. Good utilizable waste, absenc long-livef eo d high-radioactive isotopes

TABLE III. ABSORBER MATERIALS FABRICATED AT THE MOSCOW POLIMETAL PLANT DURING 1962 - 1998

Type of rods Type of Materials absorption

Automatic control rods, n, a (naturalC 4 B ) 10 manual control rods, (enricheC 4 B d B) shut-down rods MeB2 !0 6 includinB e M g enricheB d Boron steel (SBA-2, SB-2, SB-2M)

n,y Eu2O3, Gd2O3, Dy2O3, Sm2O3, Er2O3, Ho2O3 Dy2TiO5, Dy2Ti2O7, Eu2TiO5, Dy2O3-HfO2, et al. Eu2O3+Mo, Eu2O3+Al, Eu2O3+(Ni, Cr), Dy2O3+Cu, Dy2O3+Mo. al t e , Hf, Hf+Nb, Hf+Ta

Burnable poison rods n, a B4C+A12O3, CrB2+Al. 2Oal t 3e , B4C+A1, CrB2+Al, B+Zr, et al. n,y Gd2O3+Al2O3, GdAlO3, Gd2O3-ZrO2-Nb2O5, et al. Gd2O3+Al, Gd-Zr-Nb, et al.

176 TABL . CO-OPERATIOEIV N SCHEM MOSCOE TH F EO W POLIMETAL PLANT WITH OTHER COMMERCIAL ENTERPRISES AND RESEARCH INSTITUTES DURING FABRICATION OF CONTROL RODS

Technical task The Chief Designer of a reactor

^ f Project development Moscow Polimetal Plant

1. Selectio absorbef no compositd ran e materials, development MPP of a construction and fabrication method of rods. 2. Physical efficiency research. RSC CI, IPPE, EMBB 3. Compatibility research of absorber and composite materials. MPP, IPPE 4. Research of corrosion resistance. IPPE, PJNM 5. Radiation stability researc absorbef ho r materials, rodlet SSC RIAR model, RCC researcAn i h reactor theid san r post-irradiation examination in hot cells. 6. Special testing (?, vibro-, thermo-, wear stability et. al.). MPP, EMP, RCIET, RDO HP,EMBB, finae 7Th .l calculations, produc technicaf eo l documents, MPP interdepartmental tests

Production of control rods MPP

i r Post-irradiation investigatio usef no d control rods SSC RIAR, RSC CI, IPPE, NV NPP

i r Serial productio controf no l rods MPP

177 TABL . OPERATIOEV N FEATURE WWERS-100F SO 0

Parameters Unit Condition or value Coolant water Coolant pressure MPa 16 Inlet coolant temperature °C 129 Outlet coolant temperature °C • average 320 • maximal 335 Max. temperature of coolant surface °C • outside 345 • inside 375 Max. flux xlOun/cm2s • thermal -0.6 • E>0.1MeV -2.2 Max. RCCA drop time s <4

TABLE VI. AVERAGE GROUP NEUTRON FLUX OF WWERS-1000 REACTON (I R CORE VOLUME)

Group E Ft number MeV n/cm2s 1. 15 2.09xl010 2. 12.2 1. 11x10" 3. 10.0 4.34x10" 4. 8.18 1.75xl012 5. 6.36 4.24x2 1 10 6. 4.96 5.69xl012 7. 4.06 l.lSxlO13 8. 3.01 1.17xl013 9. 2.46 3.34xl013 10. 2.35 1.46xl013 11. 1.83 3.10xl013 12. 1.11 4.27xl013 13. 0.55 6.03xl013 14. 0.111 6.32xl013 15. 3.35x10° 2.54xl013 16. 583x6 10' 2.08xl013 17. lOlxlO'6 LOlxlO13 18. 29x1 0'6 5.49xl012 19. 10.7xlO'6 4.90xl012 20. 3.06xlO'6 4.64xl012 21. 1.12XKT6 2.59xl012 22. 0.0625x1 0'6 6.00xl013 >0.8 ~1.33xl014 >0.55 ~1.54xl014 >O.H ~2.19xl014 thermal ~0.6xl O14

178 F, x1021n/cm2

-3,48

-2,90

-2,32

-1,74

-1,16

-0,58

T L 0,00 0 200 400 600 800 1000 1200 1400 1600 rodlee th t bottom Lrodetm >m

FIG. 1. Distribution of thermal neutron fluence (A) and B burn-up (O) along the length of a rodlet

F, x1021n/cm2

3,5-

3,0- V \ V \ 2,5- , \ ma>x (t=327x2 eff.gays ^ at H=660 mm) 2,0- V / \ < \ \ 1,5- \ \ V \ N 1,0- ^ \ \ v \ V ^ v mm/ \ 0,5- \

^^^ n n- 0 500 1000 1500 Hrodletm m '

FIG. 2. Thermal neutron fluence on the lower part of a rodlet; operation time WWER-IOOOin years;2 is operation regime regulationis

179

-50 0 5 12 0 510 0 5 7 5 H17 15(m0 m CORE *~U/ER R CORE

F/G. 5 TTze density of fast neutron flux (E>0.111 MeV) m the upper and bottom parts of reactor core ofWWER-1000 blockV the reactorand ofNVNPP

1 - gas-collecting

volume

2 - Ni-net

3-B4C

4 — cladding

FIG The6. first generation rodlet WWER-1000of RCCA

181 1-B4C

2Dy- 2 TiOs 3claddin- g rodlet

4-filler (X18H10T)

FIG 7. The second generation rodlet of WWER-1000 RCCA

TABLE VII. MAIN SPECIFICATIONS OF WWER-1000 RCCAs

Parameter Unit Type CR-71 CR-85 CR-95 CR-97 Total length mm 4240 Numbe rodletf ro s 16 a a Absorber material Eu2O3+Al B4C B4C B4C b b Dy203-Ti02 Hf Cladding tube material OX18H10T OX18H10T OX18H10T EP-630V (EP-630V) Tube external diameter mm 8.2 Tube thickness mm 0.6 0.6 0.6 (0 5) 0.5 upper part bottom part

2. RODLETS REACTOR TESTS AND POST-IRRADIATION RESEARCH

At the rodlets designing of WWER-1000 RCCA rodlet, tests were carried out in the SM-2 a WWER-100 reacton i d an r 0 (NPP "Rheinsberg") maie Th .n characteristic e absorbeth f o sd an r cladding materials are given in Table VIII and the main characteristics of the irradiated rodlets in Table-IX. Post-irradiation tests were carried out in the hot cells of the State Scientific Centre of the Russian Federation "Research Institute of Atomic Reactor" (SSC RF RIAR, Dimitrovgrad, RF), the NPP "Rheinsberg" (Rheinsberg, GDR) and the Central Institute of Nuclear Investigation (Dresden, GDR) l designeAl . d rodlet higd sha h stability, their shap dimensiond ean t changeno d sdi . Swellinf go

Dy2O3-TiO2,1 R2O3-TiO2 and B4C pellets were 0.5%, 2 3% and 3.5%, respectively (Table X).

182 TABLE VIII. MAIN SPECIFICATION RODLETF SO S IRRADIATET DA "RHEINSBERGP NP " WWER-2

3 powdeC 4 1B . r extruded together with cladding, p=1.99 g/cm . 3 2. B4C pellets, cold pressing, p=1.75 g/cm . 3 3. Dy2O3-2TiO2 pellets, hoTpressing, p=5.49 g/cm . 3 Absorber . R4 2TiO5 pellets, R=Dy+(5-pressingt ho , g/cm6 Ho p= ) , 8% . 3 5. Eu2O3+Al powder extruded together with cladding, p=2 g/cm . 6. Ni - Hf- Sm - In rods, Ni - 70 %, Hf- 10 %, Sm - 10 %10, I%n - Cladding 1.X20H40, 08.0x0.m 5m 2. 06X18H10T 8.2x0.0 , 6 mm

TABLE IX. MAIN CHARACTERISTICS OF RODLETS IRRADIATED AT NPP "RHEINSBERG" WWER-2 (F=(2.2-9.1)xl021 cm'2, water loo 1= 0p-P MPa , T=270°C)

Characteristic Rodlet

B4C B4C Dy203-2Ti02 R2TiO2 Eu2O3+Al Ni-Hf-Sm- powder pellets pellets pellets powder In rods Ad 0 0 0 0 0 0 T'%n/ cladding

Ad o - 1.35-4-3.50 <0.5 <2.3 — —

absorber burnupB 10 % , 11.8 8.8 - - - -

TABLE X. MAIN CHARACTERISTICS OF RODLETS IRRADIATED AT SM-2 REACTOR (F=1.7xl021 nm'2, water loop-P=19 MPa, T=350°C)

Characteristic Rodlet Extruded type Pellet type

Absorber B4C Eu2O3+Al B4C Dy2O3-2TiO2 Ni-Hf-Sm-In Cladding X20H40. 0 8.0x0.5 mm p, g/cmj (common) 1.93 3.63 1.72 4.85 9.82 , g/cmp 3 (absorber) B-1.53 Eu-1.85 B-1.35 Dy-2.93 Hf-0.9 Sm-0.4 In - 0.4 Ad 0.5 0 1.5 0 0 — -,% (cladding) d He release, cm3 1.8 0 3.0 0 0 Ad 0.5 0 3.0 0 0 — —% (absorber, ) d

183 3. RESORT TESTS OF WWER-LOOO RCCA AT NOVOVORONEZH NPP

designo Tw f rodleto s s were selecte firse th tr WWER-100dfo f Novovoronez0o h NPP, They contained Al+Eu2O3 and B4C powder accordingly. Al+Eu2O3 rodlets have a very long life-time (> 20 years werd )an e replaced becaus followine th f eo g main reasons:

low efficiency (80% from natural B4C); - high cost; high radioactive b2 l:>4d Euan Eu isotopes with long half-life (13.48 8. year d an s years respectively).

Large experience of RCCA operation with natural B4C powder showed some problems. The life-tim limiteds i e burnu,B firs burnuf a ally o t b p , A (Fig. f morpo 8) . e than 30-35%C 4 B e th , powder sinters in a solid column due to high temperature, produces and releases helium and forms lithium (Table XI). At a burnup of 50-55%, the rodlet cladding begins to deform. These conditions are achieved during 2-3 years in regulation regime and 5-8 years in shut-down regime.

Since 1983, test with Dy2O3-TiO2 powder filled RCCAs were started at the Novovoronezh WWER-1000. Afte operatio4 r n year regulation si n regime, there wer problemso en maie Th .n results of the post-irradiation investigations are: the absence of rodlet shape and dimension changes;

no swellin Dyf go 2O3-TiO2; - no sintering of Dy2O3-TiO2; out-gassino n - g during irradiation;

- Dy2O3-TiO2 powder was freely extracted at the cladding cut; - low y-activity of radioactive isotopes with short half-time.

These testd investigationan s s showed that these RCCAs hav a vere y high resistanco t e radiation damages maie Th .n defect thif so s relativeldesige th e nefficiencar w ylo y (85% from natural Bburnud 4Can ) Dy-isotopef po s (Figs. 9-12).

Bu,%

80- rrax=77(100%)

70- 166(86%) 60-

50-

40-

30-

20-

10-

0—r—— -, L m ,m ttercde0 5 t 100 150 200 250 300 350 400 450 rodet bottcm

FIG. 8. B burnup in B^C powder of RCCA after irradiation during 7 years in Novovoronezh NPP unit-5

184 TABL POST-IRRADIATIOL EX N EXAMINATION RESULT CONTROF SO L RODS WITH VIBROPACKE YEAR3 AFTE C D 4 DB AN S R 2 OPERATIO POWEN I R CONTROL MODT EA THE NOVOVORONEZH NPP UNIT-5 (WWER-1000)

Irradiation time, eff. year/day 2/491 3/793 Max. neutron fluence, n/cm2 • thermal l.SxlO21 2.4xl021 • fast (E>0.8 MeV) 3.3xl021 5.3xl021 Max. boron burn-up at the 35 41 -45 normal conditions, % Helium releasej cm , 70- 130 202 - 248 Gas pressure, kg/cm2 • 20°C 5.0-9.3 14.4- 17.7 • 350°C 10.6-19.8 30.5-37.6

Condition of "B4C-cladding" B4C powder was freely emptied B4C was sintered with syste bottothe min m part at the cutting of cladding cladding Cladding diameter change% , 0 0 Oxide film thicknesn so 5-7 5-7 the cladding surface, mm Cladding mechanical good low plasticity properties in the bottom part

SECONE 4TH . D GENERATIO WWER-100F NO 0 RCCA WIT H, cc) COMPOSE(n - D AN - y) , D(n ABSORBER

Since 1995, the Moscow Polimetal Plant began to fabricate a new generation of RCCAs for

WWERs-1000 witpowdeC 4 uppee hB th n rrodleti e r parth Dyf d o t s2an O3-TiO2 powde bottoe th n ri m part (300-80 OX18H10e (Fig) Th . 0mm .7) T claddin s changegwa Cr-Na y db i alloy, EP-630Ue Th . mechanical propertie f EP-630so e showUar Tabln ni e XII. This materia vera s yha l high plasticity after irradiatio higd nan h resistance corrosion cracking (IASCC).

In some rodlets hafnium was used in the bottom part instead of titanate dysprosium. This RCCA desig s fabricate Ukrainiane wa th r fo d n WWER-1000 e MoscoTh . w Polimetal Plan s alsi t o designing hafnium rodlets without cladding.

TABLE XII. TYPICAL MECHANICAL PROPERTIES OF EP-630U

Fast neutron fluence Temperature Yield Tensile Elongation (E>0.1 MeV) xlO22 of testing stress strength cm"2 °C % 0 20 520 800 47±3 350 390 700 42±3 1.12 20 660 820 55±3 350 310 620 49±3 2.7 20 750 870 50±3 350 540 660 41±3 6.8 20 820 910 33±3 350 660 680 28±3

185 Dy 160 161 162 163 164 Initial contents, % 2.34 19.0 25.0 24.9 29.1 Microscopic absorption 130 680 240 220 2780 section, barn

:, at.%

200 400 600 800 1000 1200 the bottom rodlet -rodlet, mm

FIG. 9. Change ofDy isotopic composition along a rodlet after irradiation during 3 years in operation group of Novovoronezh NPP unit-5

5. END-OF-LIFE CRITERIA OF WWER-1000 RCCAS

The following new main criteria for the life-time of WWER-1000 RCCAs have been established: efficience th operatioe th f o y d decreasnen time th e t et shoula mor no e db tha n 10%; - the rodlets should have shape stability and there should be no cracks; increase claddinth e th f eo - g diameter (Ad/d) t morshoulno ee db tha n 0.01 mm/year (common increas less ei s tha mm)1 n0. ; the wear depth should be not more than 0.1 mm.

186 Dy 27.32

0 200 400 600 800 the bottom rodlet Lrodletm m '

FIG. Change10. ofDy isotopic composition along rodleta after irradiation during years4 in operation group of Novovoronezh NPP unit-5

187 ±Dys, % 50 -,

45-

40-

35-

30-

25-

20-

15-

10-

5- 160iDy 0- 0 200 400 600 800 1000 1200 1400 the bottom rodlet Lrod|etm m ,

FIG. 11. Distribution ofDy isotopes burnup along the rodlet length after irradiation during 3 years

188 %

60 -i

50-

40-

30-

20-

10-

0J 0 200 400 600 800 the bottom rodlet rodlet' mm

FIG. Distribution12. ofDy isotopes burnup along rodletthe length after irradiation during years4

6. CONCLUSION

The experience gathered with operation of different types of RCCA in commercial WWERs- 100 t NPPa 0 Russian i s , Ukrain Bulgarid t shoan eno wd di afailure f componento s s durin0 20 g reactor-year units7 1 n .so

The new generation of combined RCCAs containing Dy2O3-TiO2 (Hf) and B4C in EP-630U claddin life-tima s gha morf eo e the year0 n2 accordancn si end-of-lifw e ne wit e hth e criteria.

NEXT PAGE(S) left BLANK

189 FRAMATOME CONTROL ROD ABSORBER MATERIALS: XA0053641 EXPERIENCE AND DEVELOPMENT

M. MONCHANIN FRAMATOME Nuclear Fuel, Lyon

X. THffiAULT EOF Septen, Villeurbanne

France

Abstract

Phenomena limiting the lifetime of the RCCAs in PWRs have already been highlighted m the past Wear of the rodlet claddin claddine th f swellind o g an d glowe e en arisinth p n gi rti g fro structurae mth l change AglnCe th sm d absorber materiae ar l o damagtw e th e mechanisms encountered wear, potentially leading firs claddino t t g perforation secondly further lated ro r mechanical failure, and swelling, causing clad cracking and rod swelling possibly culminating m RCCA jamming in the fuel assembly guide thimble The upgrades applied to the new generation HARMONI™ RCCAs have provided a solution to the RCCA wear problem delayed san d contact between absorbe claddind ran g Better knowledg predictiod ean AglnCf no d swelling in the reactor is necessary, however, as it conditions the lifetime of the RCCAs An extensive analysis programme of this AglnCd absorbe s undertakerwa FRAMATOMy nb conjunctioEn i n with EDF orden ,i determino rt causee out-of-pils eth it f - so m d ean reactor changes (creep and metallurgical changes in the material under irradiation), predict its behaviour and make structural improvements limitin s m-reactoit g r change e outcomTh s f thio e s HARMONIe analysith s wa s desig™G 2 n featurinn a g improvement m the Ag-In-Cd material Turning to the 1300 MWe RCCAs containing boron carbide (B4C), an experiment on the behaviou f thio r s absorbe s bee s inserteha r nwa reacto F conducteC d4 ED B intrn a o m dcladding s pre-perforatee th t a d fabrication stage and subjected to corrosion by the reactor coolant The first lessons on the conditions of boron carbide dissolution have been drawn. The improved knowledge in the behaviour of these absorbers m the reactor and the design enhancements mad allowine ear g predictio lifetima f no e with respec swellino t g whic furthes hi r extende generatiow ne r dfo n HARMONI™ RCCA relative to the first HARMONI generation A development programme on an RCCA design using hafnium absorbe hafniuf alss o ri progresn e oi us me woulTh s d completel RCCAe yth f rulswellint o e d sou lowe e en th p n gi rti

1. INTRODUCTION

yearw Fe s ago, FRAMATOM developes Eha RCCw ne Ada product called HARMONI™ which definitively remedies cladding wea offerd an r reaa s l improvement against swelling desige e th Th . f no HARMONI™ RCCA combines ion-nitriding anti-wear treatmen e claddinth f o t g tubd severaan e l modifications suc s reductioa h f absorbeo n r diameter alon e loweth g r par f eaco t h rodlet, helium backfillinAISI316f o e rodlee us th d f Lgo an t claddin enhancinr gfo g swelling/cracking resistance.

accumulatee Th d in-reactor experience, under various operating conditions, confir adequace mth y of this design with respec weao t swelling/crackind ran g resistance indicatioo .N weaf ndetectes o rwa d ion-nitridee th n i d ro dy oparn an mor n o t e than 1700 HARMONI™ RCC operatioAn i dato nt e with approximatel reactorsF thirdo ED y tw maximu e n i s Th . m wear experience feedbac ion-mtriden a f ko d RCCA is seven 12 month cycles in 900 MWe and 1300 MWe EDF reactors (lead RCCA) and only four cycles regardin e clagth d swelling. Probably e swelling/crackinth , g phenomena will conditioe th n lifetime of the HARMONI™ RCCAs design.

