fuel handling systems - iaea

INIS-CA—0052

CANADIAN NUCLEAR SOCIETY Thef u e l

changingoperation is

based on thecombined use of two

remotely controlled fuellingmachines operatir ' — ' '' ~x ~'channel.New fuel!machine, are inse

stCA9700717

ProceedingsMAY 13 & 14, 1996HOTEL PLAZA IITORONTO, CANADA

direction oi The loaaingdepends upon thedirection of the coolantflow in the fuel channelbeing fuelled, whichalternates from channelto channel. The fuellingmachines receive newfuel while connected tothe new fuel port anddischarge irradiated fuelwhile connected to thedischarge port. Theentire operation isdirected from the controlroom through thep r e p r o g r a m m e dcomputerized system.The control systemprovides a printed log ofall operations andpermits the manualintervention by theoperator. New fuel isreceived in the new fuelstorage room in theservice building. Thisstorage room canaccommodate up to sixmonths' fuel inventoryand can store temporarily

all the fuel required for theinitial fuel loading. When

required, the fuel bundles are transferredto the new fuel transfer room in the reactor building. The fuelbundles are identified and loaded manualiy into the magazinesof the two new fuel ports. Transfer of the new fuel bundles into a

INTERNATIONALCONFERENCE ON

CANFUEL HANDLING SYSTEMS

ISBN 0-919784-43-7

111, 2. 8 1 1 4

FOREWARD

These proceedings record the information presented at the First CNSInternational Conference on CANDU Fuel Handling Systems held May 13-14,1996 in Toronto, Canada.

The papers for these proceedings were prepared according to guidelinessupplied by the Canadian Nuclear Society and are generally published assubmitted by the authors. Responsibility for the content of each paper restssolely with the author.

The proceedings are copyrighted by the Canadian Nuclear Society. Requestsfor further information concerning these proceedings, permission to reprintany part of these proceedings, or orders for copies of these proceedingsshould be addressed to:

Canadian Nuclear Society144 Front Street West, Suite 725,Toronto, OntarioM5J2L7

Telephone: (416)977-7620Fax: (416)979-8356

( i )

FIRST INTERNATIONAL CONFERENCE ON CANDU FUEL HANDLING SYSTEMS, MAY 1996

AN INTRODUCTION TO THE TECHNICAL PROGRAMby

Bill KnowlesTechnical Committee Chair

The analogy that fuel handling is the "Heart and Soul of CANDU" is apt in that it is themechanism that pumps the life blood, fuel, through the body of the reactor to sustain itsvitality. Should any of the major functions of the fuel handling system fail then the life ofthe reactor would soon wane. This conference gives some insight into the health andintegrity of fuel handling

Fuel handling for CANDU systems, both in Canada and abroad, has contributedsignificantly to high reactor capacity factors while having minimal effect on any forcedoutage rates. Building on this success story are the additional tasks that the fuel handlingsystems have been asked to play, particularly with respect to reactor inspection andremedial work. We have several papers that address the versatility of the system in thisrespect.

Papers on the more traditional aspects of fuel handling design discuss instrumentation andcontrol, structural analysis and several future equipment designs. Also in the future, theeffects of new fuel cycles and fuel designs on the fuel handling system are considered.

Operational considerations are covered by papers on commissioning, training andin-service repair; while the last part of the in-station fuel cycle is addressed by severalpapers on dry storage.

The welcome presence of presenters from India and Romania give our conference a trulyinternational flavour. This affords us a rare opportunity to see how the Indian fuelhandling development for their PHWR compares to the two basic Canadian designs.

I hope these papers and the poster displays excite your interest and that after thepresentations and discussions with your conferees, you will leave the conference withgreater knowledge and the seeds of new ideas.

Welcome from the Conference Chair

RONALD A. MANSFIELDPeng Ceng FIMechEformer AECL Fuel Handling Consultant (Retired)

On behalf of the Canadian Nuclear Society, I have much pleasure in welcoming you to the FirstCNS International Conference on CANDU* Fuel Handling Systems. The theme of ourconference is "Fuel Handling Systems: The Heart and Soul of CANDU - Implementation,Operating Experience and Future Designs."

The high capacity factors achieved by CANDU units is largely dependent upon maintaining thesuccessful performance of the on-power fuelling systems of these reactors. The evolution, over 34years, of CANDU Fuel Handling Systems has involved all disciplines including design,construction, commissioning, operations and maintenance. This conference brings thesedisciplines together to share experience and knowledge gained, to improve the performance ofoperating fuel handling systems, and to advance the future designs.

The four sessions of this conference cover a potpourri of fuel handling topics of interest to alldisciplines. It is possible for attendees to listen to all the lectures being presented to gainmaximum benefit from the conference. In addition, a poster session and display can be visitedduring the reception on Monday evening and in conference breaks. You are encouraged tointeract with your peers as much as possible at this conference, and learn from their invaluableexperience.

I thank you for participating in the First CANDU Fuel Handling Systems conference and wish theteam success in keeping the "Heart and Soul of CANDU" in good order for the future.

RONALD A. MANSFIELD

* CANDU IS A REGISTERED TRADE MARK OF ATOMIC ENERGY OF CANADA LIMITED

(iii)

CONFERENCE ORGANIZING COMMITTEE

Conference Chair

Technical Program Chair

Treasurer

Registration

Conference Facilities

Publications

Secretary

Conference Consultant

Technical Committee

Ron MansfieldAECL (ret.)

Bill KnowlesGE Canada

Jim GeorgasAECL

Sylvie CaronCNS / CNA

Isabel FranklinAECL

Tim McLaughlinGE Canada

Bert FonsecaAECL

Ed PriceAECL

Vijay BajajAECL

Dennis HancockAECL

Allan LaneAECL

Jerry FabianOntario Hydro

Bryan KeelanOntario Hydro

Bill KnowlesGE Canada

Randy WattOntario Hydro

( i v )


TABLE OF CONTENTS

SESSION 1 Page

Evolution of the Darlington F/H Computer Systems 1V. Leung and B. Crouse, Ontario Hydro, Darlington

Configurable Logic in Fuel Handling Control 9W. Ernst and D. Rayment, GE Canada

Pursuit of Excellence in F/H Operator PerformanceB. Keelan and B. Curie, Ontario Hydro, Darlington 15

Better Fuel Handling Performance Through Improved Elastomers and Seals 23R. Wensel and R. Metcalfe, AECL

Pickering Irradiated Fuel Transfer Conveyor Isolation 31D. Koivisto and L. Eijsermans, AECL

Fuel Handling at Cernavoda 1 N.P.S. - Commissioning and Training Philosophy 41G. Standen, C. Tiron and S. MarinescuNew Brunswick Power (AECL-Ansaldo Consortium) / RENEL / FCNE (Romania)

SESSION 2

Fuel Handling Solutions to Power Pulse at Bruce 'A' 51R. Day, Ontario Hydro, Bruce A

Fuel String Supporting Shield Plug for Bruce 'A' 65P. Henry, GE Canada

Fuelling with Flow at Bruce 'A' 75M. Gray, GE Canada

Bundle 13 Position Verification Tool - Description and On-Reactor Use 83T. Onderwater, GE Canada

Operation and Maintenance of the SLAR System at Bruce 'A' 93S. Ahuja, Ontario Hydro, Bruce A

Mini-SLAR Delivery System 103D. Alstein and K. Dalton, Ontario Hydro, Bruce A / Spectrum Engineering

SESSION 3

CANDU 9 Fuelling Machine Carriage 111D. Ullrich and J. Slavik, AECL

CANDU 9 Fuel Transfer System 121Z. Keszthelyi and D. Morikawa, GE Canada / AECL

( v )


SESSION 3 (continued) Page

Overview of the CANDU Fuel Handling System for Advanced Fuel Cycles 135D. Koivisto, D. Brown and V. Bajaj, AECL

Demonstrating the Compatability of Canflex Fuel Bundleswith a CANDU 6 Fuelling Machine 137

P. Alavi, I. Oldaker, H. Suk and C. Choi, AECL/ KAERI (Korea)

Design of Fuelling Machine and Carriage toMeet Seismic Qualification Requirements 145

A. Ghare, A. Chhatre, A. Vyas and H. BhambraNuclear Power Corporation of India

Structural Analysis of Fuel Handling System 155L Lee, AECL

SESSION 4

Pickering Dry Storage — Commissioning and Initial Operation 165S. Jonjev, Ontario Hydro, Pickering

Bruce Used Fuel Dry Storage Project — Evolution from Pickering to Bruce 175R. Young, Ontario Hydro, Bruce NPD

Pickering Fuel Handling Decontamination: Facilities and Techniques 185R. Pan, Ontario Hydro, NTS

Darlington Head 8 Commissioning 195P. Skelton, T. Sie and J. Pilgrim, Ontario Hydro, Darlington / GE Canada

Fuel Handling System of Indian 500 MWe PHWR —Evolution and Innovations 201

A. Sanatkumar, I. Jit and G. Muralidhar, Nuclear Power Corporation of India

POSTER SESSION PAPERS

Analysis of Fuel Handling System for Fuel Bundle SafetyDuring Station Blackout in 500 MWe PHWR Unit of India 211

R. Madhuresh, R. Nagarajan, I. Jit and A. SanatkumarNuclear Power Corporation of India

Evolution of the Design of Fuel Handling Control Systemin 220 MWe Indian PHWRs 221

L Dhruvanarayana, H. Gupta and M. BharathkumarNuclear Power Corporation of India

(vi)

CANDU FUEL HANDLING SYSTEMS CONFERENCE 1996

EVOLUTION OF THE DARLINGTON NGS FUELHANDLING COMPUTER SYSTEMS

by

V. Leung and B. CrouseOntario Hydro

Darlington Nuclear Generating StationP.O. Box 4000, Bowmanville

Ontario, CanadaL1C3Z8

CA9700718

ABSTRACT

The ability to improve the capabilities andreliability of digital control systems in nuclearpower stations to meet changing plant andpersonnel requirements is a formidablechallenge. Many of these systems have highquality assurance standards that must be metto ensure adequate nuclear safety. Alsomany of these systems contain obsoletehardware along with software that is noteasily transported to newer technologycomputer equipment. Combining moderntechnology upgrades into a system ofobsolete hardware components is not an easytask. Lastly, as users become moreaccustomed to using modern technologycomputer systems in other areas of thestation {eg: information systems), theirexpectations of the capabilities of the plantsystems increase.

This paper will present three areas of theDarlington NGS Fuel Handling computersystem that have been or are in the processof being upgraded to current technologycomponents within the framework of anexisting Fuel Handling control system.

1. INTRODUCTION

The Darlington Nuclear Generating Station is afour unit station with a fuel handling systemconsisting of the following primarycomponents:

• Three independent transport trolleys. Eachtrolley contains two fuelling machines for

reactor fuelling and maintenanceoperations. Each trolley is capable offuelling any of the four reactors.

• Four Fuelling Facilities Auxiliary Areas(FFAAs) each containing new fuel andirradiated fuel mechanisms to load andunload the fuelling machines.

The F/H control system that controls thisequipment consists of three independentcontrol systems connected together in anetwork to allow electronic communicationamong the three systems (Figure 1).

Each control system in-turn consists of threecomputers which work together to provide allthe required functionally for fuelling. Thesouth control system contains an additionalcomputer to control a second set of new fueland irradiated fuel mechanisms.

When the initial control system softwaredesign was completed and placed intoservice, the computers were overloadedcausing excessively display update times andfalse stops to the fuelling programs. Thisproblem was successfully resolved by acombination of software efficiency changesand the installation of faster processors thatwere capable of running the existingsoftware.

Subsequently, station requirements mandatedthat the Fuel Handling system be able tohandle both standard length and long fuelbundles to control pressure tube fretting.This additional requirement has forced the


implementation of a disk drive upgrade toaccommodate the larger software.

Lastly, the additional reporting requirementsof the long fuel bundle management programhas necessitated the installation of anelectronic communication link between theFuel Handling control systems and the Fueland Physics LAN to provide improved fueltracking capabilities by the Fuel and Physicsdepartment.

2. Processor Upgrade

The justification for the replacement of PDP11/70 processor started on June 8, 1992.The project was approved on December31,1992. Evaluation tests started onFebruary, 1993 were conducted in GE LabSystem and in DNGS site.

QED95D processor, made by QuickwareEngineering & Design Inc. was chosen.Review & specify QA, documentation &hardware tests requirements for QED95Dstarted in June, 1993. Review quotation &finalize hardware, cost & performanceevaluation was done in July, 1993. An orderof 21 sets of QED95D processor was placedin February, 1994. Unit testing for all 21 setsof QED95D processors in the GE Lab systemwas done in October, 1994. The installationof the QED95D processors in the F/H systemsstarted in November 2, 1994 and wascompleted installations in April 5, 1995.

QED95D processor is a DEC Unibuscompatible processor that is fully compatiblewith all PDP-11 Unibus hardware andsoftware. It addresses four megabytes ofmemory and implements the complete PDP-11architecture, excluding the CommercialInstruction Set.

The QED95D consists of two processorboards, a memory board, a massbus adapterboard and a battery backup & Unibusconnector board. These five processor boardsreplaced 22 boards in the PDP-11/70 CPUchassis, eliminated the PDP-11/70 memorychassis, and three PDP-11/70 battery backupunits. Power consumption of the CPU system

dropped from 40A to 13.7A. QED95Dprocessor has built-in self-check diagnostics,which are initiated when the CPU is poweredup and started.

QED95D is a high-speed CMOS processorwith a 32 KB cache memory, it has a majorcycle time of 110 nanoseconds (ns), allowingthe execution of up to 9 MIPS. Bycomparison, the PDP-11/70 only runs at 1.36MIPS.

The new upgraded processors in each trolleysystem scan and operate 4700 - 5000 I/Opoints, receive input from operator workstations, drive Ramtek Display Generators.They execute various application programswhich operate the process equipment andperform the man-machine interface.

The PDP-11/70's took up to 8 seconds toupdate CRT displays. With the QED95Dwhich is 4 to 5 times faster, the update timefor the CRT displays now is less than 1second.