The HARMONI™ RCCA was the first result of a large-scale Research and Development programme undertaken to provide a second generation of HARMONI™ RCCA with higher performance regardin phenomene gth a limiting RCCA lifetime.

Afte shora r t swelling/crackinrecae th f po g phenomena, this paper present extensive sth e analysis programm AglnCe th f eo d absorber behaviour under irradiation undertake conjunction ni : o nt witF hED

understand the creep mechanism; determine the effect of AglnCd grain size on creep law; predict the AglnCd creep deformation rate;

191 understan swelline dth g mechanism under irradiation (metallurgical change materiae th n si l under irradiation); predict the swelling volume rate and diameter increase of AglnCd under irradiation.

consequencee Th thif so s programm HARMONIe th n eo design™G 2 , featurin improvemenn ga t in the AglnCd material and an increase of RCCA lifetime, are detailed. The improved knowledge of boron carbide (B4C) behaviour subjected to water corrosion is also an important safety concern. The special experiment conducted in the PALUEL 2 reactor is described and this paper gives an outline on conditione th borof so n carbide dissolution.

Lastly generaa , developmene l th ide f o a a lea f do t hafnium RCC f s Apresentedo i e us e th ; hafnium would completely RCCAe rult swellins protecteth i eou t e f lowei o th f i sp n gi ti rd against hydridin weard gan .

. GLOBA2 L DESCRIPTIO SWELLING/CRACKINE TH F NO G PHENOMENA

Wear inspections on RCCA absorber rods have revealed swelling on several rodlets, sometimes combined with longitudinal cracking of the cladding at the bottom of the rodlet. The interpretation of this phenomena whic finding e bases h i t cel th rodlete 130d n ho dl o an n investigationi s 0MW 0 s 90 R PW f so presented below, demonstrates:

diametra longitudinad an l l variatio e absorbeth f no r cause creey db p unde effece th rf o t applied stati dynamid can c stresses; diametral swellin f AglnCgo d absorber resulting from irradiation-induced metallurgical changes.

These findings lead to the following conclusions about the cladding strain mechanism:

beginnint a lifef go , axial shortenin absorbee th f go r stack occurs unde effece rth thermaf o t l creep, caused by the static and dynamic stresses applied to the absorber. The direct outcome of this axial shortening is an increase in absorber diameter and gradual reduction of the absorber-cladding gap e thermaTh . l creep mainly govern e narrowinth s f absorbero g - cladding gap and there is progressive closure of this gap. When contact has occurred, creep stresses are not sufficient to cause cladding deformation; same th t ea time, unde effece th r f irradiationo t e chemicath , l compositio f AglnCno s di modified (transmutation of Ag into Cd and In into Sn) causing the beginning of swelling and occurrence of a second phase as soon as the solubility limits of Sn are overstepped. It leads to deformation of the cladding and to a local stress increase in the cladding; cracking occurs alon gcladdina g generatio nabsorbee linth f ei r pushing lead exceeo st e dth cladding ductility which is itself linked to irradiation; after cracking, absorber creep resumes and is superimposed on irradiation swelling.

Figur illustrate1 e absorbee effece th sth f o t r diameter evolution versu srealityn i tim t e ebu th , phenomenon is more complex and this variation is not fully linear since it is influenced by:

the operating mode, which influences RCCA insertion and therefore the heat-up and the fluence seen by the materials; spatiae th l distributio f absorber-claddinno (thp gga e off-centerin absorbee th e f th o g n i r cladding and the progressive closure of the gap contributes to a drop in temperature); changee th absorbee th n si r grai temperaturno t size edu e effect; changee th mechanican i s l properties (suc s loweha r thermal conductivity) arising from those in the chemical composition and its structure.

Better knowledg f AglnCeo d cree swellind pan predictiod gan f creen o swellind pan g ratee ar s however necessary as they condition the lifetime of the RCCAs.

192 d> c Swelling rejection criteria 0) *-• o> o o V) A Standard absorber rod (0 I diameter

HARMONI o0) c •»-n•s (A

Non-HARMONI RCCA lifetime HARMONI improvement Exposure duration

FIG Absorber-clad1 evolutiongap versus time 3. AglnCd CREEP ANALYSIS

To understand AglnCd creep and determine its thermal behaviour and creep, FRAMATOME launche extensivn da e programm thermaf eo l analysi creed san p tests.

3.1. Thermal behaviour of AglnCd absorber

This first phase of the investigation programme was to understand the out of pile behaviour, specially with a constant temperature. When the AglnCd material is subject to high temperature above half melting temperature (Tm/2=260 C) or recrystallization temperature (about 275 C), the grain size increase recrystallizatioy sb n the graie nth n coarsening continues. This evolution depend initiae th n lso grain size, the material temperature, the cold working rate and the time.

FRAMATOME ran a series of tests to measure the grain size trend for AglnCd materials which initially had a grain size included between a "small grain" size ASTM 6-8 (between 20 and 45 microns) and a "large grain" size ASTM 0-2 (between 170 and 355 microns) depending on:

in-reactoe th r temperatur materiale th f eo ; the hold time.

This thermal study showed that there was no significant change of this grain size up to 350 C and for higher temperatures this change is smaller in scale with a large grain material. Grain size measurement RCCn so A rodlets irradiate cycle9 d reactorn an si d8 s have confirmed thao t n ther s ewa major change in the grain size in the out-of-flux zones. As a result, the absorber-cladding gap has the time to be closed by in-reactor creep with a low grain size material before grain coarsening occurs.

3.2. Creep test programme

In orde determino rt e bot thermae hth l cree pbehaviouf o mechanis w irradiatee la th a f o rd man d materia lparameters y versuke e sth , FRAMATOM undertakes Eha nlarge-scala e creep test programme on Ag-In-Cd material differenn si t metallurgical state coverind san gwida e rang initiaf eo l grain sizes4 5 . creep tests were run on standard specimens machined from fabrication bars for tensile creep tests or from specimen t directlscu y from fabrication bar r compressivsfo e creep tests more representativ- in e th f eo service loadings.

tese Th t parameters were representative considereb o t s a range o dth s f actuaf eo eo l in-reactor loadings (several temperatures and stresses applied to the absorber drawn from the operating conditions) and with several grain sizes ranging from ASTM < 0 to ASTM 8. The measurements taken on a continuous basis consisted of measurement of the axial deflection of the plates applying the force, giving a deformation/time curve and grain size measurements after testing for assessment of material changes. Figure 2 shows an example deformation/time curve obtained by creep with a temperature of 350 C and a stress of 3.5 MPa.

Irrespective of the initial grain size, all the tests show a creep rate trend which can be split up into:

a period of 0 to 160 - 200h corresponding to primary creep; secondary cree rata t pea diminishin stepn g i tes r sfo t temperature. C 0 35 s

graie Th n size measurements confirmed thasecondare th t y creep tren dependens di graie th nn o t size growth durin teste gth .

194 350°C/o=3,5 MPa

c o "•S (0 E

0) Q

0 200 400 600 800 1000 1200 1400 1600 1800 Primary creep 700 h secondary creep after coarsening grain

secondary creep without coarsening grain Tim) e(h

FIG. Typical2. AglnCd thermal creep

3.3. AglnC pilf o t e creedou w pla

primars A y cree shors pi duration i t 200hn< , onl yformula secondarr afo y creep rat eg measure d experimentall soughs ywa t versu parametere sth s stress, temperature, grain standarsize timed th ean n i , d form: s =Kdncrr exp—— RT where: £ : creep rate K, n, m : constants : pre-creed p grain siz micronn ei s a : stresMP n si Q: creep activation energy in Joule/mole : idea constanR s ga l t temperatur: T n eKelvini .

A simplified way of taking into account grain coarsening was applied by discretizing the law into long-tere th ; h) implicitl w 0 m la long-tera 70 Figur- d > h ( an y0 short-ter) a w eallow2 m 70 la < ( sw mla for growth through coefficient. K d an Q s

resulte Th s show creep rate dependenc stresn e o t constan sno t wit graie hth ncreea sizd pean rate inversely proportional to the initial grain size of the material. The creep activation energy is fully comparable wit hdiffusioe th tha f o t f puro n e silver alon graie gth n boundarie short-tere th r sfo m law. This activation energy is drastically reduced when the test duration lengthens, showing the effect of grain growth.

195 3.4. Creep mechanism

interpretatioe Th creee th coefficiente f po th law performef s Q no , swa m , n scomparin y db g them wit creee hth p law purf so e silvedomaie th n r i "hig f no h temperature meltin> "( g point/2 low-stressd )an . e coefficientTh s obtaine tesa r t dfo duratio n shorter than 700h correspon plastia o dt c strain mode controlle diffusioe th y db atomf no s alon graie gth n boundaries (COBLE creep [1]) although Ag-In-Cs di solia d solution with major dilution.

3.5. Consequences

These tests showed that an AglnCd material with large grains (ASTM < 3) drastically reduces the creep rate (Figur ) relativ 3 emateriae th o et l previously used (ASTd an M0 1 6-8factoa o t y )b 5 r afterwards extends the absorber/cladding gap closure time in the period during which absorber creep predominates over irradiation-induced swelling.

• Stress 3.5 MPa - duration<700h duration<700- a StresMP 5 s7. h duratio- a StresMP 700n5 > s3. h duratio- a StresMP n5 s>7007. h

(0 a o a> O

0 1 8 Grain size (ASTM size number)

FIG 3 Gram size effect on creep rate

4. AglnCd SWELLING ANALYSIS

4.1. Examinations of irradiated AglnCd bars

To determin e swellineth g mechanis f absorbemo modeo rt rode d swellinth lan s g kinetics, FRAMATOME and EDF co-operated in a hot cell expert assessment programme using rodlets from 900 and 1300 MWe reactors.

This programme measured the diametral swelling and absorber density trend in three 900 MWe RCCA foud san r 130RCCAe 0MW s from different banks (control, shutdown, power compensation), making eleven swollen rodlets. It was supplemented by a programme characterizing the structure and the modifications made under irradiation [2] with the aid of optical microscopy, microanalysis, metallograph diffractioy ra X d nsan measurements (performe CEA)e th y db , especiall AglnCe on n yo d bar irradiated 8 cycles in reactor.

196 These examinations showe e firsth t n i dplac e thae densitth t r volumetriyo c swelline th f o g absorber changes linearly with irradiatio givea o t np fluencnu e (Figur . Figure4) alse4 o shows that beyond this limit fluence, the change remains linear but with a steeper slope.

14000

12000

HMOO

8000

•WOO

3006

IfttZ to-os

F/G. 4. Gamma rod activity versus AglnCd density

4.2. Swelling mechanism

The metallographs performed on a few samples confirm this difference in behaviour:

up to a dose limit, the microstructure is principally single-phase face-centred cubic (fee). During transmutation atome frosd th froint, n an mI int g s o d mSn A create oC more dar e voluminou e networ fe e solubl e ar thero th ks d n s i swellinean i se a transmutatio y gb n effect; beyond this dose, the solubility limit of the elements Cd, In, Sn is exceeded. There is occurrence of a close-packed hexagonal phase (hep) more voluminous than the initial fee phase. The structure is then principally two-phase (fee + hep) but at the periphery only the hep phase can be obtained [2].

The additional analyses [2] determined a solubility limit on the order of 1.7 to 2 at.% tin in the fee phase and a minimum 4 at.% tin in the hep phase.

However sampla n o , e temporaril r froba ma y subjecte higanalyseo y dt ra h X flux e s th ,evidence d the presence of singular zones highly enriched in tin up to 34%. These small-sized zones seem localized at the junction points of second-phase domains. This existence cannot be explained by global oversaturation of the material in tin, as the fluence received is not sufficiently high to cause such an enrichment and the micrographs reveal little hep phase. These findings show the importance of the role of transmutations in the material structural changes and justified the need to closely investigate the AglnCd quarternary system to:

gain greater insight into the phases forming under irradiation; predict the proportion of each of the phases in the material; simulat e developmenth e e concentratioth f o t n profilee elementth f o s so withitw e th n phases.

197 joina o Thit t d FRAMATOME/CEsle A research programm thif presentes eo a sm ai 4.4e n di Th . programme is to simulate the growth of a hep phase precipitate on an fee phase grain boundary in the absorber and thus to finally predict the irradiation-induced swelling in the hep phase. The swelling in the fee phase is deduced by the Zen law [3] using the effective atomic volumes of the elements.

4.3. Swelling modelling

densitBasee th n do y measurements taken during examinations macroscopia , c mode f AglnCo l d irradiation swelling was built, with allowance for the accelerating effect arising from the occurrence of seconp thehe d phase r eacFo .h examined irradiated RCCA volume th , e variations alon absorbee gth r (calculated fro densite mth y measurements) were correlated wit distributioha f fluenceno s calculaten do this absorber and adjusted to the values obtained from the gamma activities measured on two RCCAs. A fine fluence calculatio s als nowa referenca carrien o t dou e RCCA e insertionr whicth fo , l hal s were recorded durin cyclee gth .

In the case of RCCAs from temperature control bank R, the overall calculation method overpredicts by about 25% the value estimated by the measured activities. By refining the method with reference th e RCCA whose every insertio knowns ni accurace th ,fluenc e th f yo e calculation improves. Should the absorber be inserted into the fuel stack, the uncertainty obtained does not exceed 11% even for an RCCA frequently in motion (power compensation bank G). The density measurement accuracy leads to an uncertainty on the volume variation measurement on the order of 10 to 20%.

A two-slope swelling model was thus determined by means of linear regressions for the experimental points lyin occurrenc p eacn go he he sidth f eo e fluence. Figur showe5 nominae sth w la l obtained s definea e neutro, th n i d n energy rang correspondinV f e 0.62o e 3 5- o indiugt n ti > m- transmutations, wit experimentae hth l points. Good agreemen obtaines i t minimud dan maximud man m laws bounding swelling were determined.

Law

Fluence (10 20 n/cmV e 23 ) 0.62< E 5< A A

FIG 5. Swelling rate versus fluence

Application to the swelling calculation was performed by considering an isotropic swelling of the absorbe s lona r thers contaco ga n s ei t wit claddinge hth . Processin f examinatiogo n result r crackesfo d rodlets with hard contace absorbeth f d o claddint an r g (absorbe r diameteba r r chang d densitan e y measurements) lead thin si s cas non-isotropio et c swelling

198 An evaluation of the increase in diameter of the absorber during an annual cycle was conducted with this law and compared with the experience feedback from cladding average swelling measurements determined by UT inspection. The calculated value of 21 microns/cycle for bank R is consistent with the average swelling valu microns/cycle9 e(1 measuref o t ) obtainese e dth r RCCAdfo more i. s; e tha0 n50 RCCAs.

The combine f theso e e us dabsorbe r cree d swellinpan e gabsorbe on law n d i thermals ro r - mechanical computation software programme, allows predictio pre-contace th f no tabsorbee timth f eo r change th d clan ei an d swelling after hard contact wit absorbere hth .

4.4. Modelling of AglnCd composition under irradiation

A wide-ranging research programmunderstano t n ru s moded dean wa phase th l e transformation processe e absorberth n i s n particularI . e changth , e undee transmutation e effecth th rf o t e th f o s concentration profiles within the two phases as soon as an hep phase precipitate exists and the determination of the proportion of each of the phases allows evaluation of the quaternary absorber material swelling. The laws of variation of atomic volume with chemical composition in both phases was also determined from representative synthetic alloys.

For this purpose, this research relied upon two preliminary studies:

study of the growth of a precipitate under the effect of the transmutations in the simple case binara f o y alloy Ag-Cd [4]; quantitative modelling of the phenomena of diffusion through the quarternary alloy for both phases [5].

These studies showed that the enrichment in transmuted element at the centre of the mother phase and of the precipitate occurs to the detriment of the growth of the latter as soon as the transmutation rate (neutron flux largs i ) e compared largee wit mobilitth e l hth atomse ral th whes i f yflut e o I . nhigth s xi h diffusioe th d an n coefficient smalle ar s . Thi givn sca e excessive oversaturatio elementn ni core th f et o s a precipitatee th alreads a , y experience tinr dfo .

The effect of increasing the absorber grain size has been looked at. As a result of the diffusion coefficient phasp he e whicth n i sgreatee har abouy rb tha 0 1 tn those phasee obtainefe e ,th thern di s ei effeco n f increasino t g grai seconne sizth n eo d phase precipitation.