The installation of the QED95D was straightforward and took about 2 hours for eachcomputer. The total project cost is$910,000.00.

3. Disk Drive Upgrade

A replacement disk drive is required for theexisting DEC PDP-11 RL11/RL02 disk sub-system to provide a modern high capacitysub-system. This will increase the executableand data storage size and give a more reliableand more easily maintained sub-system. Thisreplacement sub-system must be fullycompatible with the DEC PDP-11 UNIBUShardware.

A review of the hardware options for RL02disk sub-system replacement indicates that astandard UNIBUS / SCSI interface would bethe preferred choice so that any standardSCSI storage device could be used now or inthe future and give device independence.

Three replacement disk drives are required.Two of the units must be identical fixed and


each to be divided into two 300M bytes ormore partitions. The third unit must be aremovable media unit.

The storage must be non-volatile. The read /write time must be less than or equal to theRL02 disk drive time (55m sec, transfer rate512K bytes /sec)

The chosen SCSI interface is the CMD 720/Mcard , made by CMD Technologies due to :- Low cost.- Good reliability statistics & warranty.- Excellent support from CMD Technologies.- Good operating experience with the CMDSCSI cards over the past couple years.

The final selection ofreplacement is as follows:

the disk drive

- Two Seagate 1.0 Gb Winchester type unitsto be the on-line units in each F/H computersystem.

One Syquest 540Mb removal harddiskcartridge style unit to be the off-line unit ineach F/H computer system.

These disk drive selections were made baseon:- High reliability statistics.- Long warranty periods.- Competitive pricing.- Good trade publication reviews.

The total quantities of the disk drivereplacement for the F/H computer systems asfollows:

- 23 CMD 720/M SCSI host adapters.- 45 Seagate 1.0 Gb Winchester units.- 23 Syquest 540Mb removal harddisk units.The preliminary cost estimates are:

- Hardware: 300 K$- Software: 350 K$

The Darlington Fuel Handling computerhardware configuration is shown in Figure 2.

4. Fuel Handling LAN

A number of user requirements have beenidentified that require the interconnection ofthe Fuel Handling control computers to theDarlington Information System LAN and to theFuel and Physics LAN. The requirementsinclude the ability to:

• automatically transfer long fuel bundletransaction logs from the Fuel Handlingcontrol computers to the Fuel and Physicsfuel accounting system.

• automatically transfer the fueling activitycomputer logs from the Fuel Handlingcomputers to electronic storage media forlong term archive purposes.

• automatically transfer the fueling activitycomputer logs from the Fuel Handlingcomputers to the Information System LANfor use in technical surveillance andsupport activities.

• automatically transfer the job, sequence,and operation data from the Fuel Handlingcomputers to the Information System LANfor use in configuration management andtechnical support activities.

The current control system interconnects thevarious Fuel Handling control computersthrough the use of 2400 baud serial datalinks. These links are currently running nearcapacity transferring control data among thevarious control computers. These links do nothave the capacity or speed necessary to meetthese new requirements.

An alternative solution was identified byGeneral Electric Canada after searching theInternet for potential solutions. It wasidentified that a mature product with a largecustomer base was using the TCP/IPcommunication protocol. This protocol iscommonly used to interconnect variouscomputer types using high speed ethernetdata links and to interconnect variouscomputer networks.

The product that has been chosen is calledTCPware or RSX by Process SoftwareCorporation. It is designed to run on DigitalEquipment Corporation (DEC) PDP-11computers running the DEC RSX11-M V4.0operating system software. This productprovides FTP file transfer tools, TELNET


terminal sessions, and provides facilities toallow PDP-11 application programs tointerface to the ethernet.

While the Fuel Handling computers arerunning a modified version of the DECRSX11-M V3.2 operating system, theTCPware product has been successfullyimplemented on that platform at GE Canada inpreparation for installation at Darlington inmid-1997. However, RSX11-M V3.2 cannotsupport the TELNET server software ofTCPware.

Each Fuel Handling computer will be equippedwith a DEC DELUA ethernet card thoughwhich TCPware will interface to the ethernet.The various Fuel Handling computers will thenbe interconnected via hubs and fiber opticscables to a single Sun Microsystem SPARC20workstation. The workstation will act as thegateway between the Fuel Handling controlsystem and the outside world (Fuel andPhysics LAN and the Information System LANis shown in Figure 3).

This configuration of the ethernet cablingdoes not allow for complete independence ofthe three Fuel Handling control systems. Assuch, a single failure may impact on morethan one control system. This is consideredacceptable as the information that will becarried on these data links is not required tooperate the Fuel Handling equipment. Inaddition, this configuration does not connectto the FFAA computers. However, if thisnetwork proves highly reliable, then controldata may be carried on these links once theFFAA computers have been interconnectedand the cabling arrangement has beenadjusted to provide functional independenceof the three control systems.

The workstation will be equipped with arouter between the workstation and each ofthe Fuel and Physics LAN and the InformationSystems LAN. These routers, in combinationwith the workstation, will provide security forthe Fuel Handling system from computerusers outside the Fuel Handling system.Initially, the system will be configured toblock all requests from computers outside ofFuel Handling. All communication will be oneway; from Fuel Handling to the outside world.

The workstation will be set up toautomatically transfer various files from theFuel Handling control computers each day andsend them to the Fuel and Physics LAN andthe DNGD Information Systems LAN.

To meet the stated requirements the systemwill function as follows:

• Once each day the workstation willtransfer the long bundle fueling log filefrom each of the control computers andsend it to the Fuel and Physics LAN.

• Once each day the workstation willtransfer the fueling activity log data filefrom each control computer. It will store iton large volume media for long termstorage and also send a copy to theInformation System LAN for user access.

• Periodically transfer the jobs, sequences,and operation data files from the controlcomputers and send them to theInformation System LAN for user access.

Once satisfactory performance of this systemhas been obtained, other potential uses of thislink between the Fuel Handling controlcomputers and the other site LANs include:

• Electronically sending the Fuel ChangeOrders from the Fuel and Physicsdepartment to the Fuel Handling controlcomputers for execution.

• Modification of the jobs, sequences, andoperations on the Information SystemLAN and uploading these changes into theFuel Handling control computers. Atpresent all changes must be entered usingthe control computer software.

• Transferring plant I/O status data from theFuel Handling control computers to theInformation System LAN for use intechnical support and surveillanceactivities.

DARLINGTON FUEL HANDLING CONTROL SYSTEM

NORTHWEST F/H SYSTEMNORTHEAST F/H SYSTEM

2400 BAUDSERIALLINKS

FFAA 1/2CONTROLCOMPUTER

OPERATIONALSYSTEM ICONTROLCOMPUTER

F/M

OPERATIONALSYSTEM 2CONTROLCOMPUTER

F/M 2

TRANSPORT TROLLEY 1/2



•—1F/M 5 ] :F/M 6


1FFAA 5/6CONTROLCOMPUTER


F/M 3 F/M 4


I


I



SOUTH F/H SYSTEM

FIGURE 2

DARLINGTON FUEL HANDLING COMPUTER HARDWARE CONFIGURATION

OED95D

PROCESSOR

UN1BUS

SERIAL

INTERFACE

LA 120

PRINTER

CMD 728/M

ADAPTER

SEAGATEt GB

FIXED DISKDRIVE* « &

SEAGATE1 G8

FIXED DISKDRIVE* 2 & 3

SYQUESTS40 MB

REMOVABLE DKDRIVE* 4

GRAPHICDISPLAYINTERFACE

CPIMULTIPLEXINTERFACE

DZ 11-C8 CHANNELINTERFACE

RAMTEKDISPLAYGENERATOR

VIDEO SWITCH

DZ 11-D8 CHANNELINTERFACE

LINKS TOOTHERPROCESSORS

DIGITAL 4 ANALOG

INPUTS & OUTPUTS

COLOUR

PRINTER

SEMI-AUTO

KEYBOARD #

SEMI-AUTO

KEYBOARD #2

AUTO

KEYBOARD ft\

AUTO

KEYBOARD t)2

FIGURE 3

DARLINGTON FUEL HANDLING LAN

TO FUEL & TO INFORMATIONPHYSICS LAN SYSTEM LAN

OPERATIONALSYSTEM ICONTROLCOMPUTER

DELUA - XC


DELUA XC

ROUTER ROUTER

SUNWORKSTATION

: HUB

FIBRE OPTIC CABLE


DELUA


DELUA XC


DELUA


LEGEND

- DEC DELUA ETHERNET CARD

- FIBRE OPTIC TRANSCEIVER ANDDEC ETHERNET CONTROLLERINTERFACE BULKHEAD ASSEMBLY

FIRST CNS INTERNATIONAL CONFERENCE ON CANDU FUEL HANDLING SYSTEMS, MAY 1996

APPLICATION OF CONFIGURABLE LOGIC IN NUCLEAR FUEL HANDLING

by

W.H Ernst and D.J. Rayment

General Electric Canada Inc.Power Systems Department

Peterborough, Ontario K9J 7B5

CA9700719

ABSTRACT

Control and protective systems operating in theolder nuclear power stations are nearing the endof their reliable operating life. These systems arestill subject to frequent logic changes. Testing thesoftware logic changes is becoming a significanttask with ever greater expense. The softwarebased systems can be replaced with systemsusing configurable logic. These systems providenew, more reliable technology, offer the capabilityfor change, and provide capability for completelogic simulation and test before installation. Thereis a base of operating experience with thesedevices and many potential applications wherethey can be used to advantage.

DEALING WITH CHANGE

The fuelling systems at the Bruce and Darlingtonstations are ever-changing. The systems at BruceA have been in operation since 1977, carrying outroutine automatic on-line fuelling. The design hadevolved up to that time from the initial conceptscarried through the design and manufacturingstages to actual operation of the systems in labtesting at GE. The transition from concept tosmooth operation required many changes to theoperating and protective logic. Many changeswere made during the site commissioning phase,as system parameters were fine tuned to copewith actual site parameters. During stationoperation, new modes of operation and theaddition of maintenance and inspection functionshave resulted in numerous changes to the logic.The Bruce A protective logic is currently runningat software version 35, and the count startedsome years after the commissioning wascompleted. The Bruce B protective system,theoretically a duplication of the A system, isrunning version 50. The Darlington protectivesystem, on line at the station since 1990, isrunning version 23. The certainty in thesesystems is that changes, input and output circuitsand the logic to process them, will be required andmust be accommodated.

The other certainty that we are faced with is thatsystem changes become more expensive toimplement as time passes. The requirements for amore thorough software design process havebecome more stringent in recent years. Designchanges are implemented by personnel who arenot as knowledgeable about the system as thosewho carried out the initial design. More extensivedesign reviews and independent testing have beenadded to the process to compensate.Shoehorning code into the older computers withlimited memory capacity requires reorganizingprogram and data storage locations; this changesthe program flow. This creates more uncertaintyin adequacy of test coverage when modificationsare made.

The original system hardware is becoming difficultto maintain. Is it now time to revisit the controland protective systems because of the equipmentobsolescence? Is it time to overcome some oftheir shortcomings, including reducing costs forhardware maintenance and for implementingchanges? Would a different approach offer theopportunity to more fully test design changes andreduce the site commissioning effort whenchanges are made? From the point of view of thetechnology available, the answers are 'Yes'.

For example: A hardware failure of the protectivesystem at Bruce A recently caused a breach ofstation containment. The protective equipment isover 25 years old, possibly approaching the limitof its reliable operating life. Replacement partsare no longer available. Other protective featuresare designed as backup to prevent incidents, butreliability of operation becomes a concern.

Another example: The Bruce A I/O system hasgone through periods of high unreliability, also atthe end of its reliable operational life. Equipmentoutages put greater demands on the othersystems. We have had to reverse engineer andbuild replacement parts in some instances to keepthe systems operational.


ALTERNATIVES DEVICE HISTORY

Overcoming the problems presented by thesesituations can be done in many ways. One ofthese is to use relatively simple programmablelogic controllers to implement the protective logic.PLCs offer ease of programming and programchange. For the control computers, purchase thelatest off-the-shelf model of equipment similar incapability to the original controllers, make theappropriate software conversions, and install thereplacement equipment. These are practicalapproaches, offering another 25 or so years ofmore reliable operation. The hardware capacitycan be expanded, providing capability ofaccommodating future system logic changes. Onechallenge is to rein in the costs of the softwareconversion, while doing a sufficient level ofquality assurance on the changed software.

However, another uncertainty is introduced with anew computer system. This relates to the designof recent computer systems. There are morecomplicated features used in the design of thesesystems as compared to the design of thecurrently operating 1970 vintage systems.Greater use of cache memory, pipelining of data,parallel fetching of program instructions, and amyriad of other features provide greater speed ofoperation at the cost of the greater internalcomplexity. More complex real-time softwareoperating systems are required to harness thesefeatures. Software conversion would be doneusing higher level languages. There is uncertaintywhether the new system will behave like the oldersystems. There are many examples of emulationsystems being used to reduce this risk. Moresophisticated tools are required to assist in systemdebugging. Implementation challenges are notinsignificant.

OTHER SOLUTIONS

Another approach is to look at other solutions tomeet the requirements, whereby the softwareelement might be removed, yet a changeabledesign implemented. The system could bedesigned so that testing could be far morethorough and more easily carried out than using asoftware based approach. A design offering theseadvantages could be implemented usingconfigurable logic devices. These are moregenerically termed Programmable Logic Devices,or PLDs.

Programmable logic devices have been around andevolving since the 1970s. Some of the earlierforms were application specific integrated circuits,ASICs, which were fabricated by IC suppliersbased on a customer's logic design. Later cameprogrammable array logic, or PALs, a more genericdevice that could be configured by the user.These early devices were limited in I/O and logiccapacity. They use various logic configurationapproaches such as fusible links, where a devicewas "burned" for a particular application. Laterstill came programmable gate arrays; these areconfigured using configuration data thatinterconnects the gates within the chip toimplement a required logic design. These devicesare being manufactured with as many as onehundred thousand gates per device, offeringcapabilities for highly complex logic, or many logicpaths, to be implemented in a single device. Thepinouts on these devices offer up to severalhundred configurable connections to the outsideworld, providing the capability of connecting tohundreds of inputs and outputs, and ofimplementing complex logic functions.