4.5. Consequences

Based on the knowledge of the neutron flux and of the axial force exerted on the absorber, the use full-scale th f o e swelling mode conjunction i l n wit thermae hth l creep model make possiblt si predico et t changes in absorber/cladding gap closure and subsequent clad swelling. The favourable effect of the increas grain ei n siz creen eo maintaines pi d under irradiation. Change chemicae th n si l compositiof no the absorber were also evidenced, with the aim of supplying if necessary a further lifetime gain by delaying the occurrence of the second phase.

5. NEW HARMONI™ 2G RCCA

5.1. Description

The new RCCA generation HARMONI™ 2G incorporates an AglnCd absorber material with modified grai grainw sizene n A .siz e value (AST , i.e 3) graiM a . < n diameter greater thamicron0 n12 s (instead of 30 microns for the standard material) is thus obtained and it is verified, through a controlled fabrication process, that ther erecrystallizatioo n wil e b l n risk during operation HARMONIe Th . ™G 2 RCCA present e benefit e HARMONth th l f al so s I RCC d eveAan n more regardin e AglnCth g d deformation which leadincreasinn a o st lifetimef go .

199 5.2. Absorber rod diameter evolution

The principle of Figure 1 can be retained to measure the characteristic RCCA lifetime gain supplied relative to the former design with "small grain" absorber (Figure 6) The decrease in absorber thermal creep through the increase in its grain size provides a significant reduction in the absorber/cladding gap closure kinetics at BOL at the time when the stresses and temperatures are high. The use of "large grains" has no effect on the swelling under irradiation or on the other properties and characteristics of the absorber.

During lifetime and before hard contact with the cladding, this advantage tends to diminish since the temperatures and stresses remain higher with the large grain material because of the slower change in the bar dimensions. This effect has been analyzed and is negligible compared to the gain supplied by the increase in grain size (Figure 6). After the phase of hard contact with the cladding, the time gain is therefor swelline kepth r tfo g phases (Figur. e6)

c0) Swelling rejection criteria

c 0) 0 h. .0n) n e Maxm inu o in J3 fa Standard absorbed ro r E Absorber-clad gap closure initial diameter o o •2 HARMONI 3 initial

Non-HARMONI RCCA lifetime HARMONI improvement HARMONI 2G Exposure duration

FIG 6 Gram size effectabsorber-cladon evolutiongap

6. BORON CARBIDE BEHAVIOUR UNDER WATER CORROSION

6.1. Experiment facility

experimentAle th l s performe characterizo dt leachine eth g behaviou Bf ro 4 C, use conjunction di n with AglnC blace th r kd fo 130 RCCAse 0MW , have been conducte autoclavn di representativn eno f eo in-reactor water corrosio witr no h A1 pellet2OC 4 3-B s different frodesige e 130d mth 0an MW n F [6],ED FRAMATOME decided to launch an in-reactor experiment with two purpose-built RCCAs.

e experimenTh s conductewa t d witinserteC 4 B h d into cladding tubes pre-perforatee th t a d fabrication stage for a few rods of two RCCAs in order to represent through-wall wear in reactor (guide card locatio bottonor wear)mtip . Thes RCCAetwo s have been inserte PALUEdm Lplan2 ban R in kt (subjected to a constant neutron flux) The first (named A) was withdrawn after one cycle and the second (named B) after three cycles .

RCCe Th useApre-perforateA o stw d rods firse th , t wit oblonn ha 0 widt1 m d gm hholan 3 f eo mm long positioned at the bottom of the BC stack made to look like guide card wear and the second rod with a 3 mm diameter hole at the bottom o4 f the rod (25 mm above the bottom end of the cladding tube) to simulate a tip wear. The RCCA B uses the same bottom pre-perforated rod as the A one.

200 These purpose-built rods are designed with a special AglnCd bar of 492 mm height in place of the 1016 mm classical bar. This was performed to obtain a high B'° depletion rate (approximately 4.1020 twelva r destructiofo e 3 montcm r hnpe cycle). These rods have been characterize fabricatioe th t da n stage like pilot rods located symmetricall thoso yt eacn eo h RCCA completa d an s e recordine th f go RCCAs axial position carries swa d out.

6.2. Main results

destructivn no e Th e Eddy Curren firse th t t f inspectiocyclo d een showee f RCCth no t a Ada A heighm s measurem sprine droperforatee th 0 wa t f th 50 ploso gC r 4 f locatiodfo so B a d d oblonnan g hol othee e th rod r columrC 4 oneFo .B o ,nn evolutio detecteds nwa .

The hot cells examinations on RCCA A confirmed this assessment and showed the damages:

a) perforated oblongrod stackC 4 B ;e th f o m m 3 los52 f so peripheral degradatio f som no pelletC 4 eB pellete sth (decreasf o s limitee % th 7 f eo o dt stack;

b) bottom holedrod -

no heighstackC 4 B ;te losth f so peripheral degradation of 10% of the stack-

experimene Th t will continu 199n ei t cell9 ho wit se examinationhth RCCe th n Aafteso B r three cycles.

6.3. Consequences

These in-reactor results show that leachin erosiod gan n durin RCCe gth A stepping together with the primary water media have partially affecte pelletC 4 dB s subjecte: o dt

sufficiena t B-10 depletion (abov threshold)ea ; a water circulation around the pellet.

blace Al130e th l k0MW RCCA implementee sar d wit classicae hth l 101AglnCm 6m d e barth o ,s

B4C pellet f RCCAo s s shutdown f banirradiatioo t k ou rodban R e r like nso ar k eth receive dsmala l bum-up. There would be no loss of B4C height for the usual RCCAs because the B4C stack dimension of these classical RCCAs is in fact 524 mm greater than the tested rod stack. A prolonged irradiation exposur banf eo RCCAkR s could involv esmala l peripheral degradatio pellet firsC e 4 th B t f nso without harmful consequence on the RCCA reactivity feedback.

7. HAFNIUM RCCA DEVELOPMENT

desige Th n enhancements HARMONIe madth n ei RCC™G 2 alloy Ama wfurthe a r increasn ei lifetime with respec swellino t swelline th t gbu g phenomen rulet no de outaar . Thidesigw sne n allowsa less frequent significane contro RCCAe th th o t f e o l sdu t reductio absorber-clae th f no d closure ratee Th . swelling lifetime of such a RCCA is not yet equivalent to the potential wear lifetime.

e hafniuTh m absorbe a neutro s ha r n worth equivalen e AglnCth e o t witon dt h identical dimension d benefisan t from several advantages:

no swelling under irradiation; no creep behaviour (melting temperature near 2200 C);

201 acceptable behaviour under accident conditions due to its high melting temperature; constant neutron worth during their lifetime; possibility to use hafnium rod without clad (owing to the fact of its excellent behaviour under primary water corrosion thed exten)o an n t globae dth l RCCA reactivity feedback.

FRAMATOM s decideEha desigo dt n hafnium RCCAs. Demonstratio hafniuw ne f no m RCCAs in reactor will be implemented in EDF reactor.

8. CONCLUSION

bettee Th r knowledg behavioue th absorbere f eo th f o r s presen in-reactoe th n i t r design t thisa s time and the improvements made by the new HARMONI™ 2G generation are allowing an extension of RCCA lifetime with respect to the swelling phenomenon. The modelling of the phenomena and its comparison with experience feedback will mak possiblt ei offeo et bettea r r predictio thif no s swelling. This enables the RCCA maintenance strategy to be adapted and the potential risks arising from unpredicted RCCA swellin limitede b o gt . Analysi clae th d f sperforatioo npresence th risC 4 n ki B f eo also makes it possible indirectly to extend the lifetime of the RCCAs of previous designs (possibility of accepting perforation). hafniuof swellinthe use rulm to out eg The phenomeno offeto a substantiallr and n y longer lifetime is being experimentally investigated.

REFERENCES

] [1 ASFQ3Y, M.F. firsA , t repor deformation-mechanisn o t m maps, Acta Metallurgica , (197220 l )vo , 887. [2] BOURGOIN, J., COUVREUR, F, GOSSET, D., DEFOORT, F, MONCHANIN, M, TFflBAULT, X., The behaviour of control rod absorber under irradiation, paper proposed to Journal of Nuclear Materials. [3] MASSALSKI, T.B., KING, H.W., Alloy phases of the noble metals, Progress in Materials Sciences, 10,1 (1963). [4] DESGRANGES, C., DEFOORT, F., POISSONNET, S., MARTIN, G., Interdiffusion in concentrated quaternary AglnCd alloys:modelling and measurements, Defect and Diffusion Science Forum, 143-147 (1997) 603. [5] DESGRANGES, C., MARTIN, G., DEFOORT, F., Microstructural kinetics in alloys undergoing transmutation sApplicatio- neutroC AI o nt absorbers, Materials Research Society Symp., Proc., (1997)439. ] [6 EPRINP 4533LD, Research Project X101-6, Behavio irradiatef ro d B4C, (1986).

202 MEASURE OF UB, 10B AND 7L CONCENTRATIONS IN CONTROL RODS USIN NUCLEAE GTH R MICROPROBE TECHNIQUE

D. SIMEONE, X. DESCHANELS, P. HERTER Laboratoire d'Etude Materiaus sde x Absorbants, CEA Saclay, Gif-sur-Yvette, France

Abstract

nucleae Th r microprobe techniqu bees eha n i profileuseL 7 measur o dd t an s afteB e10 r thermal neutron irradiatioa f no boron carbide absorber measuree Th . distributioB dIC n alon e pellegth t radiu s beeha s n use calculato dt e spatiaeth i L 7 l distribution with the hypothesis of no lithium migration. This result shows a discrepancy with the 7Li measured distribution, pelletC 4 B lithiu e e .indicatin th migrat th y f f o m o ma t % eou 40 go that p u t

1. INTRODUCTION

Under a neutron irradiation the IOB(n,a)7Li reaction induces swelling of the boron carbide pellet [1-3] orden I . investigato rt possiblea e migratiosamplee th f i atom10e o L 7 boront th ,f no sou , "boron and the 7lithium concentration profiles have been measured by a technique using different nuclear reaction Pierre Laboratorth e t ea sSu y Microprobe Facility (CEA/Saclay).

2. DESCRIPTION OF THE SAMPLE ANALYSED BY THE NUCLEAR MICROPROBE

Under thermal neutron irradiation and for a mean rate of capture of 20 x 1020 capt/cm3, usual boron carbide is partially fragmented near the surface of the pellet because of the high level of transmutatio thin i sB area10 avoio f T o n. d measurement difficulties neae surfaceth r composita , e material constituted of B4C fine powder (average diameter 2 um) and an organic resin elaborated at 1000° s beeCha n used porosite t e beinmosTh i . th f s o nearlyti g % ope5 y2 n porosity. Froma microstructure poin viewf o t , this materia composes i l f micrometrio d particleC 4 B c s embeddea n di carbonised resin. Some impurities like Mg and Si have been detected by a chemical analysis. This materian initiaa enrichmenB d 10 l ha l t correspondin o naturat g l boron e materiaTh . l obtained conserves its integrity after an irradiation at 20xl020 capt/cm3.

2.1. Sample irradiation

The sample was irradiated inside a steel clad filled with 4He. Two thermocouples measured the temperature on the clad's surface of the sample during the irradiation. This temperature was about e irradiatioTh 59 . 3K n laste days0 thermae 12 dTh . l s fluabou x 10xwa 3 t1 2 e fasn/cmth t d 2/an s neutron flux (> 1 MeV) was about 1 x 1014n/cm2/s. During the irradiation, IOB is transmuted in 7Li accordin transmutes i IO o gi t B(n,a) L 7 e para th 7t f dLito intBu .: H o3

n ' + H 3 + e H 4 » - n ' + i 'L

This nuclear reaction occurs only with energetic neutrons (£„> 2.8 MeV). The cross section of this nuclear reactio abous ni barns4 0. t . Some calculations have been don estimato et burn-upi L 7 e eth . This valu maximus ei periphere th valus t mpelleit e a th abous d ei f yan to t 1016 atom f tritiuo s r mpe cm3, which is about four orders of magnitude smaller than the value of the 7Li at the same position. Thus the 7Li burn-up can be considered as negligible.

2.2. Sample preparation

preparatioe Th sampla experimente f th no r efo mord an , e especiall polishins yit difficuls gwa t becaus hige th h f eleveo porositf o l y which resulte pulling-oue th n di f mano t y grains thir Fo s. reason, pellee mountes th twa d under vacuu epoxn a mn i y resi polished nan d with diamond abrasive powder.

203 3. THE MICROPROBE EXPERIMENT

During a neutron irradiation, 10B is transmuted into 7Li and 4He according to the reaction:

) (1 ]0B+ W7L 0.8i( 3 ( 1.4MeV e 8H 4 MeV)+ )

The recoil ranges of 4He and 7Li inside the pellet have been calculated by TRIM-91. These atome H i atomsallow4 d L e 7 s an th e concludo r sth t , fo r 3.3 rangefo m 4u m aboue u thasar 7 i L 1. tt7

and 10B profiles are only induced by the neutron absorption self-shielding inside the pellet. Pollard et hav] [4 el a already shown i profile 7L pelletf o s s irradiate a fas n di t reactors usin nucleae gth r reaction 7Li(p,a)a. We have carried out a similar work using other and different nuclear reactions with boron carbide pellets irradiated in a thermal neutron flux.

3.1. Determination of 10B and 7Li production profiles

Durin a gneutro n irradiation e self-shieldin th ,induce B 10 e B th a depletio10 s f e o gth f o n concentration along the radius of the pellet. The goal of this study is to estimate this depletion Na, Na = N('°B,t=0) - N(i0B,t,rr), from the 10-boron measured profile using a proton beam. The number of photons detected depends on the concentration of the elements (10B, 11B, 'Li) analysed according to the equation (2): E < j dE N =kNN(X,x,t,~ J (2) Es dx

X studie: d element (7) Li,B U 10 r Bo : cross section for the reaction considered NY : number of collected photons. dE/dx : stopping power of the matrix inside the pellet Np: number of incident protons : beaf E m energy (3.2 MeV). N(X,x,tirr): numbe f targeo r r volumpe t e th uni t a t Es: cut off energy of the nuclear reaction. iradiatio reducee nth t duratioa d abscissan a nt, . ax knumeri: c factor dependin geometre e th f th g o d yan efficiency of the detector

The calculation of the stopping power of the matrix was done along the surface of the sample estimatee foth r d concentratio carbod an 7 f Lino B n,IO usin Zieglee gth Bierzacd an r k formulad an ] e[5

the Bragg's rule. The value of the stopping power of this irradiated pellet does not differ by more than 0.5% fro stoppine mth g powe matria f o r x estimated witcarbod han 10B n only. This last e valuth f eo

stopping powe s beeha r n use o computt d e dE/dx wit a gooh d accurac e BCth . Unden i y r this 10 n 4 assumptio possibls i t ni quantifo t simplif o e) t (2 ratie . yth oyEq ( NB /sample NB)th r *fo e analysed. IO I! Then the ratio ( NB/ NB)* in the irradiated pellet is given by the following relation :

10 U 10 n 10 10 ( NB/ NB)* = ( NB/ NB)x( N7/ NT) (3)

stae rTh point t datsou a afte irradiatioe th r nucleae th n ni r reactor. This formula allows to obtain the° l B ratio value using only the gamma ratio of the detector. Figur present1 e e calculatesth measured dan boro0 1 d n profil studien i e pelletsC dB^ e anne Th . x shows further results of a lithium migration study inside B^C.

3.2. Determinatio concentratioi L 7 e th f no n profiles

A reliable standard pellet of B4C including a known concentration of 7Li could not be done to determin i profilL 7 e eth usin . (3) r thigEq .Fo s Nationaa element e us e lw , Burea f Standaruo i L d7 standar standara d( d 7f Lio containin determino )t m i concentratiopp L 7 0 e g50 eth irradiatee th n ni d pellet. The 7Li concentration in the sample is then deduced from Eq.(4) which is a ratio of two Eq.(2):

204 100-

Mesure microsonde Programme

-100 -080 -060 -040 -020 000

FIG. 1. Comparison between calculated and measured Na profile along the radius of a B4C pellet (dots refer experimentalto datafull calculated and linethe is profile)Na

The first concerning our irradiated 640 sample, the second concerning the NBS sample also measured microprobe byth e:

f i y dc - f -~•" dE o /"> N T . 4 B4C LI dQ N Es Li r "dx (4) N NBS ^ d NB j S dE Li N Li Es dQ -dx'

7 7 Es, the cut-off energy of the Li(p,p'y) Li reaction (400 keV) [6].

The 7Li standard is a cylindrical sample with the same radius as the boron carbide pellet. The number of photons measured on the standard were separately collected in the same position as the photons of the boron carbide pellet. Therefore, the photons detector ratio and the solid angle were absolutely identical.

Accordin e ablEq.(4o gt ar determino et e w ) e directl i experimentaL y7 l concentration profile insid e irradiateeth d pellet e knowledgTh . experimentan a f eo profilB 10 l e allows als calculato ot e indirectly the 7Li production. The number of 7Li atoms generated in the irradiated pellet is given by the relation :

C* __ Ck *£ N, . = T7L, (l-e*) 10B (5)

e and e* are respectively the 10Br enrichment (Nio /(Nio n irradiateno f Nnd o an )d) BD DB D B IO 3 irradiated material and N]OB is the number of B atoms per cm referred to the unirradiated pellet. Figure 2 shows the 7Li profile production deduced from the nuclear microprobe experimental profile e directh d t an experimenta , B U d an B l 10 measur f i o atomL 7 f o es profile obtaine e nucleath y db r microprobe.