PARALLEL LOGIC PATHS, FUNCTIONINDEPENDENCE

An important feature of the gate arrays is that theinterconnected gates of the logic provideindependent parallel paths for the various logiccircuits that are implemented. The transferfunction connecting an output to any number ofdetermining inputs operates independently of allother circuits within the chip. Logic timing foreach path is a function of the number of gatesand interconnections used by the particular set oflogic, with pin to pin transition times in the orderof tens of nanoseconds; the circuit can bemodeled to determine precisely the timingcharacteristics. Furthermore, predefined timedelay characteristics can be designed in.

10


CHANGEABILITY CASE STUDY

PLDs can be configured by a user who has therelatively inexpensive PC based design tools.Logic designs and changes can be implemented bythe "customer" without requiring manufacturingsupport. Device logic can be reconfigured in thefield, with a very short outage, without the needto remove the device from its application.Although logic devices can be changed in situ, thelogic can be fixed into a storage device so that itdoes not change during system operation oroutage. Non changeable devices could beselected if the required level of system securityrequires that feature. However, for a system thatundergoes frequent changes, system unreliabilitycan be easily introduced by having to removemodules or chips to make changes. Fieldconfigurability can be accommodated by providingthe on-board connections to accept the logic froma PC and "permanently" retain it in a storagedevice for transfer to the logic device on powerup.

ADVANTAGES

The flexibility and capability of PLDs offer manysignificant advantages for applications within fuelhandling, as well as other nuclear applications.The speed and logic capability offer the possibilityof completely dedicated logic for functions suchas the protective interlocks. Implementation ofthe protective logic in the older software basedsystems has worked well. The cycle times ofmost processors, including the minicomputersystems of the early 1970s, have performed thejob adequately. However, there are many logicpaths that are processed sequentially in theprotective software cycle. There is timingdependence, as well as system functioninterdependence, of all logic paths on all others.It is not practical to design tests that would testall situational combinations. There is not enoughtime in the next millennium to run such a suite oftests. At the other end of the technologyspectrum, hard-wired relay logic can be readilyimplemented using PLDs. Rewiring of relaycontacts to make logic changes would beeliminated. The logic replacement offers"programmability" for any future changes.

The Bruce A protective function can be readilyhandled by separating the protected drivesubsystems into a number of sets. Each set ofdrives could then be protected by a single logicdevice. The fuelling system drives fall naturallyinto sets: There are the four unit bridge andcarriage systems, the service area drives, thetrolley, the fuelling machine, and the systemauxiliaries. The drives controlled by the outputsfor each subsystem are independent. Inputsignals from the various subsystems are requiredfor the interlock logic of each of the subsystems.Eight identical modules would accommodate theinputs, outputs and logic for the eightsubsystems. The logic for the programmable gatearrays would be unique for each of the eightmodules. This set of modules would be identicalfor each of the six control systems that make upthe station complement of fuelling systems, onefor each of the two fuelling machines associatedwith the three trolleys. These modules could beused for other general purpose logic functions aswell.

DESIGN TOOLS

The tools for creating the configuration data forgate arrays and for simulating the performance ofthe resulting "circuit" offer full computer aideddesign capability. Schematic capture tools can beused to input the logic in various familiarrepresentations. Logic can be entered as Booleanlogic statements, truth tables, logic symbols or asladder diagrams, for example. The tools providethe capability of designing a higher level userdesign entry language that is best understood bythe designer, and can then present the logic inother formats, as preferred by the customer.After the logic design is complete, the toolsgenerate the configuration logic directly; there isno "hands on" manipulation of the "code". Then,a logic "fit" can be done, based on a target gatearray device. The gate utilization factor can bedetermined, that is, defining the spare capacity inthe device. This factor defines the growthcapacity for logic changes as the design or systemmatures. The system signals can be simulated todetermine the time response of the various circuitsimplemented in the device. All of the design andsystem testing can be simulated beforecommitting the design to actual circuitry.

11


OTHER APPLICATIONS LOGIC SIMULATION AND TESTING

PLDs are of interest at this time because of thenumbers of gates available on a single device, andespecially because of the increase in the pinout ofthe devices. The speed and logic capabilities haveled to their use in other areas within nuclearapplications. These devices form the logicelements on many modules making up moderncomputers and especially emulators of oldertechnology machines. Recently the PDP-11/70computers and memories of the fuel handlingcontrollers in the Darlington station were replacedwith a single module, functionally equivalent, butmuch faster than the original.

The need for testing of the software changes forthe Bruce stations has been met by designing aninterface module around one of these devices.This interface makes a dual port memory residentin a PC appear as a simulated fuelling system.The PC has sufficient speed to simulate theactions of the fuelling system. The control orprotective computer software can be testedwithout intrusive changes that would otherwisebe needed to simulate system operation.

This same interface module has been adapted forother functions. By changing the configurationlogic, and with minor software changes in thecontrol computer, logging data can be passed athigh speed transmission rates from that computerto a PC for transfer to a station local areanetwork, providing up-to-date log data in thetechnical office. Having this log data on-line on aPC offers the capability of easier sorting andanalyzing log to determine system problems andmaintenance needs.

The problem of the unreliable I/O equipment atBruce A can also be resolved by making use of theinterface module mentioned above. This devicecan be reconfigured to provide an interfacebetween the current control computer and arecent model I/O system. This approach permitsthe selection of the best available I/O hardware,providing an interface whereby the controllersoftware remains unchanged and continues tointeract with the system in the same way asbefore the change. System qualification andtesting can be done in the lab, reducing thecommissioning effort after installation of thereplacement system.

One of the greatest benefits of using thistechnology is the ability to more completely test alogic implementation than is the case with asoftware based design. In the case of theprotective system noted above, there are fourdrives associated with the bridges and carriagesfor each of the reactor units. There are numerousinput signals that define the protective logic foreach of these drives, many of the signals commonto all four drives. With the current softwarebased system, these four drives are included inthe same protective logic system as the otherfuelling system drives. It is practically impossibleto test all combinations of all signals occurring inthe common processor in testing the logic for anindividual drive. Limitations have to be placed onthe test cases to achieve reasonable assurancethat the logic is correct. With the alternativeindependent logic of a PLD, more complete testingcan be done to provide assurance thatrequirements are met. Since it is practical tosegregate the logic for the subsystems intoseparate modules, a change in one subsystem hasno effect on the others. Of course, multiplesoftware systems could also be used. In this casethere would still be more significant timingvariations resulting from the variable pathsthrough the software logic. Software qualityassurance objectives can be met, but there is stilla greater compromise between exhaustive andpractical testing levels.

DESIGN CYCLE TIMES

Recently there was an opportunity to propose adesign for a ventilation system control for theBruce A station. This control uses channelizeddata from 24 temperature monitors at criticallocations within the plant to activate theventilation system. The logic requirement wasthat if any of the signals was determined by twoof the three channels to be over the temperaturelimit, ventilation would turn on. The logic for thiscircuit was defined, the configuration logic for aparticular gate array was established, andsimulation tests were run to prove the designconcept within relatively short design time. Thedesign implementation cost for this control wouldbe negligible compared with a software basedapproach. The front-end testing that could bedone to prove the design would significantlyreduce the commissioning effort after installationof the equipment.

12


RISKS

Any new design approach is not without risks.The main risk in use of the technology is the factthat "custom-designed" hardware is required forits implementation. There are no off-the-shelfgeneral purpose modules available offering acustomer base of experience. The problems ofdesigning the appropriate interface for the logicdevices to withstand the environment in whichthey will be operating depends on the skills andexperience of the design group. The upside ofthis is that the design can be tailored to fully meetthe requirements of the particular applicationwithout the need for costly modifications of a"standard" product line. Reliability ofinterconnection to external devices provides adesign challenge. From another perspective, thereis the potential for delays in acceptance ofchanges to use this newer technology by theregulatory authorities.

SUMMARY

Field programmable gate arrays provide aninexpensive, yet versatile means for implementingcontrol and protective logic within a nuclearpower station. The reconfigurability of the logicdevices will accommodate the changes andexpansion found necessary in logic functions. Thesystem logic and its timing can be well definedand tested due to its deterministic nature. Thecapabilities for completeness of testing will reducethe level of effort required during the systemcommissioning phase. Separating the logic intoindividual modules greatly reduces theinterdependence of functions within the "logicprocessor". The completeness of definition andtesting of the logic functions should ease theburden of getting regulatory approval for designchanges.

Programmable Technology Evolution

Programmable System Hardware

ULSISystems

VLSISubsystems

IntegrationLevel

LSILogic Blocks

SSI/MSIGates/Fhp-Ftops

Bipolar Fuse PALSOne-Time

Programmable,

1970's 1980's 1990s

Time

2000'S

REFERENCES

Choosing FPGAs, ASICs, or Cores for DSP Based System Design, B. Tuck, Computer Design, vol 35 no 2,Feb 1996

Technology Trends, Electronic Engineering Times, March 22, 1993

In-System Programmability Manual 1994, Lattice Semiconductor Corp.

Cypress Programmable Logic Data Book, 1996, Cypress Semiconductor Corporation.

Altera FLEX 8000 Handbook, 1994, Altera Corporation.

13

tmxrSaft BLANK


PURSUIT OF EXCELLENCE IN F/H OPERATOR PERFORMANCE

Bryan Keelan and Brad CurieOntario Hydro

Darlington NGDP.O. Box 4000

Bowmanville, Ontario L1C 3Z8 CA9700720

Abstract:

Performance of any F/H system is heavily dependent on the performance of the systemoperators. Sadly, this topic often receives little attention until incidents occur.

Darlington has taken a proactive approach designed to reach and maintain excellentoperator performance.

Expectations and Standards are generated and published. A detailed, goal oriented,auditable training program which includes hands on training modules ensures operatorsreach the standards before qualifications are granted. Refresher courses maintain currentskills. A depersonalized, computerized reporting system ensures training and standards areupdated to match current situations.

F/H System at DNGD

The Fuel Handling system at DNGD canbe viewed as a special purpose roboticmanipulator encased in a pressure vessel.The pressure vessel forms an extension ofa reactor Primary Heat Transport Systemduring fueling operations. Fuel Handlingroutinely passes fuel through containmentboundaries, and handles irradiated fuelbundles, each one of which can containactive elements several orders ofmagnitudes greater than allowablereleases to the public.

The fuel handling system at Darlingtonhas three trolley systems and two FuelingFacility Auxiliary Areas (FFAA's). Thethree trolley systems run in a commonduct under the four units to visit either aunit that needs fueling, or to an FFAA toload new fuel or discharge irradiated fuel.

A trolley consists of two half trolleys, andits associated FFAA. Each trolley half hasa fueling machine, and a set of auxiliariesthat provide D2O and air at appropriatepressures and temperatures as is requiredduring the fueling process.

Control System

The operational computers scan andoperate about 5000 I/O points andoperate 4 work stations per trolley half.The main workstations have a CRT andtwo keyboards for operator interaction.

The computer controls the system in realtime. Operators insert high levelcommands called 'jobs' which define forexample, the number of bundles to bepushed in a reactor channel.

Operator Interaction with F/H

The man machine interface putsoperator in the loop in three ways:

the

1) Command - The control systemrequires operators to initiate fueling runs.Following initiation, operators monitorthe system to have a clear view of thestatus and to be able to act appropriatelyin an event.

2) Monitor - Sensor display data tell theoperator the status of the system, andallows him to take effective action tomaintain control during an event.

15


3) Log - The system logs all auto andmanual operation. Operators manually logfueling operations and problems. Thisprovides an opportunity to reviewoperator and machine performance at alater date if required.

Field Work

The Fuel Handling system requiresextensive field support. New Fuel Bundlesmust be loaded into the New Fuel (NF)mechanism, and their serial numbers mustbe entered into the bundle trackingsystem. When fuel comes out of thereactor, storage modules must be loadedonto the Irradiated Fuel DischargeMechanism (IFDM), and must be movedfor final storage when full.

Any special work requires the loading ofspecial equipment or components.Discrepancies between the expected andactual components in the FM can havedrastic consequences at the unit.

Previous Training Systems

Operators were trained by using severallevels of Training Manuals (TM's) tooperated the machine via operatingmanuals. The TM's explained and justifiedthe principles that operators must followin the Operating Manuals (OM's), anddeviate only with appropriateauthorization. Although good in theory,often there was missing or conflictinginformation. Often the TM's did notreflect the OM information, and were notkept up to date. Many situations were notcovered in either document.

The training manuals were poorly written,making information transfer even moredifficult.

As a result, much of the training wasdone via 'oral tradition' or individualbeliefs held by senior operators as tostatus of systems. Often based on faultyinformation or misunderstanding, theselegends developed a life of their ownindependent of the stationdocumentation. This process was a major

impediment in the pursuit of operatorexcellence.

Many standards and expectations wereimplied. Not formally documented,individual interpretations led to haphazardinterpretation and execution.

Requirements for Operators

The goal of any operator performanceprogram can be stated simply:

Ensure the system is operated safely andcorrectly, and correct responses are madeto fault conditions.

Excellence in operator performance isachieved when this standard is defined,and always met.

Taking these very general statements andconverting them to usable daily directionsis where the circle of excellence comesin. The circle generates standards andexpectations which cover both thephilosophy and direct operation of thejob, as no set of written instructions cancover all issues.

Circle of Excellence

The unified approach to operatorexcellence is best described in the modeldescribed in Figure 1.The circle model conveys severalimportant concepts:

The movement is continuous and selfrenewing. As individuals enter into thecircle, their actions and experiences are abuilding block for the next step.

The circle is an iterative process,adaptable to the real world, where'perfection' is a moving and changingdefinition.

The circle has mechanisms for defining adirection, determining how far off fromthat direction you are going, and it hastools to get you back on track. Since theredefinition phase repeats, the circle

16


analogy allows for changes in direction asthe work environment changes.

Initial Selection:

Initial selection of candidates is based oncorporate hiring practices, which providesassurance of a reasonable level ofcapability. Movement into Fuel Handling ismostly on a volunteer basis, which is thefirst step in having a positive motivationtowards job excellence.

Initial Training

Initial training (level IV) at DNGS startswith a common stream for all operators.Courses cover the full range of stationsystems and both radiological andconventional safety.

Fuel Handling Specific Training

Following completion of generalistoperator training, candidates may wish toenter into the F/H department. This set oftraining is performed in the followingorder:

Field Training: Operators are trained in allthe field equipment, with a'combinationof classroom and on the job training.Expectations are that they will obtain aclear understanding of all F/H fieldequipment by the time the are fullyqualified.

Main Panel Operator (MPO):The MPOlevel of training begins where the fieldstudies stop. Candidates are introduced tothe control panels and control systems.Training concentrates on explaining the'why' as much as the 'what' of controlsystem operation. Candidates areexamined to ensure they have an in-depth knowledge of fuel handling from apanel perspective. Mastery of thispackage allows candidates to be co-pilots.

Copiloting

Copiloting is the transition phase whereindividuals combine knowledge based

training with actual operationalexperience. They are allowed to performOM procedures under the directsupervision of a qualified panel operator.This allows them to gain experience in alimited risk situation..

After the specified period of co-piloting, afinal interview concentrates on OperatingPrincipals and Procedures (OP&P's), andthe Recommended Operating Practices(ROP's). These emphasize the operatingphilosophy as much as the logic.Operators at this level have the additionalresponsibility of dealing more withinterfacing work groups, i.e. maintenance,Authorized Nuclear Operators (ANO's)who operate the reactor units.

Experience

Experience is the gist of the mill in thepursuit of excellence. Operators, like mostof us, learn a great deal through properlystructured experience, in both the copilotand operational phases. Experiencegenerates the problems which measurethe successes and failures of the process.Properly recorded, experience helps pointthe direction to excellence.

Standards and Expectations:

Standards and Expectations are key tothe pursuit of operator excellence.Individuals will not strive for excellence ifthey do not know what excellence is.

Standards are measurable goals againstwhich performance of specific actions aremeasured. Expectations are therequirement that certain actions beperformed or attitudes be upheld.

The standards and expectationsdocuments are the base of the pursuit ofexcellence.

- Standards and expectations define thecurrent definitions of excellence, or themanner in which it may be approached.

-By comparing standards andexpectations against a new goal, or

17


challenge, it measures the amount ofchange required to move to the newdirection.

- Standards and expectations define therequirements of tools to get to the levelof required excellence.

Expectations come from the stakeholdersin the performance process, customers,and the group themselves. They includestechnical and maintenance groups, theoperations coordinator, and the operators.

Generation of Expectations

A major thrust at DNGD in the last yearhas been to formalize standards andexpectations.

DNGD has taken the tack that there mustbe 'buy in' for expectations to be adoptedgenerally and upheld. The first stage is formanagement to provide the expectationthat the F/H system will be operated witha high degree of professionalism.'Performance Objectives and Criteria'provide a starting point. The next step isto have operators accept that havingpublished expectations is a valuable toolto achieving the first, agreed uponexpectation.

The fact that operators provide many oftheir own performance standards is ameasure of the buy-in they already have.Operators (as with other work families)recognize any tool that helps themachieve excellence is in their own selfinterest. No one wants to initiate orworsen a caper.

The generation of 2nd level expectationswas done in a cooperative environment tomaximize buy-in.

Expectations were fleshed out by fieldstaff, and given concrete form that wasmeaningful in the work environment. Asan example, A housekeeping standard byitself is open to interpretation, as twopeople may have two different views ofacceptability. A concrete interpretation ofthat general standard came from a field

operator: "all fuel pallets must be stackedin the north end of the fuel area".

Vigilance

The lessons learned by experience areonly retrieved through vigilance. The skill,or expectation is hard to cultivate, andeasy to loose. Issues such as a positiveattitude for incident reporting, perceivedpositive effects of previous reports, mustbe addressed to have an effect vigilanceprogram. A vigilance program is oftenmade more effective by the use of areporting system:

Fuel Handling Incident Report (FHIR) andTechnical Section Input

To aid in the process of vigilance,Darlington NGD has incorporated a FHIRdata base, with the Significant EventReport (SER) database. FHIRs have aformat similar to SER's, but withcirculation limited to the F/H family. F/Hpersonnel enter depersonalized reports ofincidents that may affect F/H, but are notclassified as SER's. FHIR's are reviewedby operating maintenance and technicalsection and action items are generated.Regularly, the action items are reviewedfor completion and are closed out whencomplete.

FHIRs perform several functions: 1) theymay indicate short coming in the OM's orTM's, 2) they may indicate a goodresponse to an incident that should becommunicated to the operator group; 3)They may illustrate a new hazard that isnot clearly understood; 4) they maydocument an ongoing problem that needsattention from another work group.

Followup actions of FHIR's often involvethe Technical Support Unit, and/or design.FHIRs help those groups focus on issuesimportant to the operator group.

Internal Peer Capability:

Another important player in directing thepursuit of excellence is an Internal Peercapability. On an ongoing basis, this

18


process asks people, "Are we going in theright direction and are we near where wewant to be?". Examples of internal peertools are questionnaires and speciallytrained people who compare standardsand expectations documents with day today operating experience.

Training and Operating DocumentationUpdates

The results of FHIR's, and Internal PeerAudits must be used to ensure trainingand operating documents reflect thepursuit. Documentation that is out of datebrings the whole system into disrepute,and reduces the by-in for the rest of theprocess. Updating of the documentationdemonstrates to the operator family thattheir needs are important, and that thejob is important, besides the obviousattributes of having correct proceduresand information.

Motivational Tools

There are various tools available to shapethe performance of the operator family.Many have been tried and found wanting.The secret of successful use is acombination of how the tool itself isformed, along with its use.

Tools for the pursuit of excellence areoften straight forward. If a standardrequires that a certain action beperformed safely, a tool will be anauthorized procedure to perform thatfunction. Operating manuals, trainingmanuals are tools to reach a standard ofsystem knowledge.

Job expectation guides both inform andmeasure against the current definition ofexcellence, as embodied in the document.

The model of the pursuit of excellence initself is a tool, used to visualize theprocess.

A second key tool of motivation isrespect. Members of the F/H family mustrespect each other, and the contributionsthey bring to the table. Each job group

must realize the important contributionthe others make to the overall goal ofF/H.Respect (buy-in)

Respect must infiltrate every element ofthe model for it to be effective:

Operators must respect the aims of theprogram, the tools that are provided, andthe trainers, and standard providers.

The standard givers and trainers mustprovide tools the operators can respect,and must respect the operators ideas andopinions. Respect and disrespect areshown in many different ways. One ofthe most important aspects in this model,is appreciating and considering otherpeoples points of view, and not discardingthem simply based on the source. It isimportant to recognize that each group inthis circle brings different talents to thetable, and there is no one group that hasmonopoly on them all.

Job Performance Measures

Job Performance Measures (JPM's)provide an important tool to measurequality of the physical actions of anoperator. The JPM instructs the operatorto perform a certain sequence ofoperations, using the real equipment. Thisallows the operator to form a kinetic(body image) memory of the action,increasing the chance of correct operationin the time of need.

Refresher Training

Refresher training provides a means of theoperators meeting expectations on anongoing basis. Refresher trainingreinforces components of the jobexpectations that are not normallyperformed as a part of day to day duties.

Key aspects of refresher training are:

Selection of Training Topics

The FHIR data base helps in the selectionof training topics. New hazards,

19


challenges, or changes covered since thelast session are easily detected andcovered. Topics will be relevant as theyhave already happened.

Refresh rate

The refresh rate is affected by thecomplexity of the covered material andthe rate of use in the field, and theprojected use of the information.

As an example, refresh training onmaintenance equipment used duringoutages should be provided, and providedjust before use during the outage.

Procedures seldom used, i.e. powertransfer procedures, should have a higherrefresh rate than a procedure that is usedevery fuelling push.

Fitting in Shift Cycle

Since the customers for the training areusually on shift, multiple editions of thetraining should be considered.

Problem Reporting and Feedback

Trainers must be responsive to' coursefeedback. While it will not be possible toplease everyone, feedback can provideimportant considerations for coursedesign.

2 0

FIRST INTERNATIONAL CONFERENCE ON CANDU FUEL HANDLING SYSTEMS. MAY 1996

InitialSelection

Experience

RefresherTraining(JPM's)

Expectationsand

Standards

• CIRCLE OFOPERATOR EXCELLENCE

MotivationalTools

Training andOperating

DocumentationUpdates

InternalPeer

Capability

FIGURE 1

FHIR'sTechnical

Section Input


BETTER FUEL HANDLING SYSTEM PERFORMANCE THROUGH IMPROVEDELASTOMERS AND SEALS

R.G. Wensel and R. Metcalfe

Atomic Energy of Canada LimitedChalk River Laboratories

Chalk River, Ontario KOJ 1J0

CA9700721

1. ABSTRACT

In the area of elastomers, tests have identifiedspecific compounds that perform well in each classof CANDU® service. They offer gains in servicelife, sometimes by factors of ten or more. Moreover,the aging characteristics of these specificcompounds are being thoroughly investigated,whereas many elastomers used previously wereeither non-specific or their aging was unknown. Inthis paper the benefits of elastomer upgrading, aswell as the deficiencies of current station elastomerpractices, are discussed in the context of fuelhandling equipment. Guidelines for procurementstorage, handling and condition monitoring ofelastomer seals are outlined.

In the area of rotary seals (non-elastomeric) therehave also been significant advances. In the fuellingmachines, low friction is needed for accuratepositioning. The development of an improved sealis described, which since 1992 has saved at leastM$l per year at Pickering alone.

2. INTRODUCTION

Over the past twenty-five years the Fluid SealingTechnology Unit (AECL, Chalk River) has helpedsolve a variety of sealing problems in CANDU® fuelhandling equipment, primarily in response tospecific station requests for help. Following aresome of the component problem areas that havebeen investigated and resolved:• D2O hose failures (Darlington '95, Lepreau '91,

NPD and Douglas Point)• Marotta fill, vent and drain valve seal retrofit

(Darlington '94)• Snout plug and squeezer mechanism seals

(Pickering '91)• Darlington and Bruce flathead O-rings and quick-

disconnect seals (-'93)• NPD snout seals• Ram ballscrew face seals

We are now at a point in our knowledge ofelastomers, where we have identified specificelastomer compounds that have "superior"performance in each class of CANDU® service. Awholesale upgrade of seals on fuel handlingequipment to the "superior compounds" isrecommended. Furthermore, adoption of improvedelastomer quality assurance, procurement andhandling practices as well as improved in-servicemonitoring are recommended.

3. ELASTOMERS

Historically, equipment in nuclear plants hascontained whatever elastomer each componentsupplier traditionally used for non-nuclear service.Most plant operators do not realize the resultingproliferation of elastomer compounds, many ofwhich are far from optimal for the serviceconditions, has multiplied the costs to providestation reliability, maintainability and safety. Bystandardizing on a handful of "superior" elastomercompounds, service life can be extended,maintenance planning and safety can be improved,environmental qualification can be streamlined, andprocurement and handling of replacement parts canbe simplified. Extensive tests have identified highperforming compounds for each class of CANDU®service. These designated "superior" compoundsoffer gains in service life sometimes by factors of tenor more over others being used. Moreover, theaging characteristics of the "superior" compoundsare becoming much better known than those of othercompounds, many of which are only known bygeneric type.

4. CRITERIA FOR ASSESSINGELASTOMERS

To establish the life of any elastomeric component,the type and magnitude of deterioration that causes afailure in service must be known, as well as thedeterioration rate. There are many possible types offailure (e.g., extrusion, chemical attack, wear-out,

23


tensile cracking, load relaxation and compressionset). They are all influenced by a combination ofmaterial characteristics and service conditions.Commonly reported material properties alone cannotquantify extrusion resistance, etc. Some propertiesare closely related to common failure modes(e.g., tensile strength to tensile cracking failure), butother commonly quoted properties such as hardness,ultimate elongation, permeability and thermalexpansion are only tenuously associated. A primarychallenge in assessing an elastomeric component fora particular service is to determine when and how itmight fail in service. Relevant and comprehensivefailure criteria based on functional properties arenecessary for meaningful life prediction. Theseneed to be based on a fundamental understanding ofthe performance needs for the particular application.

5. IMPORTANCE OF COMPOUND- ANDSERVICE-SPECIFIC DATA

A database of relevant properties and behaviour iskey to elastomer selection and life prediction. Thisdata must be compound-specific because within agiven elastomer class (e.g., nitrile), the base polymeris compounded with varying amounts of fillers,vulcanizing agents, anti-oxidants, anti-ozonants,processing aids, plasticizers and accelerators fromany number of suppliers. These variables, and themethod and degree of mixing and curing, allprofoundly affect functional properties of the finalelastomer product.

Properties alone, however, are not enough for thedatabase to be usable for severe service. It must alsoinclude service-specific behavior. Most of thepublished data on elastomers is misleading because:(1) the effects of the fluid are neglected (e.g., airversus water versus other fluids), and (2) themeasured damage parameters that life predictionsare based on often have little bearing on how a partactually fails in service. Parameters are more oftenchosen for testing convenience than for functionalrelevance.

Besides choosing the most appropriate damageparameter(s), the level of damage considered toconstitute a failure must also be chosen judiciously.Compression set may be correctly identified as themost likely failure mode for a particular application,but its magnitude for failure may be poorlyestimated. For example, a compression set failurecriterion appropriate for a piston seal will severelyunderestimate life for a bolted flanged joint. A highcompression set criterion is appropriate for a flange

seal since the seal is in a static, highly squeezed faceseal arrangement, with no extrusion gap and nochanges of squeeze. In contrast, a low compressionset criterion is appropriate for a piston seal that isdynamic, lightly squeezed for low friction and wear,and has parts with tolerance stack-ups that createsignificant eccentricity between the piston and bore.