This work shows that 7Li profiles increase from the center to the periphery. At the center of the pellet, the calculation is in agreement with the measurement, and the 7Li concentration seems to be measured with a good accuracy by the microprobe. But there is a large difference near the surface of

205 pellee th t betwee e experimentanth i profilL 7 l e measure i calculateL d7 directle th d dy an production . This differenc e explaine b e pellet i th diffusioy recoie L f 7 o ma Th e. y t db l i atomou rangnL 7 f eo s calculated by TRIM-91 [5], is 1.7 (am. The recoil length of 7Li in the (n,cc) reaction is then smaller than the grain size (« 2um). That is why recoil should not be responsible for the loss of 7Li near the interface between the pellet and the resin.

i S*

i r i 020 080

FIG. Comparison2. between profiles measuredLi and alongNa radius pellet,the B^C ofa black dots refer to Na data and -white dots refer to 7Li data

The 7Li production provides an average value of 2.44 x 102! atoms/cm3 and the average value of experimentae th i measuremenL 7 l 10x 4 211. atoms/Ctrl's i t 3, indicating that approximatele th f o % 0 y4 7Li produce nucleae th y db r reaction shoul pellete artefacn da th diffusf o t o . e t Thit du ou e st losno s si during the sample preparation, because the value of the 7Li concentration at the center of the pellet is calculatee same th th r efo measured dan profilesi L d7 .

4. DISCUSSION ON 7LI DIFFUSION IN THE BORON CARBIDE PELLET

7Li profiles obtained fro nucleae mth r microprobe experiments indicate possiblsa i diffusioL e7 n boroe inth n carbide absorber.

The 7Li transport may occur in tree different steps [7-9]: (i) Intragranular diffusion by a driving force (chemical potential); (ii) Desorptio surface graie th th n 7 f nf Lino o e o ; (iii) Pore diffusion of 7Li.

Such a transport mechanism may describe the loss of 7Li observed in Figure 3. These phenomena have been observed in nuclear fusion problems for tritigen ceramics [10].

Gram

Boron carbide grain Surfac grae th mf eo

Open porosity

FIG. 3. Schematic of Li transport in the boron carbide ceramic

206 Some authors have reported the formation of Li2C2[l 1], LiBio[12], Li4B5[13] in the system Li-B Li-Cd an . Therefor difficuls i t ei kno o samplet e t chemicae th w th n suggese i i W .L 7 le fortth thaf mo t the diffusion inside the grain is due to a chemical potential gradient. At the surface of the grain, the 7Li may desorb and goes away through the open porosity of the boron carbide pellet.

Anothe r y segregatissuma s thai i eL t 7 tgraia e n boundarie d coulan s d thus modify their cohesion. This hypothesis verif e coulb t ydno becaus graie eth n siz observo t e ) wer n smalo ur eto 2 ( l a segregation at the grain boundary with our size of proton beam. If we had reduced the size of the protons bea woule mw d ver a havcoun d d yba e ha t rat photonsf eo .

5. CONCLUSION

comparisoe Th experimentae th f no calculatee th d an l d profile showB f so s tha e spatiath t l distribution of the IOB in the pellet is correctly described by a simple neutron capture calculation,

which e implieirradiateth n i o diffusio n sB d1C borof o n n carbide pellet e havW . e experimentally shown that there is La7 i diffusion in this boron carbide pellet, representing a loss of about 40 % of the 7Li atoms produced by the nuclear capture reaction. According to us, this diffusion should be essentially due to the chemical potential of 7Li atoms.

7Li diffusion may play a role in the brittleness of boron carbide absorber during a neutron irradiation i diffusioL 7 boro e e th Th .n ni carbide pellet lead relativa o st e mass los f abouso 10~x 6 t 3. This value is not very important in our case. But in Fast Breeder Reactors or for longer irradiation time PWRa n i s , this relative mass variatio muce b y h mornma e important. Therefore affecy ma t i , e evaluatioth f swellino n boron gi n carbide absorbers, since these calculation e deducear s d from density measurements. Moreover e absorbe th e diffusio f y brini th atomo , L 7 t ma rgf ou s o n thes e atoms in contact with the metallic sheath.

REFERENCES

[I] STOTO, T., ZUPPIROLI, L., PELISSIER, J., Radiation effects, 90, p 161-170, (1985). [2] KRYGER, B., COLIN, M., SPECIAL MEETING ON ABSORBING PINS AND MATERIALS, CEA-CONF- 6656, (1985). [3] HOLLEMBERG, G., BASMAJIAN, J., J. Am. Cer. Soc., 65, p 179, (1982). ] [4 MCMILLAN, J.W., PUMMERY, F.C.W., POLLARD, P.M., Nucl. InstMeth.d an . , 197 ,171-177p , (1982). [5] ZlEGLER, J.F., Handbook of stopping power of energetic ions in all elements, 5, Pergamon Press. [6] JARJIS, R., PhD Thesis, University of Manchester, (1979). ] [7 OLIVIER , PEISACHC. , . RadioanalJ , M. , . Chem. 389p , ,98 , (1986). ] [8 RAFFRAY, A.R., CH, S. , O ABDOU, M.A. . Nucl,J . Mater., 210 ,143-160p , (1994). [9] FREDERICI, G., Wu, C.H., RAFFRAY, A.R., BILLONE, M.C., J. Nucl. Mater., 177, p 1-31, (1992). [10] RASNER , CEA/CEREB. , M ref94/01T .N 4 BR/FL (1994). [II] MAHAGIN, D.E., BATES, L.L., BAKER, D.E. report HELD-TME/77-33, (1977). [12] SECRIST, D.R., J. Am. Cer. Soc., Vol. 50, N°10, p. 520-523, (1967). [13] WANG, F.E., MICTCHELL, M.A., SUTULA, R.A., HOLDEN, J.R., BENNETT, L.H., J. of Less- Common Metals, 61, p237-251, (1978).

207 ANNEX

STUDY OF LITHIUM MIGRATION INSIDE BC NUCLEAR CONTROL RODS

USIN NUCLEAE GTH R MICROPROBE TECHNIQU4 E

Neutrons

Thermal neutrons E

dpa °B(n,a)7Li reaction

V

Li,Te H , Heat

Neutron absorption boron-10by

CALCULATIO HELIUF NO LITHIUD MAN M PROFILES

• Physical phenomenon:

- Self shielding of boron-10 atoms

1,0

0,2 0,4 0,6 0,8 r/R reducee r/th R s a-dimensionai n a d s abscissi £ d aan l number

208 RESULTS OF CALCULATIONS

8E+14

6E+14

4E+14

2E+14

OE+0 1.00 0.80 0.60 0.400.20 0.00 Rayon reduit

Helium profile in PWR

Helium and dpa production rates along the B^C pellet radius

Caption :

black dots : 6.2 10m 2nc ' -2

white dots : 2.1 1021 ncm -2

white lozenges : 1 1019ncm'2

209 MEASURE OF LITHIUM PROFILE BY NUCLEAR MICROPROBE

Choice of nuclear reaction avoiding some surface effects (rugosity).

Protons beam detector

7 7 Li(p,p'Y) Li V ke 8 47 E Y= 10 7 B(p,ay) Be E 429keT= V n "B(p,p'y) B Ev = 2125 keV

210 RESULTS

PWR condition ( Bigarreau experiment)

100

so -. Mcsurc mscrosonde

20

-1 00 -0 80 -0 60 -0 4C -0 20 0 00 r/R

100

Ls mesurc 80

60

40

20

-1.0 -0.8 -0.6 -0.4 -0.00 2 r/R

Evidence of lithium migration out of the pellet

211 MEASUR LITHIUF EO M DIFFUSION COEFFICIENT

Proton beam

B4C pellet irradiated in FBR

6.00

\+

4.00

•4-

2.00

0.00

8.00E-4 1.20E-3 1.60E-3 2.00E-3 1/T (1/K) Do=10-14cmV

Lithium diffusion coefficient as a function of the inverse temperature

212 MODEL OF LITHIUM MIGRATION

gram

Interconnected porosity

Grain boundary

Schematic presentation of lithium migration

• intragranular diffusion of 7Li atoms

• desorption of lithium atoms at the grain surfaces

• gazeous diffusion of lithium atoms in the open porosities

CONCLUSION

Interest of the nuclear microprobe technique :

• Possible validation of neutron codes by measuring the effect of the neutron flux every where insid nucleaea r reacto indirecn a y rb t method;

• To obtain a better understanding of B4C evolution under nuclear conditions;

• Experimental evidence of lithium migration outside of B4C pellet, this may affect the corrosion of metallic sheaths.

NEXT PAGE(S) left BLANK 213 THE ROLE OF CLADDING MATERIAL FOR XA0053643 PERFORMANCE OF LWR CONTROL ASSEMBLIES

P. DEWES, A. ROPPELT Siemens AG, Unternehmensbereich KWU, Erlangen, Germany

Abstract

The lifetime of control assemblies in LWRs can be limited presently by mechanical failure of the absorber cladding. majoe Th r caus f failureo mechanicas ei l interactio absorbee th f no r wit claddine hth irradiatioo t e gdu n induced dimensional changes suc absorbes ha r swellin claddind gan g creep, resultin crackinn gi clade th f .g o Suc h failures occurre botn di h BWRs and PWRs. Experience and in-reactor tests revealed that cracking can be avoided principally by two ways: First, if strain rate henced san , stresse claddine th n s i kepe (welw gar lo t l belo yiele wth d strength), significant straintoleratede b n sca . This is the case for the cladding of PWR control assemblies with slowly swelling AglnCd absorber. Recent examinations of highly expose controR dPW l assemblies confirme desige dth n correlatio presentle th o t p nyu used strain limit. Second sucn i , h cases where strongly swelling absorber material like boron carbid stils ei l preferred, materials whic resistane har t against irradiation assisted stress corrosion cracking (IASCC usede influencb e n Th . )ca f materiaeo l compositio conditiod nan IASCn no s Cwa studied in-reactor using tubular sample f variouo s s stainless steel Ni-basd an s e alloys stresse swelliny db g mandrelsn I . several programme steps high purity materials with special features had been identified as resistant to IASCC. Another proces claddinf so g damage whic occuy hma PWRn i r weas si r cause frictio y controde b th f no lsurroundine rodth n si g guide structure r replacemenFo . t control assemblies this proble s msolvei coatiny de b cladding th f go . There exists meanwhile excellent operatioexperienc8 1 o t p u n f ecycleo s with coated claddings.

1. INTRODUCTION

For safe operatio shutdowd nan f nucleano r reactors control assemblies f essentia(CAo e ar ) l importance. According to the different situation regarding design limits, operating conditions and demands in the various types of reactors different solutions were found for design of CA and choice of absorber and structural materials. Common requests to all types of CA are safe function under all normal operating conditions, tolerable behaviou casn i r f accidenteo s and economir fo , c reasond san waste minimization lons a , g life tim possibles ea .

The natural end of life is determined by reduction of the neutron worth of the CA due to depletion of the neutron absorbing isotopes. The limit, normally set to 10% reduction, would allow for very long life time r examplefo ; clusted casn ro i , f eo r control assemblies (RCCA) with AglnCd absorber widely use PWR'n di operatine sth g lif highes i e r tha years0 n3 . Experience, howevers ha , shown that mechanical failures prematurcoulo t d f lifed le o mos n I d . een t case clae sth d surrounding the absorber material was concerned. Either the cladding was cracked due to interaction with the absorber or it was damaged by fretting wear.

Considerable effors spen wa n investigatioto t e rooth tf o ncause f suco s h failuresn o , understanding of the basic mechanisms and on remedies to overcome the problems. As one consequence investigation and t leas, introductioa , casHf n i tf absorbeeo w ne f no r materials began leay an significano ddt ma t improvement othee futuree th th n rn si handO . large th , e experience with absorbere th s AglnC Bd 4wels Cdan a ,certais a l n limitations regarding , e.ggiv.CA e weighe th f o t rise to stick to the proven absorber materials and to rely on design modifications and improvement of cladding, instead. In this paper, the role of the cladding for CA performance will be reviewed based on Siemens' experienc irradiatiod e an wit A hC n test sample botn si h PWR BWRsd san .

2. PAST EXPERIENCE WITH CLADDING FAILURES

Performance of CA in PWRs and BWRs was subject to many publications and reviews, e.g. [1-3]. Regarding cladding failure the observations can be summarized as follows:

controR BW l rod s• with vibro compacte powdeC 4 B d r containe stainlesn i d s steel cladding tube r holeso bladen i s s showed crackcladdine th n i swashind f gan o t gou

215 the absorber when exposed beyon certaida n limit. Crackin claddine th f go causes gi d by irradiation assisted stress corrosion cracking (IASCC). In order not to exceed 10% reduction of neutron worth neutron dose limits for the most exposed parts of the control blade usede sar ; • In several PWRs, RCCA suffered from wear of the cladding caused by vibrations of the rods in the guide structure and fuel assembly guide tubes; higt A h exposures • , beginnin about ga t 0.5xl022 cm~IMeV)> RCCR E ( 2 APW , with AglnCd absorber can have single rods with hairline cracks at the tip. The cracks are caused by diameter increase of the absorber due to swelling and deformation under axial forces.

These problems were subject to extensive investigations and improvements of design and materials followine th n I . g result in-reactof so r material test programmes relevan claddinA C r e fo t gar summarize compared dan performancA C o dt e data.

3. RESULTS OF MATERIAL TEST PROGRAMMES

o investigatT e behaviouth e f absorbeo r d claddinan r g materials under original irradiation conditions several material test programmes have been started since 1979 in commercial PWRs and BWRs [4-7] materiale Th . s tested comprised:

• Stainless steels, types 304 (non-stabilized), 316, 1.4541/321 (Ti stabilized), 1.4550/348, Nb stabilized), 1.4981 (Mb stabilized); basi N e allo • y (60%Ni/22%Cr/8%Mo/4%Nb)typ5 e62 ;

• Absorbers: AglnCd, B4C pellets and vibrated B4C powder, A^Os with 3 or 5% B4C addition, hafnium.

The test specimens consisted of tubes made of stainless steels or Ni base alloys and sealed with welde d plugsen d . They were filled with absorber materials eithe o allo t t larg a r rp wfo ga e determination of undisturbed swelling and creep properties, or at almost zero gap to study interaction between absorber and cladding. The geometric proportions in radial direction were representative for typical absorbe , wal 10.D rmm l 2R tube(O thicknes R BW s r eithePW fo r s r fo ro 0.5-0. ) 6mm application (OD 6.7 mm, wall thickness 0.5-0.84 mm). The length of the specimens ranged from 60 mm (some B4C specimens) over 130 mm (majority) to 1700 mm (BWR absorber rods filled with B4C powder).

specimene Th s were inserted into fuel assemblie irradiated san hige th hn d i flu x R regioPW f no coresR anBW d. During refuelling shutdowns examinations consistin f visuago l inspection, eddy current testing and diameter profilometry were performed. Irradiation lasted from one to eight operational cycles, dependin experimentan go l objective performancd san specimense th f eo . Parf o t the specimens was taken to hot cells after irradiation for further examinations concerning especially the internal absorber materials.

3.1. Cladding creep

Whereas under LWR conditions thermal creep of stainless steels and Ni-base alloys can be neglected, measurable irradiation induced creep can occur in the high flux region of reactor cores. Creep down of cladding tubes especially at the high pressure in PWR coolant water is significant and leads to somewhat earlier interaction of absorber and cladding [8].

The results obtained from material test irradiation under PWR conditions indicated that creep is linear with fluence and no primary creep exists. An extension of the in-pile test programme which included several type f commerciao s higd an lh purity austenitic stainless steels revealed similar results. Different alloy compositio r differenceno manufacturinn i s g process resulte variation di n ni creep strength by a factor of two [9].

216 3.2. Cladding load due to absorber dimensional change

During service absorber materials take up neutrons and thus undergo transmutations which cause severe modification materiaf so l properties. Claddin absorberf go mostls si y affecte volumy db e increase of the absorber. Swelling can use up the internal diametrical gap between absorber and claddinn subsequentlca d an g y strai e claddingnth e swellinTh . g rate depends verye mucth n o h absorbe highese th r r materiafa t y swellinb s shows a l ha Figurn C ngi 4 ratB . e1 whic casn i f s eo hi pellets directly transferre claddine th s bee o dt ha ngp filledwhega e nth . Pellets mad f A^Oeo j with small addition shoC 4 B w f o sinitiall same yth e swelling rat s purpelletsea C 4 eB . Depletioe th f no neutron absorbing isotope B-10 reduce e swellinth s g rate significantly with increasing neutron fluence powdeC 4 B . r compacte claddinn di g tube theoreticaf o abou o st % 70 t l initialldensite us n yca the internal void volum compensato et particlesC e4 swellinB e contrasth n containinI .C f 4 go B o t t g ceramic materials, the ternary alloy AglnCd exhibits swelling at a constant, low rate.

/ L / B4C Pellet

^ ^ / .^ ^ s / /

_..-; L B4C Powder 3123456 7891 Neutron Fluence 1021/cm2 (E > 1 MeV)

FIG. 1. Diameter change of stainless steel tubes due to swelling of various B4C containing absorber materials compared swellingto ofAglnCd rods

As a consequence of the different swelling characteristics and the mechanical properties of the ceramic B4C and the metallic AglnCd interaction with the cladding, the stress on the cladding is different. Figur showe2 s example f stainlesso s steel clad samples strainepelletC 4 B y sdb (a)C 4 ,B AglnCn a powded an d ) rodle(b r t (c). Obviously pelletC B , s have more strength than stainless steel

claddin able ar modeo e t d claddine gan th l g accordin theio g4t r local modification diametere th f so r Fo . example, enhanced diameter increas pellee ende th th f st o e a stac k indicate peakind en n sa g effecf o t neutroe th n flulocad xan l enhanced swelling. Despit f suceo h local effects, however, overall straining claddine th f o isotropis gi axia n ctangentiai d an l l directio thin ni s case.