Naturally, the amount of testing must always be acompromise with the value of the results. Considerthat accelerated thermal aging requires data at fourtemperatures for reasonable extrapolation, and ateach temperature the duration of the test must beiterated to obtain the desired level of damage (seeFigure 1). Consequently, developing a compound-specific database can become very expensive. Thisis another reason to rationalize the number ofcompounds used in the field to the fewest thatadequately cover the required range of applications.

6. "BATCH" APPROACH TO QUALITYASSURANCE

To ensure relevance of elastomer test data, theingredients and processing variables for each chosencompound must be closely controlled forconsistency, both in original qualification testing andin subsequent service. To ensure that the correctspecific compound is received, purchasingspecifications must not open the door to othercompounds in the same class of elastomer.Otherwise, performance in service may beunacceptable (i.e., low safety margins, unreliabilityand frequent replacement). If alternative compoundsare needed as back-up, each must be separatelyqualified. Purchase specifications, as a minimum,must require that each elastomer seal be of aparticular compound, traceable to the particular"batch" of ingredients, mixed and processedtogether to form the unvulcanized stock from whichthe part was made. A certificate of conformanceshould be supplied specifying the compound, itsbatch number and date of cure, along with thehardness, specific gravity and tensile strength ofsamples from that batch, as compared with themanufacturer's expected values. In this manner theelastomer compound can effectively becharacterized.

7. DEFECTS IN ELASTOMERS

Inspection methods and rejection criteria for defectsare often neglected in quality assurance programs forelastomeric parts. This can compromise theirintegrity and reliability. Surface defects in most

24


parts can be detected by the unaided eye. In the caseof elastomeric parts this is most effective when theelastomer is strained appropriately (i.e., by bending,pulling, pressurizing, etc.), since many defects suchas cuts and tears are difficult to see in unstretchedparts. Size can be compared to reference standardsthat correspond to acceptance limits, and (ifnecessary) measured using optical and mechanicalaids (e.g., calibrated magnifier, depth-measuringmicroscope, stylus profilometer).

For detection of internal defects (inclusions, voids,discontinuities in reinforcing materials, etc.), as wellas for monitoring state of aging (on the shelf or inservice) a non-destructive technique calledelastodynamics has been found most useful. A toolhas been developed, employing this technique forinspection of O-ring seals. Basically the toolmeasures reaction force on two pinch rollers whilethe seal is driven and squeezed between them.Localized defects are signalled by spikes in reactionforce. Any generally low or high force, or variationaround the seal, signifies abnormal properties whencompared with a known baseline. A version of thistool, developed for monitoring the condition ofelastomeric parts of all types, is described inSection 9.

8. STORAGE AND HANDLING

Elastomers are subject to deterioration with time,temperature, and other environmental influences.Ideal storage conditions are cool, dark and free fromcontaminants (such as ozone, solvent vapor, etc.).Elastomer parts should be stored in a relaxed state,free from strain (i.e., not folded, twisted, or hangingon a rack). Their shelf life (expressed as expirydate) should be stated and be rationally based (e.g.,if 90% of "as-new" lifetime for the particular serviceis deemed acceptable, and proper storage at themaximum allowable temperature is known to cause1% loss per year, then shelf life is 10 years). Manyelastomers are very stable under store-roomconditions. Measurement of critical functionalproperties (e.g., compression set, extrusionresistance, hardness) of elastomers of certainethylene-propylene and nitrile compounds storedunder proper conditions has shown them to beessentially unchanged after more than twenty years(for example, see Figure 2). Unfortunately, not allelastomer compounds are this stable.

Many elastomers are incompatible with commonsolvents. Ethylene-propylene elastomers are notablefor their lack of resistance to petroleum-based

products. Some elastomers have very poorresistance to cutting and tear propagation. To coverthe many facets of proper use and handling ofelastomeric parts, a one-day on-site training coursefor mechanical maintainers has been developed.This has been very well received by stationpersonnel, and has been presented at eleven venues.

9. MONITORING IN-SERVICEDEGRADATION

Nuclear plants contain many examples of vaguelydefined elastomer compounds that make accurateprediction of service lifetime impossible (yet manyof these parts are accessible for interim inspection).Even in applications where well-defined compoundsare used, the service conditions may not be wellknown. There are also cases where station personnelwould like to assess the state-of-aging of elastomericcomponents held in inventory for long times.

In applications such as these, a recently developedtool called an elastodynamic spot tester promises tobe very helpful. Sufficient data is needed to showhow a particular elastomeric compound's propertieschange qualitatively with time in the serviceenvironment. Also, the starting quantities andminimum required quantities for these propertiesmust be known. Then, making an interimmeasurement of these quantities using theelastodynamic spot tester gives a non-destructivemethod to pinpoint the current "effective age" of thepart in terms of the percentage of service lifetimeexpended.

For example, most elastomers age-harden. Ifhardening is the failure mode and functionality ofthe part requires hardness less than say80 durometer*, with new parts being 70 durometer,then an interim measurement of hardness of75 durometer suggests that half the service liferemains (50% effective age), assuming thedegradation rate is linear with time. However, agingdata may also show that hardening accelerates andthat the part's age is therefore already say 90%.

The advantage of elastodynamic spot testing is thatit measures much more than hardness. It alsomeasures stress relaxation and recovery, and canmeasure these as a function of mechanical stresslevel in order to discern the onset of permanent

"Durometer" is a measure of rubber hardness, asdetermined by indentation depth of a stylus under agiven load after a given period of time.

25


damage. Essentially, the stiffness, damping andstrength are quantified to effectively "fingerprint"the material. It then only remains to relate these toage, in the same way that measurements of weight,running speed and subsequent pulse rate, forexample, might pinpoint the effective age of anindividual in a population of genetically andenvironmentally similar humans.

10. FUELLING MACHINE RAM SEALS

For fuelling machines in Pickering and the 600 MWCANDUs, the severe requirement is for seal torqueto be no more than 0.6 N.m—low enough to achieveaccurate positioning of the rams. Seal glandconditions are:

-25to55°C- 0 , 3 or 10 MPa (gauge)- 24, 36 or 246 rpm, reversible

Reliability is essential. Many seal types were testedduring early development. The most acceptable wasa six-pocket-with-orifice hydrostatic design, similarto those used in the first CANDU Main CoolantPumps at NPD (Nuclear Power Demonstration),then at Pickering. Its face materials were bronze vs.carbide-coated stainless steel. Leakage was about2 L/min at full pressure, and its main failure modewas dirt or erosion affecting the orifice and de-stabilizing the seal, thus causing face rubbing,friction and wear. It was also difficult and expensiveto manufacture, with quality assurance of theorifices, face coating, and the bellows for axialmovement being particular problems.

Performance was tolerable throughout the 1970s andearly 1980s because each plant kept a spare machineavailable, but the days of generous spares came to anend in the mid-1980s—higher reliability wasdemanded.

A different type of hydrostatic seal (Figure 3:CAN13 seal) was developed using analysistechniques and materials not available to the earlyCANDU designers. The bellows was replaced by anO-ring of material with demonstrated capability forthe service. Silicon carbide was chosen for the newseal faces based on extensive hydrostatic seal testingin the 1970s, when it was found to be both resistantto erosion and tolerant of moderate nibbing contact.The need and means to hold it securely incompression (by shrink-fitting) to prevent breakagehad also been understood.

Requirements for low torque and low speednecessitated a hydrostatic seal; the extreme range ofpressure required a seal face that would neithererode while on pressurized standby, nor wear whenrubbing at pressures as low as atmospheric.

The flat-conical seal face configuration wasdeveloped so that the flat region of the face wouldrub lightly and not wear out-of-flat while running atatmospheric pressure and 246 rpm. The conicalregion was optimized to meet the torque requirementwhile having minimum leakage at the higherpressures. There were three reasons for this:

(1) to minimize the erosive flow,(2) to minimize the amount of dirt

consequently filtered out between the sealfaces,

(3) to reduce the 2 L/min leakage per sealsuffered with the previous seal design.

CAN 13 seals were first installed in Pickering in1988 after extensive testing in laboratory testers andfuelling machines. Full commitment to this designwas made in 1989 and completed by 1992. Therehave since been no failures of the 64 seals running inthese eight reactors, despite dirty operatingconditions in Pickering-B that have caused manypump seals to fail. The 600 MW CANDUs havesimilarly changed to the new seals.

Savings due to installation of CAN13 seals atPickering are estimated to be at least MSI per year,with about a third coming from maintenance savingsand two thirds from extra kW-h of production.Extrapolating over the expected lifetimes of theexisting plants now using this seal, the savings are ofthe order of M$20. The economic viability of futureplants has also been much enhanced—all thiscoming from an original 1985 investment of $53,230for design, development, rig-testing and supply ofthe first two seals.

11. ACKNOWLEDGEMENTS

The contributions of many colleagues and fundingfrom the CANDU Owners Group towards the workpresented here is gratefully acknowledged.

26


Compound "A'

V)co

•55CO

£Q,oo

in air 180°C 170°C 155°C 135°C

Time for85% Setat ServiceTemperature

-11/Temperature (K ) ServiceTemperature

Figure 1: Arrhenius Thermal Aging Tests. This figure illustrates the number ofdata points required for a reasonable extrapolation.

27


<D

CJOV)<n

Eo

O

100

8 0

—O*L |As New

— 9 — A s New

---Q---After 20

- - • - - A f t e r 20

(121°C

(177°C

years

years

in air)

in air)

Natural

Natural

Aging

Aging

(121

(177

°C

°C

in

in

air)

air)

Naturally aged O-rings were in afree state, in a dark or dimly litarea and below 30°C at all times

25% squeeze0.139" sectional

diameter O-rings

10 100

Time at Temperature (h)

1000

Natural aging comparison

Figure 2: Effect of Twenty Years of Natural Shelf Aging on a Specific Nitrile

O-Ring Compound.

28


STA7OR

''p^^*X>^X3

AXIALLYMOVINGO-RING

ROTOR

Figure 3: CAN13 Seal for AECL Fuelling Machine Ball Screw Rams. The rotatingseal face is flat-conical to give hydrostatic lubrication when pressurized.

29

KEXTion ei«W;iiiS5SSSSxb:i>jj:.iaffiS-


PICKERING IRRADIATED FUEL TRANSFERCONVEYOR ISOLATION

by CA9700722

D.J. Koivisto and L.J. Eijsermans

Atomic Energy of Canada LimitedSheridan Park, Ontario, Canada

L5K 1B5

ABSTRACTPickering A NGS has been in operation for25 years and is one of the longest inservice CANDU1 stations. Someunderwater fuel handling equipment,notably the conveyor stops, have beenwithout maintenance throughout thattime.

This paper describes the concept of aconveyor isolation system that permitsdraining of a single or multiple elevatorcolumns and also the early stages of adevelopment program for the elastomericsealing element. The prototype sealelement has been proven in lab tests tobe capable of limiting leakage to 0.5IGPM at the design pressure of 6.5 psi.

The design of a sealing element isparticularly interesting because theconveyor tube is a square cross-sectionwhich contains an additional obstruction ,a conveyor drive cable.

A seal delivery, actuating and positioningsystem has been conceptually laid outand the design is proceeding withprojected implementation in 1998.

1. INTRODUCTION

The Pickering A irradiated fuel transfersystem has been in service for more than25 years and some components have beenalmost completely inaccessible formaintenance. Some of these items are theconveyor stops and possibly the conveyortubes. To perform this maintenance, partof the conveyor tube must be drained anda temporary maintenance bung put in placeto maintain containment for the reactorbuildings connected to that conveyorsystem. The conveyor system requiresconfined space work rules whileperforming any maintenance in the tunnelsbelow the elevation of the bay watersurface.

The Pickering A irradiated fuel transfersystem includes a long conveyor systemthat transports irradiated fuel bundles fromthe reactor buildings to the irradiated fuelbay. The conveyor system and elevatorsare filled with bay water which providespart of the reactor building containmentsystem.

The conveyor system comprises ahorizontal, watertight, square stainlesssteel tube; a conveyor cart, which carriesthe fuel, and the drive system for the cart.The conveyor tube is approximately 14feet below the surface of the bay waterand runs though a long concrete tunnel 4feet wide and 4 feet high which connects

1 CANDlT- Canada Deuterium UraniumRegistered trademark of AECL

31


the elevators of both RB2s to the IFB3. Thedetails are shown in Figures 1,2 and 3.

To drain the conveyor tube formaintenance, a means of isolating thereceiving bay water is required. The baycan not be drained because of thepresence of irradiated fuel particles whichare currently shielded by the water.

If only one isolation point is provided itwould be impossible to isolate the Unit 2West elevator without draining all the otherelevators in Units 1 and 2.

2. REQUIREMENTS

The conveyor tube sealing/isolationfixture must permit isolation of any singleelevator bottom housing. The sealingsystem must have built in redundancy.The system must by mechanically lockedinto place when isolating the watercolumn during personnel access.

Although not required to be absolutelywatertight the seal must limit leakage to0.5 IGPM*. The seal must be able totolerate occasional exposure to gammafields of up to 10 Rads per hour. Thesealing element must have a service lifeof 30 days.

The sealing system must not requireassembly or set up inside the smallconveyor tunnel.

The sealing/isolation fixture shall bedesigned to conform with Ontario Hydro'sergonomic design guidelines.

Compressed air shall not be dischargedunder water from any pneumaticequipment used in the fixture design.

3. CONTAINMENT

Pickering A System

The reactor building is a part of thecontainment system designed to preventthe escape of activity in the event of anypostulated accident to the reactor. To keepthe size of the RBs to a minimum, reactorauxiliary systems which do not containhighly radioactive fluids and which,therefore, do not require the containmentand shielding provided by the RB, arelocated in the Reactor Auxiliary Bay, whichincludes the IFB.

One of the functions of the spent fuelelevator in normal operation is to provideRB containment by means of a hydraulicseal between the RB and the IFB.However, when an isolated elevator isdrained of its water and an opening iscreated in the lower housing or in theconveyor tube the containment boundaryis broken via the openings in the tophousing assembly. The openings in the tophousing must, be sealed before drainingthe elevator. The major opening can beconveniently sealed by clamping the fueltransfer mechanism to the elevator. Thetwo access holes at the top of the housingcould be sealed using commerciallyavailable expandable bungs. This leavesthe openings for the mechanisms (such asfuel stop and return ram assembly, itsdrive motor and gear box, etc) which arefastened to the top housing. It is unlikelythat these connections are air-tight. If not,then this equipment must be removed andthe openings temporarily sealed withblanking plates.

The sealing/isolation fixture provides thecontainment function for the other unitthat is not opened up and may be atpower. This is the reason that the bungmust withstand the accident condition of12.5 psig which is 6 psi greater than thedesign pressure.

2 RB - Reactor Building3 IFB - Irradiated Fuel Bay4 IGPM - Imperial Gallons Per Minute

32


Pickering B System

Pickering A and B spent fuel elevators aresimilar in design dimensions and inconcepts with respect to reactor buildingcontainment. The B system has individualconveyors for each reactor building,therefore the draining of one elevator doesnot affect the other building. The conceptfor containment maintenance applied in Ais also applied in B. That is to make theelevator top housing and upper column thecontainment boundary when the water isdrained. This is accomplished by pluggingall the leakage paths into the top housing.The top housing and column are designedto withstand the +/- 6 psi differentialpressure The elevator lower columnembedded part is already containmentboundary.

The major leakage paths are the elevatorport and the access holes. There are manyminor miscellaneous leakage paths wherevarious actuators, shafts and cylindersenter the top housing.

Depending on the concept for closing theport there are some differences betweenA&B. The A port is sealed by the transfermechanism. In B there is a gap betweenthe transfer mechanism and the port. Thisgap was intentionally provided to allow theoperation of a "sniffer" for defected fuel.The B port would have to be sealed with aspecially designed bung conforming to thecode requirements.

Both A and B top housings have twoaccess holes on top which have to besealed so the concept would be similar,differing in details

The miscellaneous leakage paths in the tophousing of A and B may differ in detail butthe concepts could be common.

Therefore the containment conceptproposed for A is also applicable for B

4. SEALING SYSTEMSALTERNATIVES INVESTIGATED

Several alternatives were considered andevaluated before setting out on the designand development of the reference design.Freezing was considered as a commonmethod of isolation used in CANDU plants.However freezing had been unsuccessfullytried on the conveyor tube in the past andthe large amount of liquid nitrogen requiredin a confined space presented a hazardunacceptable to Station operatingpersonnel.

Air-inflatable bags such as used in otherindustries for isolating conventional pipingwere considered. This method has severaldisadvantages. The bag is unlikely to sealthe inside corners of the square tube.Worker safety would not be sufficientlyguaranteed by the bag which would besubject to sudden loss of air pressure orpuncture.

A concept was considered of injecting athermoplastic material which would hardenin place forming a plug. It could later beremoved by melting. The drawbacks forthis type of system are the complexequipment for installation and removal andthe possibility of incomplete removal.

The concept selected was the bung basedon an expandable elastomeric seal. Thisestablished method of fluid sealing offers anumber of advantages in this application.Custom moulded elastomer seal elementscan be made to conform to the shape ofthe tube. Most elastomers can surviveradiation doses up to 106 rad withoutlosing significant physical properties. Manygrades of commercially available materialcould be safely used in this ambienttemperature application. Mouldableelastomeric material can offer morefreedom in design than the other conceptsconsidered.

A two point isolation system was devisedthat is capable of isolating one singleelevator. The system consists of twobungs with a linkage and a delivery

3 3


system. The bungs have double rubberseals with remote actuators. See Figure 4

5. POSITIONING AND ACTUATINGSYSTEM

When the design concept for the isolationfixture was established the considerationwas given to the method of its delivery tothe correct location in the conveyor tube.The minimum dimension of the fixturewould not allow its insertion through theopenings in the elevator top housing. Thefixture could be lowered through theaccess hole in the floor plate in spent fueltransfer room to the bottom of the elevatorlower housing. Once in the lower housing,however, it would be impractical to dragthe isolation fixture inside the conveyortube.

The most convenient means of installingthe isolation fixture would be through thebay water, into the conveyor unloader.Cutting a portion of the conveyor tubeextension from the bay wall to theunloader was also considered but wasdeemed unnecessary.

A system was conceived that wouldpermit loading of the isolation fixture fromthe IFB. The conveyor cart would be driveninto the bay where an adapter would belowered onto the cart . The adapter locksinto the slots of the cart. The first bungassembly which has wheels to permit it totravel down the conveyor tube, is attachedto the adapter. See Figure 4.

The cart would then move forward topermit the installation of a connector pieceand the second bung assembly. Then thecomplete isolation fixture can be moveddown the conveyor tube to the appropriateelevator bottom housing. The connectingpiece is of sufficient length that the twobung assemblies can effect a seal on bothsides of the elevator housing including theconveyor stop.

The bungs are accurately positioned at theelevator by the irradiated fuel conveyorcart. A solid metal backup system locksthe bungs into position to prevent slippageThe isolation fixture position feedback isprovided by the conveyor cart positioninginstrumentation. The conveyor stop is notrelied upon because it may be out ofservice.

The actuating system is still in the earlystage of development and concepts suchas hydraulic, electric and ballscrewactuators are being discussed.

6. DEVELOPMENT AND TESTPROGRAM

The development of an expandingelastomeric seal to form a watertight sealinside a square tube is unique. The cablelying on the bottom of the tube presentsanother problem. A small stabilised leakrate is desired.

During the concept study severalalternatives for sealing the conveyor tubewere looked into. Of the methodsconsidered, the approach based onexpanding a solid elastomer seal wasselected for further development. Theseal configuration was designed inconjunction with a simple expansionmechanism consisting of a movabletruncated pyramid plug. By moving theplug axially the seal shape can bemodified in radial and axial directions.

During the design of the seal it wasrecognised that the large radial gap of Vi"required a substantial cross-section andconsequentially a large force to expandthe seal. Additional expansion by a factorof 1.4 will be required at the sharpcorners. To avoid expanding a largecross-sectional volume all at once the sealwas given a square tubular shape. In theaxial direction the wall-thickness wasvaried from a thin section that allows pre-engagement with the expansion plug to agradual thickening matching the slope ofthe plug to provide the bulk for

3 4


expansion. The thick straight sectionfollowing behind was tapered on theoutside and inside to a flat sectionsuitable for mounting of the seal. Thethick section therefore could expand byriding on the slope of the plug but alsoflex outwards providing bulk forcompression. The clamping bars served toprevent the seal material from flowingaxially. The seal shape thus was modifiedby expansion in the radial direction andcompression in the axial direction. Figure5 shows a sketch of the seal cross-section before and after engagement. Theseal is designed to be self-locking in theinstalled condition.

The additional material required at thecorners of the seal can be provided inseveral ways. The most simple manner isby modifying the inner mould shape togive more material on the inside corner ofthe seal where it rides on the slope. Theslopes of the truncated pyramid can besculptured or the outside dimensions ofthe corners could be increased. However,both these methods require more complexmachining and will be difficult to modifyafterwards. Accordingly, for thedevelopment work the preferable way ofchanging the seal shape as required wasby modifying the mould. The moulddesign was modular and accommodatedeasy modifications by allowing re-machining of parts only. The mould wasmade from sections of aluminum anglefixed to a base plate in the form of asquare annulus. Strips of aluminum barwere machined and attached to outside ofthe inner square and the inside of theouter square to produce the requiredcross section

The material selected for the seal waspolyurethane. It is a very tough andstrong material with a high tear strength.It is one of the better materials forradiation resistance and the temperaturerange of application is quite acceptable.To keep the expansion loads reasonablethe hardness of the seal material waskept in the low (40 to 55) Shore-Adurometer range. For the development

seals two urethane materials were used;Adiprene L-83 and Conathane TU-4010.Both materials can be cast relativelyeasily using the same mould. TheAdiprene L-83 was polyol cured to lowerthe hardness to around 50 Shore A. Thismaterial also required multi-componentmixing and a curing temperature of 100°C. The Conathane TU-4010 is a twocomponent material which can be mixed,poured and cured at room temperature.The Shore A hardness of 40 can belowered further with the addition of amodifier.

To test the seals a section of theconveyor tube was constructed and aseal expansion mechanism was designedand built. Figure 6 shows the completetest arrangement. The leakage flow ratewas measured at the design differentialpressure of 6.5 psi and the accidentcondition differential pressure of 12.5 psi.

The details of the seal expansionmechanism can be seen in Figure 7. Thedrive rod and expansion plug are movedaxially by means of a double acting nuton a threaded portion of the rod end. Thisnut is held captive on the end of thesupport tube. All leakage paths at theplug, drive rod, seal flange and blindflange are sealed with O-rings. The travelof the expansion plug is 2V*".

A number of seals were manufacturedfrom Adiprene L-83 or Conathane TU4010 materials to Shore A Hardness 55and 40 respectively. The first seal wasmoulded with a constant cross-sectionwithout allowance for extra material atthe corners and without a chamfer orgroove to accommodate the cable. Thisseal provided an initial assessment of themagnitude of the expansion force,required stroke length and flow of sealmaterial with corresponding change inshape.

The second L-83 seal was dimensionallychanged to fit better on the mountingflange and the shape was changed bytapering the inside bars towards the

35


corners giving more bulk at that location.The expanded seal did fit tight inside thetest tube except for gaps of about 1/16"x V2" at three corners.

To improve material flow and sealing atthe corners the next seal was made fromthe softer Conathane TU - 4010 material.The shape was also modified toaccommodate the cable and to give evenmore material towards the inner corners.This seal was tested at a differentialpressure of 6.5 psi and yielded a leak rateof 3.4 IGPM.

To provide visual assessment of the sealshape, a transparent Lexan conveyor testtube was constructed. A 1 inch squaregrid pattern was drawn on one face ofthe seal and when expanded the seal firstmade contact with the centre of the faceof the conveyor tube and then thecontact area spread out towards thecorners. The early clamping at mid-planecaused the seal material to be held thereby friction and be stretched towards thecorners. This helped to visualise thatpremature clamping at mid planeprevented full sealing at the corners and itillustrated the need to make first contactat the corners. The grid pattern on theseal face became barrel shaped andshowed that seal material was beingextruded forward of the expansion plug.The expansion plug, seal and walls of theLexan test jig were lubricated with asilicone mould release agent to reduceand delay the friction factor. The sealmaterial did indeed flow much earliertowards the corners and the contact areawith the conveyor tube was increasedespecially in the corners. The mould wastherefore modified to move the cornermaterial forward in the next seal and atthe same time to reduce material at mid-plane. Figure 8 shows the shape at theinside corner of the seal.

The result of pressure testing the finalseal made from Conathane material was aleak rate of 0.9 IGPM, all of it at thebottom corner. When the test wasrepeated with a %" solid rod instead of

the cable, the leakage rate was reducedto 0.6 IGPM. The seal was leak tightexcept for a small leak at the uppercorner.

The seal was installed again with everyeffort taken to position it concentric.Tested with the cable in place the sealproduced leak rates of 0.4 IGPM in onedirection and 0.45 IGPM in the oppositedirection. The leak rate at 12.5 psidifferential pressure was measured at 1IGPM while the seal maintained itsintegrity. This was the test requirementfor accident conditions of overpressure. Inthis condition a higher leak rate ispermitted temporarily.

The torque required to drive theexpansion plug over the full stroke lengthwas measured at 29 foot-pounds.

7. ACKNOWLEDGEMENTS

The authors would like to acknowledgethe contributions by Rathin Chakrabortyto the project concept and design phases.Barrie Keal of the Sheridan ParkEngineering Laboratory also contributed tothe development of the working prototypeseal. The overall project was funded byOntario Hydro's Pickering ND.

3 6


FIGURE 1FUEL TRANSFER ELEVATOR AND CONVEYOR

164'

•

219'

STOP MECH i(TYP)

2WW2WE

2EW2EE

253'-6"

1YVE '' 1EE1WW 1EW

FIGURE 2CONVEYOR ELEVATION

37


FIGURE 3PICKERING A CONVEYOR

• CONVEYOR CART

ADAPTOR

BUNG ASSEMBLY

WIRE ROPE-

FIGURE 4CONVEYOR CART WITH BUNG ASSEMBLY

38


TruncatedPyramidPlug Seal

ClampingBar -

ExpandedSeal

Square Conveyor Tube Section

FIGURE 5CONVEYOR SEAL SCHEMATIC

FIGURE 6CONVEYOR SEAL TEST RIG

39


Seal Support

Stationary Mounting Flange

nDouble Acting Nut

FIGURE 7CONVEYOR SEAL EXPANSION MECHANISM

FIGURE 8URETHANE SEAL ELEMENT

40


FUEL HANDLING AT CERNAVODA 1 N.P.S.- COMMISSIONING ANDTRAINING PHILOSOPHY

by

G.W. StandenAECL-Ansaldo Consortium

Cernavoda 1 N.P.S., Cernavoda, 8625, Romania(on attachment from NB Power)

and

CA9700723

C. Tiron and S. MarinescuRegia Nationala de Electricitate(RENEL),Filiala Centrala Nuclearo Electrica(FCNE),

Cernavoda 1 N.P.S., Cernavoda, 8625, Romania

ABSTRACTEfficient operation of a Candu nuclear powerplant depends greatly on the reliable and safeoperation of the fuel handling system. Successfulcommissioning of the system is obviously a keyaspect of the reliability of the system and thiscoupled with a rigorous training programme forthe fuel handling staff will ensure the system'ssafe operation.This paper describes the philosophy used atCernavoda 1 N.P.S. for the commissioning of thefuel handling systems and for the training of stafffor operation and maintenance of these systems.The paper also reviews the commissioningprogramme, describing the milestones achievedand discussing some of the more interestingtechnical aspects which includes some uniqueRomanian input.In conclusion the paper looks at the organizationof the mature fuel handling department from theoperations, maintenance and technical supportpoints of view and the long term plans for thefuture.