217 Diameter Chang) e(% after cycle 7

40 60 80 100 120 Axial Position (mm)

a) B4C Pellets

iameter afte irradiatior5 n cycles (spiral trace)

Diameter (mm) 6,8

as fabricated diameter

6.7 ~ 430 mm

PowdeC 4 bB ) ) D. r . (CompacteTh f o % 70 do t

Diameter Change (%) after ryde

-1 40 60 80 100 120 Axial Position (mm) c) AglnCd Rodlet

FIG. 2. Diameter profiles of test specimens filled with different absorber materials

218 powdeC 4 B usualls i r y filled into cladding tube section f certaiso n compactes lengti d han y db vibration to reach the required average density. Due to this process some axial fluctuations in density may occur which are reflected by the diameter change profile, see Figure 2b. Maximum strains occur over considerable axial length of the rod and are higher than average strain. Measurements of rod length sho increasw an lengtrod ehof approximately equa averagto l e diameter increase.

In contras ceramice th o t metallie sth c alloy AglnC rathes di r soft. It's swelling behavioun rca therefor e influenceb e e retardinth y b d g momen claddine th f o t g supporte outey db r coolant over- pressure. Thi s obviousi s fro e diametemth r profiles measured successively during irradiatioa n no short stainless steel specimen filled with AglnCd at very small gap, see Figure 2c. In the middle part specimee th f o diametee nth r chang equivalens ei unrestrainen thaa o t f o t d AglnCd sample. Towards both ends, however, widening of the cladding decreases and about 1 cm from the ends of the AglnCd rodlet even creep down of the cladding is observed. Apparently, the AglnCd extends in axial direction filling available void volumes (chamfe f AglnCo r dplud rodleen ghollo d e facan tth t bottof ea wo m sidep claddinthue d to th t ,) an a s axiao t give p y gga lwhic swa presses hi d externae dowth y nb l coolant pressure. In such a case, the stress distribution in the cladding tube may differ from the isotropic casstatn i f swellins eeo a morge b ceramicy e pronouncema d san tangentian di l direction.

3.3. Deformability of cladding materials under irradiation

To prevent cladding failures of CA in operation and to introduce improved materials for new CA it is necessary to know to what extent cladding materials can be deformed under irradiation without fracture.

In several joint EPRI/Siemens programmes, deformabilite th f austenitiyo c stainless steeld san Ni-base alloys under original irradiation conditions in PWRs and BWRs was studied. In the first programm i containinS ed phasan P g e w stainlesonllo ya s steeresistans e founb wa lo t d t against environmenR PW IASCd an R bot Cn tsubsequeni e [4]th h BW n .I t second programme phasth f eo e these results could not be verified with other high purity material charges [5]. After the influence of Ni was ruled out in a further test run [6], the experiments finally succeeded in finding an appropriate microstructure of austenitic stainless steels which makes the alloys insensitive to IASCC. A twofold approach turned out to be successful for alloys of Types 304 and 348. First, the process temperatures during tube manufacturing were kept low resulting in materials of fine grain size. Secondly, the Typn i alloy 8 C e 34 d contents an wer b N e f varieso thren di e steps [7].

The specimens were tested first in a PWR. The results revealed that charges of Type 304 and 8 allo34 y which failed belo e earliestrai% th w1 n i nr test when teste coarse-grain i d n condition survived more than 2% strain in the fine-grain condition. High concentrations of Nb and C were found havo t effeco en r Typ allofo 8 t graie e34 s lon yth a ns g a keps siz ewa t low sucn .I h materia fina l e grain size can be reached with standard fabrication routines, also [7]. In these swelling mandrel tests, e e claddinstresth th n i s g resulting from straining depend e creeth n po s strengt f claddino h d gan mandrel. For ceramic mandrels the creep strength can be assumed to be considerably-higher than for claddinge th thao s , t mandree cree neglectede th b f po n ca l . Sinc in-reactoe eth r claddine creeth f po g materials tested hers beeha e n measured also, stress historiee calculateb n ca r ssevera fo d l combination f claddinso swellind gan g mandrel materials.

Figure 3 shows for example stress histories of three Type 304 specimens. They can be compared wit e yielth h d strengt f thio h s alloy which increases with neutron fluence. Calculated stresses above yield strength are not meant real, of course, but indicate plastic deformation. In the non-failed specimen the applied stress clearly remained below yield strength during all the irradiation time n anotheI . r specimen e calculateth , d stress reache e yielth d d strength earl n irradiationi y . Inspections and examinations of the specimens could not unambiguously exhibit cladding failure. The third specimen, however, which clearly exceeded the RT yield strength, failed with long intergranular cracks contrasn I . thato t t , alloys whic resistane har IASCCo deformee t t b n ca , d plastically without fracture showas ,5. Fignin and .4

219 1400 - failed probably failed - sound 1200

1000

_____-- Yield Strength 800 0) •*-* CO 600 T3 *•0*) JO 5% B C 4 400 ¥O (C O C 4 ----B -

__VT__-.T56~-«-. AI 0 200 2 3 Fine Lin Diametee= r Change Bold Lin Strese= s r 3 10 11 Neutron Fluence (1021/cm2)

MeasuredFIG.3. diameter change calculatedand stress history of three tubetype304 samples with different swelling mandrels; a) rapid diameter increase ofAl2O3+5%B4C caused stress in cladding tube reachto sample the yield and failed; similarb) rapid stress increase caused sample probably failto quickly, thereafter washout ofB4C slowed down further diameter increase; slowlyc) swelling AlzOj pellets resulted stresslow in well below yield strength claddingthe and remained sound

1200

Fine Line = Diameter Change Bold Line = Stress

0.0 10 Neutron Fluence (1021/cm2)

FIG. 4. Measured diameter change and calculated stress history of a type 348 tube sample with Al^O^5%B4C pellets

220 After the successful tests with IASCC resistant materials in the PWR environment, the same tests were repeated in a BWR. The results were about the same; the fine-grain materials again were abl reaco et h high deformations without fracture.

6- 1400

1200 B.C-Pellets 0) D) 1000 | 4-1 O 800

E v 05 b 600 -o £ 2-| 3 400 ^ cOuJ O ^ H Fine Lin Diametee= r Change Bold Lin Strese= s 200

T i 2 10 Neutron Fluence (1021/cm2)

FIG. 5. Measured diameter change and calculated stress history of a type 348 tube sample -with B^C pellets

4. IRRADIATION TEST RESULT CONTROD SAN L ASSEMBLY PERFORMANCE

As showe swellinth n ni g mandrel tests e straith , n rat r streso e s level govern e damagth s e process. Furthermore knows i t i , n that IASCC needs certain neutron exposure; dose threshold values of 0.05xl022cm"2 for BWR (non-hydrogenated water chemistry) and about 0.2x1022cm~2 for PWR environment have been derived [10].

An appropriate way to summarize the data and to define failure and non-failure regimes was shown in [11], where experimental results of the first irradiation tests were presented in a map of tangential strain and fast neutron fluence. In Fig. 6 this map is extended to higher neutron doses to include recent data of the irradiation tests [5-7] and of examinations of RCCAs in several PWRs. Figa show6 . s data from commercial purity heat f typo s e 1.4541 (321) teste n irradiatioi d n programme used s RCCan sda A cladding. Data from several high purity austenitic stainless steels (types 304 1.4981d , an 348 6 )31 , founsusceptible b o dt IASCCo et e Th . , also6b showe g ar ,Fi n ni critical strain at which fracture can occur decreases with increasing neutron dose to very low values below 0.2% above 0.2xl022cm"2(E > IMeV). The lower boundary curve for failures shown in Fig. 6 was derived for commercial purity alloys, The results obtained for several high purity alloys indicate in some case slightlsa y higher threshold generae th t bu , l sametrene th s di .

shows A [11n ni t lowea ] mediud an r m neutron doses, there exist criticasa l strain rate; below that the non-failure/failure threshold curve can be exceeded without failure. The data indicate that the critical strai slightle b n y rateyma highe r higrfo h purity tha r commercianfo l purity alloys.

221 K) O o OQ 3 T3 3 8 -1 Average Diameter Increase o Average Diameter Increase p p N to E.' p o b bi en b 01 X! b en o Ol b 01 c p I i I c p i I I i v: o b

O p , o to vc: KJ C/l

I

o O P ^ p (D CD (D CD

i-t- r-»- I O 3 p 1 P. fc CO 3 CO c c (D (D 3 o 13 O O (D o M M O o 1 • O *-^. 3 § p p "P 7 •b -p CD I (D I (D (D 00 a. Q. O to to b b 4.1. Slow strain rate (PWR RCCA with AglnCd absorber)

mediud an w mlo t neutroA n doses performance th , RCCf eo A wit slowle hth y swelling AglnCd absorber is well in agreement with the conception of a ,,safe strain rate", which allows to cross over the failure threshold (see Fig. 6a). It is known, however, that also AglnCd can cause cracking of stainless steel cladding of PWR RCCA [2, 12, 13]. Such hairline fractures have been observed on a few control rods in one Siemens plant. The RCCAs containing single fractured rods were operated beyon conservative dth e design limit [8]problee Th . recognizes mwa timn d i earl y eb y examinations and, afte year1 2 r f operationo s RCCe th , f thaAo t plant were replaceonew ne s y withoub d y an t complicatio r reactonfo r operation fracturee Th . s occurre t verda y high neutron dose indicates sa n di Fig. 6a. Part of the diameter change measured in the fractured part of the rods is probably caused by crack opening. Accordin [2]o gt , crack widt f sucho h hairlinm m e 2 crack0. s betweed swa an 1 n0. which correspond 0.6o t 3 %0. increaso st diametern ei . Assuming similar behaviou subtractind an r g those values from the failure data would bring the RCCA failure range shown in Fig. 6a down in contace non-failurth o t t e range. Hence e datth ,a suggest that crackin havy gma e occurred there despite of the slow straining rate applied by the AglnCd absorber.

Accordin preseno gt t knowledg irradiatioe th f eo n behaviou f AglnCo r ] swellind[8 lineas gi r with fluence. knowe Baseth n do n creep propertie claddine th f so g materia claddinstrese e th th l n si g calculatee b n ca functios da neutrof no n fluenc initiad absorbee an th l f fillino p r gga (se e 3.3). Fig.7 shows the resulting stress histories of RCCA rods with various as fabricated gaps. When the gap is filled by absorber swelling and cladding creep, stress increases to a certain level and remains constant during further exposure under these assumptions. The maximum stress levels are well below the yield strength which increases with fluence. The maximum stress is the same for rods with small and nominal gap ;uppee gapth n sri tolerance range would have avoide stressy assumes dan i t I . d that fractured rod smald sha l initial gaps strese laste;th d sha d long time, therefore, until rupture occurred.

1.0- 1000

0.8' Fine Lin Diametee= r Change 800 Bold Line = Stress

E E o> 0.6- 600 en CO ro CO Q) O w o> 0.4- 400 o> £ co Nominap Ga l Q .o - Minimap Ga l ro 0.2 200 O

0.0- 0.0 2.5 5.0 7.5 10.0 12.5 15.0 17.5 20.0 Neutron Fluence (10 MeV211 /cm> )E 2( )

FIG. 7. Calculated stress history of a RCCA rod with AglnCd absorber

Presently, it is not exactly clear what mechanism caused fractures of RCCA cladding. The fracture e Siemenmodth f eo s RCC s probablAwa y also intergranular s founa , examinationn i d n so fractured RCCA elsewher , 13]12 ] , state. [2 [2 eSipus l da t thae h t austenitic stainless steey ma l

223 become susceptibl creeo et p ruptur radiatioo t e edu n enhanced segregatio f impuritieno graie th no st boundaries. It is not clear if any corrosive attack by the water environment and, hence, lASCC plays a role in the fracture process at all.

In case of Siemens RCCA, fracture occurred only after favourably long life time when significant strain had been accumulated. Besides, some kind of creep rupture another possible explanation of the late fracturing might be an acceleration of clad straining at very high exposure. Irradiation of AglnCd results in building-up of Sn resulting in a) modification of the chemical composition due to neutron capture leading to expansion of the crystal lattice and b) formation of a second phase with lower density nowo t p ,U .neithe r dimensiona d densitan l y measurementn o s absorber materials [2, 8, 12, 13] nor crystallographic examinations [13] give any clear evidence of an increase in the swelling rate at high fluences due to influence of the second phase.

e dimensionaTh l dat afro w availablmno o Siement p u e s t RCCindicatno o Ad e significant contributio plastif no c deformatio axiao t e l ndu forces . Thi concludes si d fro observatioe mth n thae tth measured date consistenar a t wite assumptioth h f lineao n r absorber swellin e rangth f n o ei g experience up to 6x1O22 cm"2 (E > 0.6 eV), see Fig. 8. Nevertheless, it can not be excluded that at very high exposure r rodfo sd witan s h unfavourable smal sizp rate f diametega eth leo r chango t e edu interaction with the AglnCd increases for reasons mentioned above.

Diameter Change n A o o v RCCA Average Values of Various Plants and Range

minp ga .

J_ 2 3 45 6 ) eV 6 0. Neutro > E ( n m Fluencc 0 1 e*

FIG. 8. Measured and calculated diameter changes of Siemens RCCA rods

For other RCCA designs and operating conditions, cladding fractures have been connected with enhanced diameter increase at the bottom of the absorber rod caused by axial forces due to rapid movement of RCCA (,,slumping"). This might have caused enhanced rate of diametral straining so that cladding fractures could occur at medium neutron doses (0.5 - 0.7xl022cm~2, E > 1 MeV), already, totaw lo anl t dstraina reportes a , , 13]d[2 .

224 4.2. High strain rate (IASCC resistant materials for BWR application)

Requirements for BWR applications, mainly certain weight limits and required neutron worth, make it hard to design control rods without any B4C. The high swelling rate of B4C in the pellet or powder for presentls ma y used lead prematuro st e IASCC failur f claddino e g made from commercial purity alloys. EPRI/Siemens IASCC test programmes [4 - 6] have shown, that also with conventional high purity materials the failure threshold can be increased only slightly as indicated in Figure 6. In these programmes, however, appropriate microstructures were found which make austenitic stainless steels resistant against IASCC [7] s mentionea , 3.3n i d Fign .I , diamete9 . r changes measuren o d IASCC resistant cladding sample plottee sar . neutrodvs n fluence e datTh a. shown here come from PWR irradiation but are thought to be valid for BWR environments also, since it was found in the irradiation programmes that the differences between PWR and BWR environments did not influence the behaviour of high purity stainless steels under high neutron exposure.

D 5- Non Failed Test Specimens Irradiated in PWR Environment 4- 0) D) c CD .c 3- O n

CD 2- E 03 b 1 - B4C Powder Rods of BWR CA Dimensions Tentative Failure Threshold of c.p. Alloys (see Fig. 6a) i 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 Neutron Fluence 1022/cm2

of FIG.experienceMap 9. with IASCC resistant austenitic stainless steels irradiatedPWR in irradiation tests. Range of diameter change for BWR control rod type test rods shown for comparison

To compare the experimental results of the improved alloys with the loads which are to be expected for BWR control rods in service, experimental data from another irradiation programme performed in a BWR with test rods having the original dimensions of B4C rods in a BWR control assembl e showyar Fign ni also 9 .claddine Th . f lattego r test rods madswa e from high purity Type t witye ht optimualloy8 no t 34 bu , m microstructure rode th s, reacheSo . d remarkable high neutron doses, but exhibited cladding cracks at end of life. At that time, the B10 burnout was 45% for the thin- wall clad rods and even 60% for the rods with the thicker wall. A prototype control assembly [14] made from the same alloy and having improved design features like Hf rods in the outermost positions of the blades, was operated successfully during seven years at maximum flux positions up to the high neutron dose of 3x1021 cm"2 (snvt) in the uppermost quarter.

The results show, that significant improvements of the performance of BWR control assemblies could be reached by appropriate choice of the material condition for the cladding of B4C rods combined with a design using Hf absorber at the most exposed positions.

225 5. ELIMINATIO CLADDINF NO G WEA COATINY RB G

In several PWR plants fretting of the cladding due to vibrations of the absorber rods in the guide structure and the guide tubes of the fuel assemblies was observed. Criteria for maximum allowable wall thinning were established accordind an , replacee theso gb t o t e d criterid ha [e.gA aC . , 15] 8 overcom o , .T 2 e this problem, coatin claddine th f go g with hard layer introduceds swa . Siemens uses Cr3C2 coatings applied by a detonation gun process [8]. Other methods to protect the cladding from platinr weaC e rar g [16 nitridinr ]o g claddine [15th f ]o g surface.

The experience with these surface treatments seems to be generally good [8, 15, 16]. Meanwhile, Cr3C2 coated RCCA were in service for up to 18 operational cycles without any indication of fretting attack.

6. SUMMARY

The various mechanisms which can cause mechanical failures of control rod cladding and limit the lif rathew e timeno re welar , l understood.

In PWRs, experienc shows eha n that very long lifetimreachee b n eca d using AglnCd absorber in conventional stainless steel clad with appropriate design. Wear problem e avoideb n f ca si d necessar propey yb r surface treatment.

Control rod design for BWRs apparently can not completely give up B4C absorber, which is putting high demands upon the cladding material. There might exist a potential for further improvement by careful choice of cladding material condition with regard to IASCC resistance.