1. INTRODUCTIONIn August 1992 a group of 14 Romanianengineers and technologists arrived at the PointLepreau N.P.S. in New Brunswick, Canada tostart an extensive training programme into theoperation, maintenance and technical support of

the Candu 6 fuel handling system. Thisprogramme, which was 11 months long fortechnical personnel and fifteen months for fieldpersonnel, included hands-on operations andmaintenance training as well as considerablesystems and skills training in the classroom.At about the same time a group of expatriateCanadians from operating CANDU plants inCanada were assembling in Cernavoda, Romaniawith other Romanian staff to form the nucleus ofthe fuel handling commissioning group. Some ofthe expatriate members of this group wereinitially assigned to the construction departmentto provide technical assistance. Working underthe direction of the expatriate fuel handlingsuperintendent this group developed thecommissioning philosophy, prepared the overallcommissioning plan and started on thepreparation of detailed procedures. In November1993 the Romanian trainees returned toCernavoda to take up their assigned positionswithin the fuel handling commissioning group.As well as preparing for system commissioningthe fuel handling group were required to supportthe following associated construction activities:

1) The verification of the Romanian built channelclosure installation equipment.2) The installation of the closure plugs at eachend of the fuel channel.3) The removal of these same plugs for draining

41


and drying of the fuel channels following the hotperformance tests.4) The manual installation of the first fuel loadtogether with the installation of the shield plugsand once again the closure plugs.

2. COMMISSIONING PHILOSOPHYIn order that the Romanian fuel handling staffacquired as much operating and maintenanceexperience as possible it was vital that they bepart of the commissioning process. To this end4 composite teams of Romanian technical andoperational staff were formed, each supervised byan expatriat. Each team would be responsible forcommissioning a number of fuel handling systemstogether with the preparation of the associatedprocedures and other documentation. In the earlystages all team members worked on theproduction of documentation, but as constructionprogressed and turnovers were imminent the fieldtechnicians were released from the technicalgroups and formed into their own operationsgroup to carry out the hands on commissioningwork in the field. Initially the operations groupwas supervised by expatriate with a Romaniandeputy who would later become the supervisor ofthe group.Once the field work was underway the remainingteam members formed the technical unit witheach engineer retaining the same systemsresponsibility to support the commissioning workin the field. As well as leading each team it wasthe responsibility of each expatriate to train thesenior Romanian staff to perform the systemengineering function so that as time progressedthe Romanian staff could take over the systemresponsibility thus ensuring a smooth transition asthe expatriate staff leave the project.To gain even more hands-on experience it wasagreed with the construction department that thepre-commissioning activities would be under thedirection of one of the fuel handling expatriatesusing the Romanian fuel handling staff to performthe field work. Through the experience gainedwhile pre-commissioning, preparing documents,commissioning and operating the hardware andthe training given by the team leaders, thisphilosophy has provided a solid foundation for theRomanian staff in the design, operation andmaintenance of the CANDU 6 fuel handling

system.

3. COMMISSIONING ACTIVITIES3.1 PRESERVATION CHECKSOne of the first tasks to be done was to evaluatethe condition of each of the three fuellingmachines as they had been under a state ofpreservation for over five years. This was muchlonger than the two years storage time expectedwhen the machines were shipped from Canada.For preservation the fuelling machines were filledwith demineralised water and pressurised to10PSI with nitrogen. The complete assemblywas then encapsulated in a foil wrapper whichwas filled with a nitrogen atmosphere.A clean room with temporary lighting and apressurised ventilation system, was established inthe warehouse where the fuelling machines werestored. The cold test facility was moved into theclean room and its oil and water systemscommissioned to provide adequate, if rudimentarypressure sources to operate the fuelling machinesystems.After removal of the foil encapsulation aninspection was made of the external surfaceswhich were found to be in good condition. Somedecomposing of the pipe sealant used on the oilsystem pipe threads was observed but this hadno effect on the capability of the sealant toperform its function. No sign of any oil or waterleaks could be observed.A simple 110VAC electrical test box wasassembled enabling the operation of each oilsystem solenoid valve to be verified. Samples ofthe hydraulic fluid and demineralised water wereobtained for analysis before operation of the fluidsystems. These analyses showed that thehydraulic oil was in good condition but thedemineralised water was in poor condition andrequired changing.The cold test facility was connected to thefuelling machine oil and water systems and inspite of a serious shortage of tubing and fittingsoperation of all aspects of the machine wereverified. At the same time an inspection wasmade of the accessible surfaces of the fuellingmachine and the components, as they passedinto or through the snout. After the inspectionthe snout plug was replaced and the machinerefilled with water to enable a pressure test to be

42


completed.The result of this inspection showed a littlestaining on some of the components otherwisethere was no sign of deterioration on the inside ofthe fuelling machines. The operational checksalso verified that all systems were still operatingcorrectly and hence the three fuelling machineswere considered ready for commissioning withoutany rehabilitation. The fuelling machines werereturned to the proper storage environment untilrequired for installation.The D2O supply pumps for the fuelling machinesalso required inspecting after being in a state ofpreservation for about ten years. The preservationfor this equipment was not satisfactory as thepumps had been shipped and stored with theplungers installed. Although these plungers hadbeen coated with preservation oil, significantcorrosion had damaged the surfaces which hadan adverse effect on the life of the pump packing.Although commissioning of the D2O supplysystem continued with the damaged plungers thefrequent seal failures caused frustration whichwas only solved when the fluid end of the pumpswere rebuilt with new seals, plungers andbushings.

3.2 PRE-COMMISSIONING PROGRAMMEBefore each system was turned over the pre-commissioning team had identified all systemcontrol components, thus identifying deficiencies,completed a wire-by-wire verification of thewiring and cable runs in the system and set upthe MCC for each electrical load. Allinstrumentation devices were calibrated by thepre-commissioning group as part of thisprogramme.

From a mechanical point of view all pressuresystems were hydrostatically tested beforeturnover in a cooperative programme between theconstruction group and the responsible fuelhandling commissioning team. This was also asuccessful programme as it familiarised the fuelhandling team with the layout of each systemprior to the start of commissioning and alsohelped in identifying further deficiencies.

3.3 COMMISSIONING PROGRAMMEA simplified diagram showing the commissioningprogramme is detailed in figure 1 with the

completion date of major activities shown abovethe activity. The programme spans just over threeyears from the assembly of the team to the off-power refuelling demonstration. Some of thecommissioning work was still outstanding at thistime, particularly in the irradiated fuel dischargesystem where testing was held up due to delaysin filling the discharge and storage bays.Following the turnover the commissioning teamcompleted equipment tagging, flow sheet checksand load verification before energising thesystem. Similarly in the field the mechanicalchecks and control valve set-ups were completed,also the electric motors were bumped, whenpermissable, to verify the direction of rotation.Pre-operational checks of the control logic andinstrumentation were being completed at thesame time in preparation for initial operation ofthe system. During the initial operation thesystem operating parameters were establishedand alarms verified. The system was then readyfor the final operational tests in which the systemperformance was verified under all conditions andfully automatic operation established undercomputer control.

The control consoles and termination racks werethe first systems to be turned over and this wasin May 1994. It was necessary to have thesesystems first to ensure that control power wasavailable for all the remaining fuel handlingsystems. Commissioning of the fuelling machinebridges, carriages and catenary systems togetherwith the fuelling machine auxiliary systemsincluding the F/M oil, the F/M air and the F/M D2Osystems followed in logical order. Commissioningon all these systems proceeded relativelysmoothly so that by October 1994 they wereready for the fuelling machines to be mounted ontheir carriages.

Commissioning of the two fuelling machines alsoproceeded very smoothly and by March 1995they were ready for initial operation at therehearsal facility. This facility is a full scale fuelchannel complete with end fittings and shield andclosure plugs. The complete assembly ismounted between the two fuelling machinemaintenance locks. The channel, which can bepressurised to reactor pressure, is supplied withD20 from the fuelling machine systems and thuswith the exception of high coolant flow, the

4 3


rehearsal facility is an accessible fuel channelwhich is readily available for testing and trainingpurposes. The rehearsal facility contains twelvedummy fuel bundles that are identical to theCandu 6 fuel bundles and these are used for allrefuelling operations at this facility. For fuellingmachine testing or operator training eight moredummy fuel bundles are placed in one of thefuelling machines with the other machine leftempty to receive the fuel bundles from thechannel. To compensate for the lack of coolantflow to move the fuel string towards thedownstream fuelling machine it is necessary touse the computer refuelling jobs designated forreactor shutdown operation.It was necessary to have the fuel handlingsoftware in a fully operational state at an earlystage to support the commissioning of the othersystems. To gain experience in the organisationsand operation of this software the fuel handlingstaff assisted the software group in thiscommissioning process. Some changes to thesoftware were required during the commissioningof the individual systems when some smalloperational defects were found.For the initial operation of the fuelling machinesthe refuelling was completed under computercontrol in step by step, mode to allow theoperators to carefully monitor the performance ofeach fuelling machine. As the initial operationwas very successful a second test wasimmediately made, but this time the system wasoperated fully automatically in run mode. Thesuccess of this test proved that the fuellingmachines and auxiliary systems were capable offulfilling their design function.The new fuel system was being commissioned atthe same time as the fuelling machines and wasready for operation by April 1995. This alloweda complete test of the refuelling operations to bewitnessed by the regulators, except for the spentfuel system which was not yet ready.The heat transport system was being drained oflight water at this time, thus it was also aconvenient point to drain and dry the fuellingmachines and auxiliary D2O systems which so farhad been operating with light water. This was atedious process especially for the fuellingmachines and the D2O valve stations with theirmultiple circuits, each with many smaller tubes.

However the operation was successful and thesubsequent D2O fill followed with a minimum ofdowngrading in the fuel handling system.Due to the success of the automatic refuellingoperations a decision was made to complete theoff-power refuelling demonstration during theheat transport system hot performance tests.This was a CANDU first as no other CANDUstation has had their fuel handling system in anoperational state at this stage of stationcommissioning. This is a tribute to the Romaniancommissioning staff as well as to the AECLdesigners.

Following the off-power refuelling demonstrationthe miscellaneous supporting systems such asgrappling and special tooling were commissioned.The grappling system was tested at the rehearsalfacility and during these tests the operation of thefuel grapples were proven as well as the pressuretube seal and latch mechanism. A further test ofthe grappling system proved that a channel couldbe completely defuelled in the shutdown modeusing both fuelling machines and grappleextensions to push the fuel into the downstreammachine.

in August 1995 the commissioning of the spentfuel system was sufficiently far advanced toallow the fuel to be discharged from the offpower refuelling demonstration, but the final testof this system did not take place until December1995. Following this test, which was alsowitnessed by the regulators, the fuel handlingsystem was declared ready for the on-powerrefuelling demonstration scheduled to take placein October 1996.

3.4 COMMISSIONING PROBLEMSAs in any complex system there were someproblems to overcome during the commissioningprocess. Most of these were of a minor natureand were easily solved with a simple mechanicaladjustment or wiring change. One of the morecomplex problems was a pressure instability inthe two Romanian designed oil hydraulic powerunits. The specification called for the use ofpressure compensated piston pumps in theseunits but in this case the improper valve type wasused to operate the compensation system. Aftermany tests and discussions with the power unitmanufacturer the correct type of valve was finally

44

01

w

Oa

HU

iV

iH

CO

5HCO«H

9 3 - 8 6

yCOMMISSION THEo93-87

PREPARE FOR CHANNEL

'CHANNEL CLOSUREINSTALLATION TOOLS

94-85

kCOMMISSION A 8

'CONTROL CONSOLE& TERMINATION RACKS

94-86COMMISSION F/H SOFTWARE

CLOSURE INSTALLATION

94-88^COMMISSION

O93-98

COMPLETE CHANNEL THE

CLOSURE INSTALLATIONo93-89

HEAT TRANSPORT SYSTEM

'F/M BRIDGESÂ & C SYSTEMS

94-89COMMISSION

'F/M CARRIAGEÂ 8 C SYSTEMS

94-89

COMMISSION

AIR HOLD TEST

94-12VERIFY OPERATIONOF THE BRIDGE

o94-12

HEAT TRANSPORT SYSTEM

HOT CONDITIONING

F/M CATENARYA R C SYSTEMS

O OF THE "BRIDGE T /^>kC0MMI SSION A 8 C/^~"N F~~\

CARRIAGE 8 CATE-V^ÂuELLING M A C H I N E V - X I /CARRIAGE|\ NARY, A & C SYSTEMS

9 3 - 8 7

kCOMMISSION COLD/""">VCHECK[T/O) — (T /0

OTEST FACILITY

93-11

CHECK PRESERVATION;

'F/M'S NOS

SERVATIONy*^.

. 1 ,2 & 3V. /

9 4-10TRANSFER F/M'S

9 4-18

COMMISSION

A & C TO RB.OIL HYDRAULIC

PU'S, A & C SYSTEMS

OF F/M 020 PUMPS

94-87yCOMMISSION A 8 C.