REFERENCES

] [I STRASSER, A.A., SHEPPARD, K.D., ,,Light water reactor reactivity control", Nucl. Energy 23 (1984) 169-178. ] [2 SIPUSH, P.J. t al.e , , ,,Lif silver-indium-cadmiuR e timPW f eo m control rods", EPRI-NP-4512 (1986). [3] EICKELPASCH t al.e , , N. ,,,BW R control rod s- operationa l experienc posd ean t irradiation examinations", Nucl. Technol. 60 (1983) 362 ff. ] [4 GARZAROLLI t al.e , , ,,DeformabilitF. , f austenitiyo c stainless steel Ni-basd san e alloye th n si core of a boiling and a pressurized water reactor", 3rd Int. Symp. On Environmental Degradatio Materialf no Nuclean si r Power Systems-Water Reactors (1987) 657-664. [5] DEWES, P., et al., ,,Measurement of the deformability of austenitic stainless steels and Nickel-base alloys in light water reactor cores", ASTM STP 1210 (1993) 83-101. [6] GARZAROLLI, F., et al., ,,Deformabiliry of high purity stainless steels and Ni-base alloys in corPWR"e a th f eo h t Int6 , . Symp Environmentan O . l Degradatio f Materialno Nuclean i s r Power Systems-Water Reactors (1993) 607-613. [7] GARZAROLLI, F., et al., ,,In-reactor testing of IASCC resistant stainless steels", 7* Int. Symp. On Environmental Degradation of Materials in Nuclear Power Systems-Water Reactors (1995) 1055-1065. ] [8 HEINS t al.e , , L. , ,,Desig f controo n l rodr pressurizefo s d water reactors with special consideratio absorbef no r swellin clad gan d creep down", IAEA-TECDOC-813, Vienna (1995) 15-36. [9] GARZAROLLI, F., et al., to be published. [10] SCOTT revieA ,, , f irradiatiowo P. , n assisted stress corrosion cracking" Nuclf o . J , . Mat1 21 . (1994) 101-122. [II] GARZAROLLI , F., STEHLE, H., ,,Behaviour of core structural materials in light water cooled power reactors", IAEA-SM-288/24 (1986).

226 [12] MATSUOKA t al.e , , ,,IntergranulaT. , r crackin claddinn gi RCCR g tubPW A f eo rodlets", h t 6 Int. Symp. On Environmental Degradation of Materials in Nuclear Power Systems-Water Reactors (1993) 765-772. [13] BOURGOIN t al.e , , J. Contributio, connaissanca l na comportemenu ed t sous irradiatios nde crayons de grappes de commande", Proc. of Int. Symp. Fontevraud III (1994) 715-725. [14] GROSS, H., et al., ,,The BWR Hybrid 4 control rod", Nucl. Eng. Des. 108 (1988) 433-436. [15] HERTZ , PEYRAND. , , J.C., ,,La nitruration ionique remed e1'usura crayons ede grappee sd s de commande", Proc Intf o .. Symp. Fontevrau (1994I dII ) 752-759. [16] MOON, J.E., et al., Jmproved control rod performance through design enhancements and operational wear management strategies", Proc Intf o . . Symp. Fontevrau (1994I dII ) 726-735.

NEXT PAGE(S) left BLANK

227 XA0053644 BWR CONTROL BLADE REPLACEMENT STRATEGIES

M.W. KENNARD Stoller Nuclear Fuel, NAC International, Pleasantville Yorkw ,Ne , United State Americf so a

J.E. HARBOTTLE Stoller Nuclear Fuel, InternationalC NA , Thornbury, Bristol, United Kingdom

Abstract

The reactivity control elements in a BWR, the control blades, perform three significant functions provide shutdown margin during norma accidend an l t operating conditions, provide overall core reactivity control providd an , e axial power shaping control As such, the blades are exposed to the core's neutron flux, resulting in irradiation of blade structural and absorber materials Sinc absorbee eth r depletes s usedi wit t C alshi 4 , B timo f e structuraswells(i eth d an ) l components undergo various degradation mechanisms (eg, embnttlement, corrosion), the blades have limits on their operational lifetimes Consequently utilitieR BW , s have implemented strategie maximizo t s tham ai t e blade lifetimes while balancing operational costs, suc s extendinha grefuellina g outag shufflo et e high exposure blades This paper examine blade th s e replacement strategie utilitieR s useBW sy db operatin , EuropS U Asi d n gi assembliney an a b g information relatee th o dt utility's specific blade replacement strategy impace th , newee th t r blade design d changean s corn si e operating mode were havin thosn go e strategies e mechanicath , nuclead an l r limits that determined those strategies e methodth , s employeo t d ensure that lifetime limits wer exceedet eno d during operation bladd an , e designs used (curren replacemend an t t blades)

1. VENDOR DESIGNS AND LIFETIME LIMITS

1.1. Overview

Definition of Terras Several terms common to the discussion of control blade exposure and limits are defined below.

1) exposure Nuclea f th lifo e- d eren correspondin reductio% 10 a initian i no g t l cold reactivity worth, Ak/k, averaged over any quarter axial segment of the blade.

2) Quarter axial segment - the significant control blade segment relative to the core cold axial power distribution.

3) snv t- smeare d thermal neutron fluence exposure baseth foue n do th r f fueeo l assemblies adjacen controe th o t l blad interestf eo , unit f 10so 21 n/cm2.

addep Ti r- 4exposur ) e increment applie blada o dt e fully withdrawn fro core mth e (i.e., absorber tip located below the active fuel region) to account for B4C swelling and absorber tube sensitizatio regiop ti consequenc a e s na th n ni thermaf eo l neutron leakage and fast neutron thermalization belo core wth e region.

Background The original equipment blade designs offered by the various vendors consisted of all B4C absorber contained within commercial purity stainless steel materials althoug e designth h s varied slightly Siemend ,an witE hG s encapsulatin e borogth n carbid sealen i e d tubes while ABB-Atom loads the absorber into gun-drilled holes in the blade wings. The mechanical limits of these early designs were base loadn do s expected during normal operatio accidend nan t condition included san d criteri wearn ao , fatigue, seismic loadings internad an , lE pressureG e th e nuclea f lifr o Th fo e. d en r original equipment blades was calculated to be 42% 10B depletion.

229 With irradiation, it was quite apparent that mechanical limits could not be met by the early original equipment designs. In 1978, for example, GE observed absorber tube cracking and boron carbide washou hign i t h exposure blades vendoe Th . r determined that crackin washoud gan t occurred loca % depletioB 50 10 l t a n (Siemens observed similar degradatio 6 snv r 1. ~40o t t a %n local depletion) e mechanisTh . s determinemwa e irradiatiob o t d n assisted stress corrosion cracking (IASSC absorbee th f )o r tubes ligamentd an , s betwee ABB-Atoholee e nth th n si m design, with tubing stresses generate 0 swelling84 y db . Figur providee1 worte sth . depletio hvs n originacurveE G r sfo l equipment blades for the cases where 1) the nuclear lifetime is limiting (42% 10B depletion) and, 2) the mechanical lifetime is limiting due to washout of boron carbide (34% 10B depletion).

0.15

NEW CURVE laC DEPLETION H.OSS! 4

CURVD OL E (B<• * C DEPLETION ONLY)

10 20 30 40

AVERAGE FRACTIONAL DEPLETION (PERCENT!

FIG. Changes1. relativein blade worth functiona as of fractional depletion (Reference:GE)

Subsequent to the observation of cracking and absorber washout in the early blade designs, the vendors have offered mor mord ean e advanced design firso t s t address blade integrity issued an s secondly, to provide designs that could achieve higher exposures. Remedies included three major changes: f higo e h us purity • stainless steel r absorbefo s r tube wingsd an s , thereby improving material ductilit crackind yan g thresholds; • use of hafnium in the high duty blade areas (wing tip and outer edge) to reduce absorber swelling effects; and • increased boron carbide loading effectively reducing the 10B depletion rate and the attendant swelling rate.

Also included wercobalw lo e t material reduco t s e activation e latesTh .t generation blades using these features have effectively tripled control blade lifetimes relative to the original equipment designs. Figur graphicalle2 y show exposure sth e advantag f simpleo y increasin loadinC 4 gB a n gi given blade design.

230 1.0'

0.9' I O.a 0.7 15.

CO 0.5

0.4

0.3 1.00 1.05 1.10 1.15 1.20 1.25 1.30 1.35 1.40 1.45 1.50

Absorber Diameter Ratio

FIG. 2. Impact on depletion ratio of increasing absorber loading

Hafniu s beemha n introduced inte designoth eithen i s rgeometriesd platro r boten o i hd an , cases, unclad. This is also true of the all hafnium Hitachi and Toshiba designs which are available with the absorber configured as either as full length plates or rods contained within a sheath similar to DuralifE G e th e series exampler fo , exceptioe Th . ABB-Atoms ni , which maintain solie sth d stainless steel wing feature hybri e eveth n ni d configurations thin I . s case, solid hafnium rodlet plugd se an sar loaded into selected horizontal holes in each blade wing to provide the hafnium tip and edge regions, respectively.

Table I summarizes the blade designs offered by vendors. Note that Siemens does not actively market replacement control blades. Siemens will, however, provide blade economicaf si . so o d o lt

1.2. ABB-Atom

originae Th l equipment blade (CR-70 all-Bn a s )i 4C design wherei boroe nth n carbid loades ei d into holes bored horizontall stainles4 y 30 int e oth s steel wings. Like other vendors' designs- CR e th , 70 was susceptible to IASCC of the ligaments between the holes.

ABB-Atom introduced the CR-82 (hybrid with hafnium tip) in 1982 and the CR-85 (hybrid with hafnium tip and edge) in 1985. Although lifetimes were improved with the use of hafnium in the highest exposure region designse th f o s , the stile e yar th l s prona S crackino S et 4 30 s thega e yus blade wing material n 1992I . , ABB-Atom introduce CR-82e dth CR-85d Man M designs. These ear simila CR-8e th CR-8d o rt 2 an 5 designs except that absorbee 316 winge th useth s Ld i r san dfo r hole geometry has been modified. PIE examinations have confirmed improved in-reactor performance of 316L vs. 304. Note that all blades manufactured post-1992 use 316L for the wings.

ABB-Atom currently offers two additional designs, the CR-82M-1 and CR-85M-1. These are similar to the CR-82M and CR-85M designs, respectively, except for the absorber hole geometry which is consistent with that of the original CR-82 and CR-85 designs.

231 TABLE I. SUMMARY OF CONTROL BLADE FEATURES

Blade Absorber Type and Placem. Vendor Series TABLE HfTip HfEdge Comments I ABB CR-70 X Original equipment design CR-82 X X ~OriginaI equipment bladf H e+ CR-82M X X M series uses 316L, revised hole geometry CR-85 X X X -Original equipment blade + Hf CR-85M X X X M series uses 3 1 6L, revised hole geometry CR-85M-1 X X X CR-85 with 3 16L; used in Japan GE OE X Original equipment design D120 X -Original equipment blad HP304e+ ; susceptibl crevico et e corrosion D140 X X -Original equipment blad HP304e+ ; susceptibl crevico et e corrosion D160 X X HP304 + Hf D190 X X X HP30f H 4+ D215 X X X HP304 + Hf D230 X X X HP30f H 4+ Marathon X X X desigRaw ne d n+ Resis f absH .t+ tubes Siemens OE X Original equipment design Hybrid-1 X X HP SS + Hf Hybrid-4 X X X HP SS + Hf Toshiba OE X Original equipment design All-Hf X X HP SS + All-Hf

The vendor has developed inspection-based guidelines addressing blade reuse and replacement a culminatio thae ar t f examinationo n s performe o datet r dU.S Fo . . utilities e guidelineth , e ar s relatively conservative. For example, for blades with 304SS wings which have reached an exposure of 3.0snvt:

• visually inspect blade wing cracksr sfo ; • discharge blad crackedf ei ; • if the blade is uncracked, reuse for one cycle and re-inspect applying the above criteria.

If the inspection at 3 snvt shows a blade to be uncracked, it can be shuffled to a shutdown bank. No further examinations are required. However, tip adders are recommended for evaluating CR-70 exposures.

For post-1992 blades using 316L, ABB-Atom recommends visual inspection of only the lead blades of each design. Currently, the CR-82M lead blades are in Unit R2. Inspections have identified limite highese th d f crackino t o exposurtw n gi e blades which reside hign i d h duty control cellr fo s three cycles. Although the root cause analysis is not complete, the location of the cracks leads ABB- Ato mbelievo t e tha failure tth e caus manufacturins ei g related.

Although guidelines for European utilities are similarly based on observations of cracking, ther a muc s i e h larger experience bas f operatioeo n with limited cracking somn I . e cases, utilities reload blades observed to have minor cracking for an additional cycle of operation.

Current replacements are primarily the CR-82M and CR-82M-1 designs.

232 1.3. General Electric

Table I provides a summary of the of the various blade designs offered by GE beginning with the original equipment blade offered in ~1960 to the most advanced, Marathon. One observes a progression in features based on the mechanical limitations identified in the original equipment, Duralifd Duralifan designs0 0 14 e12 e . These advancements include high purity 304SS, redesigned handle-to-sheath and sheath-to-structure joints, and the use of hafnium absorber. The Duralife 200 series used highe loading C Marathone rB th d san weldea , d absorber design exteno t , d lifetimes even

more. 4

Figure 3 provides the 10B depletion rate curves, as a function of exposure in snvt, for two blades operate lata n edi generation BWR/5 plant originae Th . l equipment blade sdepletio % hav34 ea n limit based on cracking while the advanced design blades are limited to 58% 10B depletion. Depletion rates differ because of the higher boron carbide loading in the advanced design blade. The 'real' lifetime advantage of the hybrid blade is the ratio of the EOL exposures in snvts, i.e. 4.39 4- 1.96 = 2.24.

70 - 60 Advanced Blade Limit = 58% - /* S g 50 is - g- 40 / LimiE O t =34% ^ - As c? 30 > # J s + ^ - * j * Bladp eTy e - i § 20 s ,'' ^ ...... Origi ^ nal Equipment (OE) CD * Q. ; '-'^ —— Advameed Blade 10 t * '^^ , *

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

Exposur snven i t

FIG. . Depletion3 ratelifetimed an comparison, original equipment . advancedvs blade design

The original equipment, Duralife 120, and Duralife 140 designs (see Table I) have mechanical limits observatioe baseth n do f crackingno Marathoe Th . n desig limites ni d mechanicall internay yb l pressure at high exposures. The remaining designs have nuclear limits.

GE provides guidelines related to shuffling original equipment blades in Service Information Letter (simila7 (SIL15 ) designs)0 r guideline14 Duralifr d fo an 9 : 0 providee 57 12 es ar L SI n di

exposure th e• limi depletio termB n peai t10 % f sko 34 s quarteni r segment corresponding snvBWR/2 a ~ n i to t 5 plant; exposure th f i depletionless B ei cor10 sw • thamovee ne % e b blade a n20 n th , o dt eca positio d remaian n n there unti s exposurit l e t whicreachea % h 34 stim t musi e e b t dischargedd an ; • if the exposure is >20% and less than 34%, the blade may be shuffled but its residence time is limited, being the greater of:

233 one cycle, or e equivalenth t numbe f cycleo r e bladth s e would have remaine s prioit n ri d position without exceedin limit% 34 .e gth

This methodology introduced a 'phantom' exposure increment. That is, the exposure increment received in the new position was based on the exposure increment consistent with the prior core position. Ther apparentle ear restrictiono yn shufflinn o s g advanced blade design s lona s nucleas ga r and mechanical limits are met.

GE recommends tip adders for non-hafnium tip designs when placed in shutdown positions. Utilities have been reluctan thee us m o t tsinc ) ther1 e e wer requiremento en r theifo s r) 2 use, d an rigoroua t no ther s scorrelationse ebasiwa th r sfo .

1.4. Siemens

Siemens has long had a co-operative research and development program with GE. It is not surprisin vendorso g tw tha e tth ' design similae sar shard an r e similar failure mechanisms (IASCd Can crevice corrosion cracking). Based on absorber tube cracking observed in the original equipment blades in early cycles in several German BWR plants Siemens characterized the cracking phenomenon as a function of exposure (see Table II) and developed the following strategy for shuffling and replacement:

• original equipment blades can occupy high duty control cell positions to an exposure of < 1.8 snvt projecte witL hdEO subsequent discharg movemenr eo cora o t e edge position; exposure th f i >1.s ei 4• snv t beginnina t cyclf go e (BOC) placee y blade b an th ,n n di eca shutdown positiond an ; • if the exposure is >1.6 snvt at BOC, the blade can only be placed in a low duty shutdown position (not near core center).

TABL . SIEMENEII S ORIGINAL EQUIPMENT BLADE LIMITS

Local B- 10 % Nuclear Exposure, Characteristic Depletion% , Lifetime snvt Begin cracking 30 37 -1.0 Begin TABLE losI s 40 49 -1.3 10% S/D margin reduction, 54 66 -1.8 B4C + washout 1 0% S/D margin reduction, 82 100 -2.7 B4C without washout

Original equipment blade e stilar s l used quite extensivel German i y n reactors e currenTh . t advanced design offeree Hybrid-th s i d 4 blade (since -1988) using high purity stainless steel, redesigned welds to minimize crevice corrosion, and hafnium at high duty blade tip and edge locations currene Th . t limit givee sar Tabln ni (-BWR/I eII 6 plant design):

TABLE III. CURRENT BLADE LIMITS

Blade Design Nuclear, snv) (% t Mechanical, snv) (% t Original equipment 3.2(45) -2.1 (-30) Hybrid-4 5.3 (62) >4.0 (>47)

234 e %10Th B depletion value e peaar s k segment average values. Operating experience supports blade exposure morf so e tha snv7 ncracks 3. a t s hav t beeeno n observe t thesda e exposure levels.

2. UTILITY BLADE REPLACEMENT STRATEGIES

Data has been analyzed from a total of 23 utilities, operating 45 BWRs in the U.S., Europe, and r EasttheFa , concerning their blade replacement strategie operationad san l limits responsee Th . e sar believed to be representative of the industry. The geographical breakdown of the utility companies consist: sof

U.S9 . utilitie • plants)3 s(1 ; Europea0 1 n • utilitie plants)9 s(1 d an ; • 4 Far Eastern utilities (12 plants).

As summarized in Table IV, these represent BWR/2 to BWR/6 plant designs and GE, Siemens, and ABB-Atom NSSSs operated in control cell, conventional, and mono-sequence control cell modes. The table shows that the utilities have implemented a full range of blade designs, from all-B4C to hybrid B4C/hafniu all-hafniumo t m absorber materials.

2.1. Control Blade Shuffling Strategies

Frequenc Bladf yo e Shuffling Table V summarizes blade shuffling activities performed by the responding utilities. More than one-half (58%) shuffle blades. Most obviously, this is to extend blade in-core residence times. Blades are generally moved from high duty positions (i.e., control cells in CCC cores) to shutdown positions where they remain withdrawn during operation. In such positions, the incremental exposure is negligible. Additionally, as noted by one utility, shuffling allows the high exposure but low cobalt blade replaco st e original equipment designs incorporating high cobalt materials.

Slightly less than one-half of the utilities do not shuffle blades, choosing instead to leave the blades in the same position from one cycle to the next. These utilities note that the life extension benefi marginals i t , especiall increasee lighn yth i f o t d cost shufflo st e blades whic onln donhe ca y b e during a refuelling outage. One domestic utility estimates the cost to be as much as $10,000 per outage hour wit takint hourhi 8 o t shuffl o sgt 4 paia e bladesf o r . With utilities movin shorteo gt d ran shorter outages, there is more pressure not to shuffle blades. Over the last 18 months in the U.S. there has been a 117% increase in the number of plants with outages of 40 days or less and a 93% increase in the number of plants with 50 day outages or less.

Shuffling Strategies Vendors have a minor influence on the strategies implemented by utilities especially as the original equipment blades with their short mechanical lifetime e replacear s y hybrib d d blades incorporating advanced material d absorbersan s . Siemens r examplefo , , developed guidelinen o s where a blade could be placed based on its exposure level (see Table II) while GE proceduralized originae th r fo l9 equipmen57 L SI thei Duralifd d an tan r 7 guidelinee15 120/14L SI n si 0 series blades, respectively.

The strategies with the various designs are, of course, to move high burnup blades to low exposure region vicd an se versa e actuaTh . l strategy depends primaril design o y n featureE G C s original equipment blades r examplefo , e stilar , l swellinC pron4 B o e t t bladga positionp eti s even when fully withdrawn from the active core region in shutdown banks. This necessitates the use of 'phantom' exposure subsequenn si t cycles.

Figur showe4 exampln sa e strategf cyclo th f d eeo en (EOCy e use2 th 1 Utility r )db fo yD refueling outag t Uni e. a Fou Dl t r different type movef so s were planned specifie th , c move depending on blade type and exposure:

235 to Tabl . SUMMAReIV PLANR BW TP Y O FEATURE S

NSSS Plant #Fuel # Control Blade Types Operating Utility Plant Vendor Type Ass'ys Blades Currently Used Mode USA A Al GE BWR/4 764 185 OE, D140, D215 CCC B Bl GE BWR/6 800 193 OE CONV B2 GE BWR/6 624 145 OE, CR82M CONV C Cl GE BWR/2 560 137 , D160OE , D190, D230M , CCC D Dl GE BWR/4 560 137 OE, D190, D230, M, CR82M CCC E El GE BWR/2 532 129 OE, D190,D230 CONV E2 GE BWR/5 764 185 , D190OE , D215M , -CCC F Fl GE BWR/3 580 145 D140, CR82, CR82M,M CCC G Gl GE BWR/4 560 137 OE.D 190,0230 CONV G2 GE BWR/4 560 137 OE, D190, D230, CR82 CONV H HI GE BWR/4 764 185 OE, D160 CCC H2 GE BWR/4 764 185 OE CCC I 11 GE BWR/5 764 185 OE, D215 CCC Europe J Jl KWU N/A 592 145 OE, CR70, CR82, CR85 MSCC K Kl GE BWR/4 240 57 D23 other0+ s CCC L LI GE BWR/6 648 149 OE, D230, CR85 CCC M Ml KWU N/A 532 129 OE, CR82, CR82M, CR85, M MSCC M2 KWU N/A 840 205 , CR82OE , CR82M, CR85M , MSCC N Nl GE BWR/6 624 145 OE CCC O O1,O2 KWU N/A 784 193 OE, Duralife series, M CONV P PI KWU N/A 592 145 OE, CR82, M CCC Q Q1.Q2 ABB N/A 444 109 CR70, CR82, CR85 MSCC 5 Q , Q4 , Q3 ABB N/A 444-700 109-169 CR70, CR82, CR82MM , MSCC R R1,R2 ABB N/A 500 121 CR70, CR82, CR82M, SR88 MSCC S S1.S23 S , ABB N/A 676-700 161-169 CR70, CR82, CR82MM , MSCC r EasFa t T T1,T2,T3,T4 GE BWR/4, BWR/5 N/A N/A OE, ail-Hf CCC U U1,U2 ~GE BWR/2, BWR/5 N/A . N/A , CR85OE , CR85M1 CCC V VI, V2 GE BWR/4 408 97 OE.D215 CONV V3, V4 GE BWR/6 624 145 OE, D230 CONV w W1.W2 ~GE BWR/4 N/A N/A OE, all-Hf CCC OE = Original equipment 0140 = Duralife 140 016 Duralif= 0 0 16 e D190 = Duralife 190 0215 = Duralife 215 D230 = Duralife 230 M = Marathon CR70 = CR-70 CR8 CR-82= 2 CR82M = CR-82M CR8 CR-85= 5 CR85M1=CR-85M- contro= 1C lCC cell core CON Vconventiona= l mode MSC Cmono-sequenc= e control cell Table V. CONTROL BLADE SHUFFLING - UTILITY RESPONSES

Shuffle Utility Blades? Comments USA A Y Shuffle to extend blade lifetimes B N Increases outage costs C Y Shuffle to extend blade lifetimes; high duty blades moved to shutdown positions D Y Shuffle to extend lifetimes & to replace high cobalt original equipment blades E Y Limited shuffling performed; extends lifetimes; blades moved from high duty positiono st peripheral & core edge positions F N Increases outage costs G N Limited shuffling in past; used as a last resort H N Increases outage costs I Y Shuffl exteno et d blade lifetimes; high exposure blades move peripherao dt l positions, peripheral blades moved to high duty positions Europe J Y Shuffl exteno et d blade lifetimes; high exposure blades moved to peripheral shutdown positions, shutdown blade moved to high duty positions K Y Shuffle to extend blade lifetimes; high exposure blades moved to peripheral shutdown positions, shutdown blades discharged L N Life extensio minimals ni ; increases outage costs M Y Shuffle to extend blade lifetimes; high exposure blades movedutw lo y o shutdowdt n positions N Y Shuffle to extend blade lifetimes O N Increases outage costs P N Increases outage costs Q Y Limited shufflin r lifgfo e extension R Y Shuffle to extend blade lifetimes S Y Shuffl exteno et d blade lifetimes; typically move blades from control cells to shutdown positions Far East T N Life extension is minimal; replace with all-Hf designs U N Life extension is minimal V Y Shuffl exteno et d blade lifetimes w N Life extension is minimal; replace with all-Hf designs

• Move A - selected original equipment blades were moved to interior shutdown positions after having occupied shutdown positions nea core rth e peripher eighr yfo t cycles;

• Mov - selecte eB d Marathon reside0 s1 inserte C shutdown di EO t da n positiono tw r sfo cycle werd san e then moved into control cells;

• Mov - selecte eC d Duralife 230s resideinserte8 C contron i ddEO l cell positionr fo s four cycles and were then moved to shutdown positions; and

• Mov - selecte eD d Duralife 230s insertereside9 C shutdown di dEO n positions neae rth core peripher threr yfo e cycle werd san e then move controo dt l cell positions.

Examples typica thesf o l e move showe sar Figurn ni . e4

237 10 14 18 22 26

Blad- e Typ eRefuelin( g Outage Loaded) - Core Position Shuffled From - Refueling Outage Shuffled

- Control Cell

Blade Type Description OE Original Equipment D190 Duraiife190 D230 Duralife 230 M Marathon 82M CR82M

FIG 4 Planned blade shuffling for unit Dl at EOC12

2.2. Control blade replacement strategies

Replacement Designs Table VI provides a summary of the replacement blade designs used by the responding utilities e ancalculationath d l methods use o evaluatt d e exposure. Quite apparen e fulth l s i rangt f o e replacement designs, from original equipment blade n threi s e utilitie o all-hafniut s m designn i s severa r EasterFa l n plants e majoritTh . f utilitieo y s (87%) replace discharge blades with advanced designs incorporating improved structural t leassteela d t shafniuan m tips.

e U.S.I th nmajorite th e ,utilitie th f e o usiny 0 seriear s 20 g d sDuralif bladean 0 19 es with hafnium tips and edges. Less common are Marathon designs and ABB-Atom's CR-82 and CR-82M designs. In Europe, most replacement blades are ABB-Atom's CR-82, CR-82M, and CR-85 designs

238 with GE providing a relatively small number of Duralife and Marathon blades. In the Far East, replacement designs incorporate either hafnium tips and edges (Duralife 200 series, CR-85, CR-85M- 1) or are all-hafnium designs.

TABL CONTRO. EVI L BLADE REPLACEMENT UTILIT- Y RESPONSES

Exposure Adderp Ti s Replacement Utility Calculations Used? Designs USA A D140, D215 B PowerPlex, Y OE, CR82M 3D MONICORE C RODEX Y D160,D190,D230,M D 3D MONICORE N D190, D230, M, CR82M E 3D MONICORE N OE 190,02D , , 1D2305 M , F —— Y D140, M, CR82, CR82M G 3D MONICORE N D160, D190, D230, CR82 H 3D MONICORE Y , D16OE 0 I PowerPlex N D215 Europe J CR82, CR85 K ..... D230 L Core Master/Presto Y D230, CR85 M —— N CR82, CR82M, CR85, M N OE O Duralife, M P CR82, M Q —— CR82, CR82M, CR85, M R POLCA4 Y CR82, CR82M S CR82, CR82M, M Far East T All-Hf U 3D MONICORE N CR85, CR85M1 V PowerPlex, —— D215, D230 3D MONICORE w —— N All-Hf

As noted above, three utilities replace discharge blades with original equipment blade- s Utilities E, H, and N. In these cases, the utilities have purchased (or received) the blades from canceled plants and use them as inventory for replacements. High cobalt pins and rollers are replaced in all cases to reduce primary system activation.

As these blades incorporate all-BC absorber, commercial purity steels, relatively high cobalt

components d geometriean , s that promote coolan4 t stagnation, thee pronar y o absorbet e r tube cracking. Consequently, lifetime restrictee sar mechanicay db l limits (exposures <34 depletion0 1 %B ) requiring that more blades be replaced over the life of the plant. This requires more handling of blades and the use of SIL 157 restrictions in shuffling. Additionally, the higher cobalt content relative to modern replacement designs increases primary system radiation fields.

In spite of these disadvantages, economic analyses tend to favor the use of the original equipment blades simplo theit e r du yextremel pricew lo yorden I . o characterizt r e economieth c advantage of original equipment vs. advanced blades, we performed an analysis based on the following assumptions:

• the cost of the original equipment blades is $5,000 (replacement of high cobalt pins and rollers);

239 the purchase price of a new advanced design blade is $75,000; nuclea depletioB 10 r lifetime% 34 n e (peaar s k segment e originath r fo ) l equipment advancee th r bladesfo d% blades54 , ; disposal cost $25.00e sar blader 0pe ; outage costs are $10,000 per hour; control cell core operation with typical blade exposure increments per cycle; allowanco n r enhanceefo d primary system radiation higheo t field e du sr cobalt blade materials; no blade shuffling; and no escalation, future worth, etc. corrections.

Result e providear s n Figurei d s 5 and 6 through fifteen cycles of operation under these conditions.

Figur showe5 resultine sth g blade replacement schedule woule on s d A .expect , more original equipment blade needee sar simple ddu theio yt r shorter lifetimes origina4 tota14 A . f o l l bladee sar advance3 neede10 . dvs d design blade increasn a s- 30%f eo . Figur providee6 cumulative sth e cosf to using each blade typ . eEve throug15 n C wit largee hEO hth r numbe f bladeo r s neede highed dan r costs due to having to purchase and dispose of more blades, the economic advantage of the original equipment blades is substantial, with the cumulative EOC 15 cost being approximately 70% of that of the advanced blades. When averaged over the fifteen cycles, the cost advantage is about $300,000 per cycle.

Advanced (54%) TSJ Orig Equip (34%)

0 •o JS m

CD .a E 3

5 1 4 1 3 1 2 1 1 1 0 1 9 8

f Cyclo d eEn

FIG Impact5. of blade type replacementon schedule

Replacement Criteria - Utilities Using ABB-Atom Blades The early blade designs incorporating 304SS for structural materials have lifetimes limited by IASCC . e CR-70Thesth e ar e, CR-82 CR-8d an , 5 series e exposureTh . e advance th dato t sf o e d designs are not sufficient to determine if they are also limited by absorber integrity concerns.

240 1.6E+07

1.4E+07 Advanced (54%) Orig Equip (34%) 1.2E+07 •stO> oo 1.0E+07 0) 8.0E+06 u E 6.0E+06 O 4.0E+06

2.0E+06

O.OE+00 5 1 4 1 3 1 2 1 1 1 0 1 9 8

End of Cycle

FIG. 6. Cumulative costs as a function of cycle completed

U.S. utilities have adopted ABB-Atom's guideline criterid an s a relate blado dt e replacement. As such, lead blades are inspected and discharged if cracks are observed.

European utilities, on the other hand, have adopted criteria that are, at times, more restrictive than those suggeste ABB-Atomy db exampler Fo . , Utilit implementes ha yJ - d CR snvlimit9 r 1. tfo f so 70s and a modest 3.0 snvt for the hybrid CR-82s and CR-85s. These are based on their own experience crackinf so Unin gi . tJl

e majoritTh f otheyo r European utilities visually inspect blade frequena n o s t basi asseso t s s cracking. A typical example is Utility S which operates several units in Sweden. The utility has developed exposure-related criteria for their units based on their experience with cracking in the CR- 0 contro7 l blades. These criteria, noted below, apply specificall e CR-7th o t y0e ar blade d an s recommende CR-82e th r dfo , CR-82M Marathod ,an n designs:

• blades wit p exposureti h s exceedindepletioB 10 % n57 g shal e inspecteb l d before continued irradiation; • blades with maximum quarter segment exposures above 18% 10B depletion shall be inspected before continued service; and • blades schedule contror dfo l positions with quarter segment exposure exces% n si 45 f so 10B depletion shall be inspected before insertion in those positions (these particular blades are also inspected at EOC after their use in the control positions).

An additional criteria is applied to blades exposed in Units SI and S2. Blades that are free of cracks with 1) tip exposures between 57% and 60% 10B depletion, and 2) having a maximum quarter segment exposure below 18% '°B depletion are allowed to continue operation for a maximum of two years in a shutdown position. These blades are, however, considered to be cracked during their second yea f exposurro e with penalties applie shutdowe th n di n margin evaluations.

241 Although most European utilities discharge blades when cracked, Utilit allowS y crackea s d blade to be reinserted for an additional cycle only if it is moved to a shutdown position.

Replacement Criteria - Utilities Using GE Blades With very few exceptions, vendor criteria and guidelines are implemented by U.S., European, and Far Eastern utilities. Inspections are not performed due to the presence of the sheath covering the absorber tubes. Blades are discharged when exposure limits are approached.

Two utilities have discharged blades well before their lifetimes. At EOC 2, Utility A discharged 165 original equipment blades and replaced them with Duralife 140s to reduce cobalt activation and, consequently, primary system radiation fields. Utility L, at EOC 3 in Unit LI, replaced 16 original equipment blades with Duralife 230 CR-85d san orden si quicklo rt y gain experience with hybrid designs.

Replacement Criteria - Utilities Using Siemens Blades U.S. utilities respondin e surveth o gt y hav experienco n e e with Siemens blades. Limitn o s original equipment blades use Europeay db n utilities var ysnv7 fro snv2. t Unit6 a t m o t 1. Unit a t I sP t Ml and M2. Referring to Table II, these limits approximate the 10% reduction in blade worth due to depletioC 4 B n plus washout (~1.8 snvt reductio% , BWR/2-510 e bladn th ni d ean ) wort onle hdu o yt B4C depletion (-2.7 snvt, BWR/2-5).

Replacement Criteri a- Utilitie s Using All-Hafnium Blades All-hafnium blades are used in several Far Eastern plants. The utilities implement the vendor

criteri guidelinesd aan . Utilit limityT s original equipmen designsC t4 (i.e.B l al ,<1.7o )t 4o t snve du t cracking. All-hafnium blades are limited to <9.0 snvt at the top axial segment.

Use of Tip Adders Nearly one-half of the responding utilities use tip adders to account for B4C swelling and absorber tube sensitization occurrinregiop ti blada e f no th en gi whe n locate shutdowa n di n position (i.e., fully withdrawn during operation). ABB-Atom and GE recommend that tip adders be used for all-B4C blade designs. Tip adders are not needed for designs incorporating hafnium tips as hafnium experiences little swelling and growth during irradiation. With the exception of one utility, all responders still have original equipment blades under irradiation. Tip adders are typically on the order depletioB IO % cycler 3 U.Sne pe o t 1 On . . utilitf o depletiop y ti use % s 2 r 10,00npe 0 MWd/st.

Reasons given for not using tip adders include:

• blades located on the core periphery receive effectively zero exposure; • with operation in a conventional (i.e., sequence exchange) mode, all non-peripheral blades receive some exposure during the cycle and that exposure can be estimated and tracked; • vendors do not require the use of tip adders; • correlations for tip exposure are poorly benchmarked; and • no trend has been observed in measured shutdown margin to indicate absorber loss.

Handle Embrittlement Issues Although ABB-Atom and GE indicate that handle embrittlement is not a threat to blade integrity, several utilities have directly addressed this issue , dischargeQ2 . d Utilit Unitn an i , l sysQ Q blades when they have reache a fluencd f ~10o e 22 n/cm2 base n calculationao d l methods. This corresponds to about 20 years residence time. Hot cell examinations have been performed on blades from Units Q3, Q4, and/or Q5 and from these some correlations were developed based on fast fluence utilite Th . y replaces blade cycles5 2 s o afteoperationf t o 0 1 r .

242 Blade Exposure Calculational Methods Utilities use vendor correlations in calculating blade exposures. The methods consist of evaluating the exposures of the fuel assemblies adjacent to the blade of interest and converting the smeared thermal fluenc 0 depletion1 n B snvt(i eo t ) . Most utilities D witNSSSE 3 G he us s MONICORE from GE, or PowerPlex from Siemens, to estimate and trend blade exposures. Utilities with ABB NSSSs perform the exposure calculations using similar methods (i.e., COREMASER/PRESTO). In some instances, ABB-Atom performs the exposure evaluations for the utility.

The correlation between fluence in snvt and percent 10B depletion is dependent on blade design (see, for example, Figure 3) and on the following factors:

• averag eadjacene voith n di t assemblies; • design of the adjacent assemblies; and • relationship between thermal fluence and core power.

e mosTh t sophisticated method o datt e scorrelation us e e aforementioneth r fo s d factors, evaluating exposure e fouth r n i saxia l blade segment bladd an s(typicallp eti axiam m in.y6 l2 ,15 region).

3. CONCLUSIONS

On Shuffling

1) A total of 58% of the utilities shuffle blades at least on a limited basis, most often to extend lifetimes. Blades are generally moved from high duty positions (control cells, for example o lowet ) r duty shutdown positions e utilitOn . y shuffles preferentiall o positiont y s occupied by original equipment blades to accelerate the removal of those blades containing high cobalt materials;

2) Slightly less than one-half of the utilities do not shuffle blades. Although blade lifetimes can be increased by moving them to low duty positions, shuffling increases outage costs through additional manpower and time requirements.

On Replacement Blade Designs

1) Utilities replace discharged blades with ones having a wide range of design features, from original equipment blades (simila thoso t r e the discharginge yar t threa ) e utilitie allo t s - hafnium design Easterr Fa so nusetw utilitiest da . However majorite th utilitie,e th f yo s insert designs that employ at least crack-resistant structural materials and hafnium tips;

) 2 Ther a strone e tendb go t scorrelatio n between NSSS vendo replacemend an r t blade supplier. Utilities with GE NSSSs (or NSSSs based on GE's designs) tend to load Duralife and/or Marathon designs whil NSSSB e utilitieAB sd mosan s U twit oftehKW n replace with ABB's designs (Siemens designs are essentially no longer available). However, it is quite apparent that both vendor activele sar y pursuin replacemene gth t blade marke havd an t e been successfu loadinn i l g their designs int othee oth r vendor's NSSSs;

) 3 Three utilities hav a elarg e inventor f originao y l equipment blades available from canceled plants tha usee ar treplacementss da .

243 On Blade Replacement Strategies

1) Utilities Using ABB-Atom Blade Designs

• ABB-Atom recommends that utilities using 304SS blade designs (CR-70, CR-82, CR-85) begin inspecting blade peaa t sa k segment exposur snvt3 f eo ;

specifio N c • mechanical limits have been establishe r bladedfo s using 316L wings i.e., CR-82M, CR-82-M-1, CR-85M, CR-85M-1. The strategy followed worldwide is that the lead blades are visually inspected each year. In this way, the observation of cracking will establis mechanicae th f hi l lifetime more sar e limiting tha nucleae nth r lifetimes.

2) Utilities Using GE Blades

• Witexceptionsw hfe , vendor limit e implementear s utilitiesy db . Original equipment, Duralife 120, and Duralife 140 blades have a mechanical lifetime equivalent to 34% 10B depletion due to absorber tube cracking and crevice corrosion. The Duralife 160 through 230 series, incorporating better materials, hafnium, and higher boron carbide loadings (Duralif230)d , an 5 hav21 e e resulte incrementan a n di l improvemen lifetimen i t a o t s maximu abouf mo t three time originae s thath f to l equipment blade;

• Marathon design limitee sar d mechanicall consequenca s ya internaf eo l pressurizatiot na high exposure heliuo t e msdu generation fro e (neutronmth , alpha) reaction with 10B. correspondine Th g exposure limi abous i t snvt0 5. t ;

3) Utilities Using Siemens Blades

• Designs offere susceptible Siemenar y db o s same , similae th sear o eGE t degradatioo rt n phenomena. Althoug nucleae hth r limi originar fo t l equipment blade BWR/2-n si 5 series plant snv7 wortd 2. ro t s si (10hdepletio %C 4 reductioB o t e n ndu only) , cracking begins snv0 1. t t whila losC 4 initiates sei B snvt3 1. t ;da

• Exposure limits employe utilitiey db s using original equipment blades range fro6 m1. snvt to 2.7 snvt; e Hybrid-Th 4• design, incorporating hafnium, improved material d desigan s n modifications, has a mechanical lifetime limit of >4.0 snvt. Lead blades have reached exposure snv7 3. tf withouso t indication crackingf so .

On Tip Adders

• Nearly one-half of the responding utilities use tip adders to explicitly account for the exposure in the tip region of shutdown blades. Tip adders, used only for original equipment designs, range from 1 to 3% 10B depletion per cycle;

• Reasons for not using tip adders are varied, but most common are 1) the vendor does not require them to be used, and 2) no trend is observed from BOC shutdown margin measurements to suggest absorber loss.

244 HIGH TEMPERATURE STUDY OF THE CONTROL ROD XA0053645 BEHAVIOUR UNDER ACCIDENT CONDITIONS

V. TROYANOV, A. POMESCfflKOV, V. SOUGONJAEV Institut Physicf eo Powed san r Engineering, Obninsk

V. PONOMARENKO, A. SCHEGLOV Moscow Polimetal Plant, Moscow

Russian Federation

Abstract

Control rod behaviour under accident conditions is considered. A comparative assessment of various types of absorbers is carried out in this respect. It was found that the best absorbers are dysprosium based absorbers. Dysprosium based control rods show no failure at temperatures below 1300°C. Boron based control rods may be liquefied at a temperature of 1150°C. Besides, the cladding may burst at lower temperature, due to the released helium. Experiments indicate thaabsorbee th t r material initiaty sma e severe core damage under severe accidents beginnine Th . corf go e structure degradation is caused by the boron control rod liquefaction.

1. THE MAIN CORE DEGRADATION PROCESSES UNDER UNCONTROLLED TEMPERATURE INCREASE

The temperature range of core material liquefaction for WWER is between 650°C (melting poin f burnablo t e absorber base aluminiumn do , MPA 2700°d an ) C (melting poin f uraniuo t m oxide and zirconium). Thi dividee s b rang y dema conditionall temperaturw lo e th o yt e range (not more 1500°C) and high temperature range (up to 2700°C). Let us consider the stainless steel melting point conditionae asth l border.

stude Inth y considere temperaturw lo e dth e phenomena have been mostly analyzed. Jus thin i t s temperature range, the core degradation processes occur under severe accident conditions. These processe e fued ballooninar sro l failured gan , contro d failur ro ld liquefactionan e , liquid eutectic formation, chemical interaction of differential types of materials, melt flow down, etc. Several schemes of core structure degradation are presented in chapter 2. These schemes are based on the experimental study of the phenomena analyzed.

2. EXPERIMENTAL STUDY OF THE CORE STRUCTURE DEGRADATION, ABSORBER CONCERN

2.1. Methodological approach

Contact interactions as well metallurgical liquefaction have been studied. Mock-ups of spacer grids, fuel rods, guide tubes have been used. Standard fuel, absorber and cladding materials have been applie r mock-ufo d p fabrication gammA . a facilit d furnacean y e use r testingar s fo d e gammTh . a facilit higa s yi h temperature steam loop with tes d steaan t r temperaturemba pressur2 - 1 o t f ep o su 1900°C. The temperature rise is approximately 1.5°C/sec. Mock-ups have been surveyed after testing and cutting and are examined metallographically by micro X-ray phase analysis. Mechanical material tensile testin bees gha n carrie wells a t .dou

2.2. Fuel cladding/spacer grid interaction

A comparative assessment of the fuel cladding (FC) and spacer grid (SG) interactions in vacuum and in a steam environment has been done (see Table I).

245 TABL E. FUEI L ASSEMBLY DEGRADATION; INTERACTIONF SO THE FUEL RODS ( FR) AND SPACER GRIDS ( SG)

Temperature range, approximately Phenomena ______^C______600-800______Fuel cladding ballooning______800-900 Fuel cladding failure Fissio releass nga e ______Beginning of internal cladding oxidizing 1200 and higher Cladding oxidizing acceleration ______Hydrogen release______1450 Melting of spacer grids Melt flow down ______Blockage formation______1900______Cladding melting______Remar eutectio n k c formation observee sar stainleso t p du s steel melting FR cladding - Zr,l%Nb, Fuel - U02, SG - stainless steel

visiblThero n e ear e interaction testinn s aftemi 980°t 5 ga 1 r vacuumCn i . Metallography shows a 15-60 micrometer interaction layer. After 15 min testing at 1030°C in vacuum, the interaction zone is increase 25-36o t p du 0 micrometer fac e spitn si th t f thae o visibl o n t e interaction observede sar t A . 1100°C and higher temperatures, significant interaction and liquefaction is observed. It was expected that liquefaction would happen approximately at 950°C, since the liquid eutectic temperature would be exceeded. Nevertheless t happenno d initiae di th ,t i , l oxid zirconiue e th fil n mo m cladding prevents it from liquefaction.

In the steam environment, the oxidation of zirconium is higher. A thick zirconium layer prevent chemicae sth l interaction with 1500°Co steet p u l . This phenomeno s illustrateni Figsn di 1 . and 2.

2.3. Control rod cladding/absorber interaction

2.3.7. CRC/boron carbide

Pressurized mock-ups have been studied at a temperature range of 800-1200°C (see Table II).

TABLE II. FUEL ASSEMBLY DEGRADATION; INTERACTIONS OF BORON CARBID CONTROD EAN CLADDIND LRO G GUIDE TUBE

Temperature range, approximately, Phenomena ______^C______65 higher______Diffusiod 0an intcladdine B od th an C g f o n highed an 0 r 80 Contro Claddind Ro l g ballooning He released from B4C ______Stee interaction______C 4 B - l _ 1080 Ni/B eutectic formation 1174 Fe/B eutectic formation 1150 and higher CR cladding liquefaction Guide tube liquefaction Melt flow down, blockage formation He released ______H2 B20, 3CO ,released______Remark pressurized control rod cladding is not failed by the guide tube protection and guide tube strength Contro claddind ro l gstainles- s steel, Guide tub estainles- s steel

246 VACUUM STEAM 980°C 1030°C 1100°C 1300°C 1420°C

1600°C 1700°0/--C1 1800°C 1900°C

/•/(/ / Space?' grid/fuel toil cladding in/den lion FIG. 2. Fuel assembly mock-up

248 significano Thern e ear t chemical interaction temperature th n i s e range between 800-1150°C. explainee b Thie shorn th ca s y t b dperio f temperaturo d e increas e temperaturth t a e e rise ratf o e 1.5°C/sec. The beginning of liquefaction is observed at temperatures above 1150°C. Penetration of carbon and boron into the austenitic through diffusion is observed. Complex carbides type MeyCs and borides type Me s foundBi . Fatal liquefaction occurs above 1200°C. e mock-uViewth f o se ar p

shown in Figs. 3-62 , before and after testing at 1200°C and 1350°C.

2.3.2. CRC/dysprosium titanate

Non-pressurized mock-ups were teste temperaturet da 1800°o t p su C (see Table III).

TABLE III. FUEL ASSEMBLY DEGRADATION; INTERACTIONS OF THE DYSPROSIUM TITANATE AND CONTROL ROD CLADDING GUIDE TUBE

Temperature range, approximately, Phenomena ______^C______.______1200 and higher______Guide tube oxidizing acceleration______1450 Melting of the CR cladding Melt flow down within guide tubes ______No Di2TiOs relocation due to its densification ______1900______Guide tube material melting and flow down Control rod cladding - NiCr alloy; Guide tube - Zr,l%Nb.

No significant chemical interactio s beeha n n observe o 1400-1450°Ct p u d e viewTh .f o s mock-ups tested at 1200°C and 1400°C are shown in Fig. 7 and at 1800°C in Figs. 8-10.

2.4. Burnable absorber rod behaviour

The absorber alloy MPA is based on an aluminium/boron mixture. MPA is molten at temperatures higher than 650°C (see Table IV). The specific volume of MPA increases when melting be mely Th abou t. interact7% t s wit e zirconiuhth m cladding. Intermetallic compound e formedar s , but the solubility of Al in Zr is not significant.

TABLE IV. FUEL ASSEMBLY DEGRADATION; BURNABLE ABSORBE BEHAVIOUD RRO R

Temperature range, approximately, Phenomena ______^C______650______MPA melting______650 and higher______Cladding - MPA melt interaction___ 70highed 0an r Intermetallic compounds Zr/Al protect from further chemical interaction ______Cladding oxidizing______800 and higher Cladding failure for pressurized BAR Molten absorber extraction and flow down ______No failure for unpressurized BAR Burnable absorbe raluminiu- m based boron alloy, MPA; Contro claddind ro l gZr,l%Nb- . BAR burnabl- e absorbed rro

e intermetalliTh c laye s formee innei r th n r o d surfac e claddingth f o e . Thia barrie s i sr fo r further aluminium penetration into the cladding. The reaction rate depends on the diffusion through this layer fatae Th .l liquefactio f zirconiuno m cladding doe t occurcladdin R sno failure BA Th . f o eg takes plac r pressurizefo e d BARs only r faileFo . d s pressurBAKsga e th ,meleA pushetMP e th s through the hole and the melt flows down.

249 ini

FIG. 3 Fuel assembly mock-up before testing

250 A

c

FIG 4 Boron contained mock-up after testing at 1200 °C

251 . < i sit.•• * g "l

FIG. 5. Boron contained mock-up after testing at 1350°C

252 C FIG. 6 Boron contained mock-up after testing t 1350a C 253 sprosmmy D 7 G containedFI mock-up aftei / >C ° ighfytesting 1400d leftjC an ° t 1200a

254 FIG. Dysprosium8. contained mock-ups after testing 1800at °C

255 FIG. 9. Dysprosium contained mock-ups after testing at 1800 °C

256 F/G 10 Dysprosium control rod after testing at 1800 T1

257 SUMMARY

1. The liquefaction of boron absorber rods may happen between 1150-1200°C. It leads to liquid eutectic formations, flow dow f melblockagnd o an t e formations mainly neae th r spacer grids.

2. Control rod liquefaction initiates core degradation in the melt formations observed. The temperature of absorber liquefaction is about 200°C less than fuel rod liquefaction.

3. Absorber relocation within the core may happen while the core structure is not destroyed.

4. Dysprosium absorbers are much more firm. Dysprosium absorber rods are not destroyed up to 1450°C. Only the bottom part of the absorber rods of the new Russian control rod desig mads ni dysprosiuf eo m absorber. supposes i t I 5. d that under severe accident conditions controe th , l rod insertee sar d inte oth core. The most heated part of the core is the central cross-section. Thus, the unirradiated absorber is heated more.

Therefore, previous irradiation of absorbers before carrying out tests is not necessary for the above discussed experiments.

258 LIST OF PARTICIPANTS

Aden, V.G. Ministry of Russian Federation for Atomic Energy, Moscow, Russian Federation

Afanasyev. A , Ukraine State Department of Nuclear Power, Kiev, Ukraine

Andersson, T. RTH, Varobacka, Sweden

Bart, G. Institut Paul Scherrer, Villigen PSI, Switzerland

Bechade, J.-L. CEA/DTA/DECM/SRMA, Gif-sur-Yvette, France

Bjornkist, L. Vattenfall AB, Varobacka, Sweden

Date, V.G. Bhabha Atomic Research Centre, Trombay, Mumbai, India

Dewes, P. Siemen KWU, sAG , NBTW, Erlangen, Germany

ElSayed. A , Atomic Energy Authority, Mugls El Shab-Cairo, Egypt

Harbottle, J. Stoller Nuclear Fuel, Thornbury, Bristol, United Kingdom

Herter, P. CEA, Centre d'Etudes de Saclay, Gif-sur-Yvette, France

Hinttala. J , STUK, Radiatio Nuclead nan r Safety Authority, Olkiluoto, Finland

Hoglund, L.M. Barseback Kraf, AB t Loddekopinge, Sweden

Jendrich. U , Gesellschaf Anlagen-unr fu t d Reaktorsicherheit, (GRS)mbH, Cologne, Germany

Jonsson. T , Studsvik Nuclea, rAB Nykoping, Sweden

Kawashima. ,N Hitachi Engineering Co., Ltd, Ibaraki-ken, Japan

Kennard, M.W. Stoller Nuclear FuelInternationalC NA , , New York, United States of America

Monchanin, M. FRAMATOME, Lyon, France

259 Muttilainen, E.A.T. Teollisuuden Voima Oy, Olkiluoto, Finland

Odeychuk, N. Kharkov Institute of Physics and Technology, Kharkov, Ukraine

Poignet, B. DRN/DER/SPRC, Saint-Paul-lez-Durances, France

Posselsky, V.B. Ministry of Russian Federation for Atomic Energy, Moscow, Russian Federation

Rebensdorff. B , ABB-ATOM/BU, Vasteras, Sweden

Risovany, V.D. Research Institute of Atomic Reactors, Dimitrovgrad, Ulyanovsk Region, Russian Federation

Ritchie. I , International Atomic Energy Agency, Vienna

Sanovic, J. Institute for Metal Structures, Ljubljana, Slovenia

Scheglov. A , Moscow Polymetal Plant, Moscow, Russian Federation

Schulz, W. Preussen Elektra Aktiengesellschaft, Hannover, Germany

Shoaib, K.A. Permanent Mission of Pakistan to the IAEA, Vienna, Austria

Simeone, D. CEA, Centre d'Etudes de Saclay, Gif-sur-Yvette, France

Troyanov, V. Russian Federation of Atomic Energy Ministry, Obninsk, Kaluga Region, Russian Federation

Zakharov. A , Research Institut Atomif eo c Reactors, Dimitrovgrad, Ulyanovsk Region, Russian Federation

260