[T/O) — [T/0]'F/M AIR SYSTEMS

94-86 94-87VCOMMISSION F/M/—\COMMISSION D20

[T/0'D20 SUPPLY 'VALVE STATIONS

A 8 C SYSTEMSO

94-89CHECK OPERATION

NOTE: DATE SHOWN (E.G. 94-12) ISTHE APPROXIMATE DATE THATTHE ACTIVITY WAS COMPLETED

94-89COMMISSION REHEARSAL FACILITY

'F/M AIR SYSTEM,D20 VALVE STATIONSFOR A 8 C SYSTEMS 8REHEARSAL FACILITY

AND A 8 C SYSTEM SERVICE PORTS E>

E>95-83

INITIAL OPERATION

OF BOTH F/M'S ATREHEARSAL FACILITY

o 95-83CALIBRATION OF

REACTOR LATTICEPOSITIONS. A 8 C

o 95-83CHECK OPERATION OF

a 9S-02PREPARE FOR ANO

RECEIVE THE N/F

95-84

DRAIN & DRY THE

REACTOR PRESSURETUBES 8 FEEDERS

95-83COMMISSION A 8 C

A 8 C N/F SYSTEMSUNDER AUTO OPERATION

95-04

vCOMMISSION A 8 C

oCHECK OPERATION OF

A 8 C S/F SYSTEMSUNDER AUTO OPERATION

o 95-84TEST AUTO REFUELLING

OPERATIONS <» REHEARSALFACILITY

ON-GOINGFUEL HANDLING

6-BN/F SYSTEMS

95-06

COMPLETE THE

S/F SYSTEMS

FIRST FUEL LOADo 95-0 7D20 FILL OF

95-88vHEAT TRANSPORT SYSTEM

OPERATOR TRAINING

HEAT TRANSPORTSYSTEM

E>95-06

FUEL HANDLING SYSTEM

DRAIN AND DRY O95-07

FUEL HANDLING SYSTEM/

FILL WITH D20

'HOT PERFORMANCE TESTS

95-08

OFF POWER, FAF 8 .FARE

REFUELLING DEMO.

95-89

COMMISSION THE

FIGURE I FUEL HANDLING SYSTEM COMMISSIONING PLAN

GRAPPLING SYSTEM

95-11

WET TEST OF THE S/F

SCHEDUL ED 96-10ON POWER, FAF 8

FARE REFUELLING[\ DEMOMSTRATION

END)

DISCHARGE SYSTEM

ON-GO I N<3

PREPARE FUEL HANDLING

OPERATING DOCUMENTATION

FIRST CNS INTERNATIONAL CONFERENCE ON CANDU FUEL HANDLING SYSTEMS, HAY 1996

installed and the problem resolved.A more serious instability problem occurred in theF/M D20 supply system when a loose connectionin the pressure control system resulted in severevibrations in the piping to the high pressurefilters. The resultant pipe movement broke a tubeweld at the filter input. While investigating thefailure it was realised that the tube was tooconstrained and a coiled length of tube was usedto repair the weld to allow freedom of movementof the filter and pipe without straining the tube.The tubing connections on the other filters in thesystem were similarly modified.Although both fuelling machine bridges workedwell, some problems were experienced with thebrakes as they occasionally failed to release. Ifthis failure went undetected by the operator thebrake was very quickly burnt-out and requiredrebuilding. To overcome this problem each brakewas fitted with a limit switch to close on brakerelease and wired as a permissive in the bridgedrive motor control circuit.The Romanian manufactured bridge ballscrewjacks were supplied with inadequate lubrication tothe upper bearings. The jacks operated well inthe short term, but because this equipment is notaccessible for maintenance with the reactor at fullpower it was necessary to request the supplier toredesign the gearbox to provide ample lubricationto all bearings . The eight replacement gearboxeshave since been received and installed.On the electrical side the Romanian panel lightbulbs on the control panels were so unreliablethat it was often difficult to know if a light wason or off, but lit by the warming buss. The bulbswere eventually replaced from Canadian supply.A number of the Romanian designed powersupplies were also unreliable and were notsuitable for use in the specified application. Insome cases a small modification was all that wasrequired to correct the problem, in other casesreplacement units were ordered from Canada.

3.5 SUPPORT ACTIVITIESThe fuel handling group enjoyed a goodrelationship with the construction department andother commissioning groups throughout theproject and lent support in many areas as theneed arose, of particular interest are the activitiesassociated with the heat transport system and

the fuel channels.The first of these activities was the installation ofthe channel closure plugs prior to the heattransport system air hold test. Specialpneumatically actuated tooling had been designedby AECL for this purpose and manufactured inRomania. The operation of the tooling duplicatedthe action and forces of the fuelling machinerams while installing or removing the channelclosure plugs. Commissioning of this tooling wasthe first operation and this took place in June1993. During commissioning of these tools someproblems were experienced wi th themanufacturing tolerances causing difficulties inachieving the correct operating parameters butthese were overcome. The tools performed wellfor this programme but required extensiverebuilding and seal replacement before they couldbe used later for the removal of the closure plugs.This was partly due to the use of an incorrectlubricant that also damaged the seals.In order to complete this programme efficientlywhen most of the fuel handling staff were inCanada, it was necessary to use personnel fromthe construction, operations and maintenancedepartments to form the two teams required todo the work. A training programme was set-upto train the teams on the operation of the tooling,the careful handling of the closure plugs, thecleaning and inspection of the end fittings and theverification of the plugs before and after theirinstallation. To make the training more effectivea spare reactor end fitting was mounted in thetraining area and a special test closure pluginstalled into and removed from this end fitting bythe team members. This was repeated until theywere completely familiar with each operation.A platform was mounted on each fuellingmachine bridge to provide access for the toolingand operators to the reactor end fittings. Due tothe design of the bridge structure it was onlypossible to access the upper sixty percent of theend fittings from this platform and as in previousCANDU projects, facilities were provided to hangthis same platform below the bridge to reach theremaining end fittings. Moving the platforms wasa time consuming operation and interrupted theflow of the work part way through theprogramme. To overcome this problem one of theRomanian senior fuel handling staff requested

46


that the platform structure be redesigned to havetwo platforms mounted on each bridge. Thisdesign called for one platform to be on top of thebridge and the other platform suspended at asuitable distance below the bridge so that theentire reactor face can be accessed withouttransferring the platforms. Due to fabricationdelays the lower platform was not available forthe closure plug installation programme but thisinnovation was to prove very effective when thesame platforms were used for later programmes,when many platform position changes wereeliminated, with the resultant time saving for theproject.

The channel closure installation programme wascarried out in August 1993. It took thirteen daysof twelve hour shifts to complete the workincluding the time lost for transferring theplatforms from the top to the underside of thebridges. During the heat transport system airhold test that followed immediately, six closureplugs were found to be passing slightly. In all buttwo cases the plugs were reseated using manualtooling. The two plugs that could not bereseated were removed and replaced with sparesusing the pneumatic tooling.The second support activity was the draining anddrying of the reactor fuel channels and installationof the downstream shield and closure plugs priorto the first fuel load. This programme utilised thesame pneumatically operated tooling, but thistime the channel draining system built into thetool was also required.

By April 1995 when this work was carried out,the trainees had returned from Canada, but it wasstill necessary to form a similar composite teamas this task was part of the first fuel loadprogramme and required considerably moremanpower than the fuel handling group couldprovide. In order to provide extended twentyfour hour coverage in two, twelve hour shifts itwas necessary to assemble three separate teams.A dummy fuel channel assembly was set up fortraining each of the teams for both the drain anddry and the closure and shield plug installationoperations. For convenience the manual fuelloading equipment was set up in the same areaand training for this operation completed at thesame time.While testing the draining operation it was found

that the there was too much of a flow restrictionwithin the channel closure installation tool whichcaused the draining time for each channel to bemuch too long. As it was not possible to modifythe tool to correct this deficiency, a highcapacity, self emptying wet vacuum unit wasconnected to the tool to increase the flow rateand thus reduce the channel draining time.The reactor fuel channel draining programme wassuccessfully carried out in ten days. Each teamworked in two groups, one at each reactor faceand thus, with a set of equipment at each facethey were able to drain adjacent channelssimultaneously. When all the channels in onerow were drained the two groups would worktogether to pass the drying swabs through eachchannel after which each end fitting would bedrained using the wet vacuum with a specialwand attached. To ensure that all the water wasremoved from the end fitting liner tubes, thewater was trapped in a calibrated flask so thatthe volume retained could be verified. Once theend fitting liner tubes were dried the closure plugregion and the external surfaces of the end fittingwere thoroughly dried and cleaned, and then thedownstream shield and closure plugs installed.The third support activity was the first fuel loadprogramme which followed shortly after the drainand dry programme. This programme alsoincluded the organization of the arrival of the firstfuel shipment. Most of the fuel came fromCanada although some Romanian made fuel wasincluded in the first fuel load to verify thereliability of this fuel.

During the commissioning of the two Romanianbuilt, pneumatically actuated fuel loadingmechanisms, problems were experienced with thespeed control of the actuating cylinder during fuelinsertion. Some redesign of the control systemand control valve modification was required toobtain the correct operating parameters.The training for the first fuel loading concentratedon the handling and inspection of the fuel, theoperation of the manual fuel loading mechanismand the record keeping. The dummy fuel channelassembly was used in the training sessions andeach team operated the fuel loading mechanismmany times to practise loading twelve dummyfuel bundles into the channel.The programme started in late May 1995 and

4 7


was completed in early June. The actual fuelloading process took just under 5 days tocomplete which was another CANDU first. Onceall the fuel was loaded the fuel loading equipmentwas removed from each face and replaced withthe channel closure and shield plug installationtooling. The upstream shield plugs and channelclosure plugs were then installed in the endfittings and the equipment and platforms removedfrom the bridges. Again the success of thisprogramme was largely due to the fact that twoworking platforms were mounted on each bridge.For this programme each team was divided intothree groups, one at each reactor face and asmaller group to provide support to the groups oneach face for materials handling and radiationcontrol. To minimise congestion on the platformsthe fuel was unpacked and inspected at floorlevel before being lifted to the platform.

4.0 TRAINING PHILOSOPHYIn any group of people there will always be arange of age, experience, technical capability anddexterity even though all the people would havebeen selected to perform specific functions witha common interest. This fuel handling group isno exception and thus it is the job of the fuelhandling training officer to arrange training forindividuals at different levels so that each personcan progress to the maximum of their ability.From a training point of view the fuel handlingoperations group falls into three categories. Thefirst category is for the six operations staff thatwere trained in Canada (three others had beenpromoted) and hence able to take lead roles incommissioning activities. The second category isfor the eight people that remained in Romaniawhile their colleagues were in Canada, and thusonly qualified to take supporting roles in thecommissioning activities. The third category isfor the five people that were more recently hiredand had minimal or no prior exposure to fuelhandling operations. The training programme wasset up to meet the needs of all three groups whilealso having to meet the departments objective ofhaving sufficient fully trained staff ready to safelyoperate the system by the time the first refuellingis required. At the same time the department stillhas to complete the day to day operation andmaintenance of the fuel handling system.

The training required for the first category wasone of refreshment of systems training in theclassroom with extensive fuelling machineoperations training at the control panel. For thesecond category more formal systems trainingwas required both in the classroom and in thefield, to be followed at a later date with trainingat the control panel. For the third category indepth systems training will be given in the field,together with skills training, to be followed laterwith formal classroom training. When sufficientknowledge and experience has been gained thecontrol panel training will commence. It shouldbe noted that in each category the examinationshave an 80% pass mark Once the control paneltraining is complete a final very comprehensiveexamination has to be passed before the traineecan be given interim authorization to refuel thereactor and then only as a co-pilot for twentyrefuelling operations. At this point a practicalexamination with oral questions is given at thecontrol panel by the fuel handling superintendentand the operations supervisor. If this examinationis passed then the trainee is classified as afuelling specialist, and fully authorised to refuelthe reactor. All fuelling machine operationstraining takes place with the machines on therehearsal facility, usually under expatriatesupervision and once reactor refuellingcommences the co-piloting requirement for thecategory one senior specialists will be supervisedby the expatriate staff. Once fully authorised andexperienced the senior fuelling specialists willsupervise the co-piloting of the fuelling specialistswhen they are qualified.

The formal training programme started in October1995 after the intensive commissioning periodwas over. To avoid a minimum of disruption tothe day to day operations the operating groupwas divided into two teams, each with its ownsupervisor and leaders. One team was scheduledto be the operations crew for the week lookingafter the routine operation and maintenance ofthe system. They were also performing refuellingoperations at the rehearsal facility and completingfield check outs. At the same time the otherteam spent the week in the training centre takingthe extensive fuel handling systems training. Thefollowing week the two teams exchanged roles.For the team at the training centre the first two

4 8


days of the week were spent attending lectureson a specific system given by the responsiblesystem engineer for that system. On the thirdmorning a comprehensive examination waswritten after which a second course would begiven, with another examination written on thefollowing Friday afternoon. The second teamwould take the same two courses the followingweek. The examination was only written by thecategory one and two students, the categorythree students attended the lectures as anintroduction to the systems, they will study themin depth later. This phase of the training wascompleted by mid-March 1996 (at the time thispaper is being written). For the category onetrainees the next phase of training will beconcentrated on panel operations. For thecategory two and three trainees the emphasis willbe on field operations training with oral fieldcheck outs for each system given by theoperations supervisor. Coincident with thistraining some classroom training will occasionallybe given to category one and two trainees onspecial subjects such usage of the operatingmanuals.

At the end of April the category two trainees willwrite a comprehensive examination covering allfuel handling systems. The successful candidateswill be classified as fuel handling field specialistsand authorised to carry out all the field operationsassociated with fuel handling. The moreexperienced field specialists start their controlpanel training programme as time permits. Atthis same time the category three trainees will beclassified as fuel handling specialists-in-trainingand commence skills and formal systems training.In mid-June the category one trainees will writea comprehensive examination on the operation ofthe fuel handling system from the control panels.After this a practical examination with oralquestions will be given at the control panel by thefuel handling superintendent and the operationssupervisor. The successful candidates will thenbe classified as fuelling specialists authorised foron-reactor fuelling machine operations underexpatriate supervision. Once reactor refuellingcommences and their co-piloting is complete thisgroup will be authorised for on reactor refuellingand the more suitable will be selected as seniorfuelling specialist

It should be mentioned that throughout thecommissioning process the field personnel weregaining experience in all disciplines, is wasdeliberately organised this way as it is necessaryfor fuel handling personnel to be cross trained inall disciplines. In the fuel handling group allpersonnel are expected to be able to fulfil themechanical, control, electrical and operationsfunctions at any time although some technicianswill take lead roles in certain disciplines based ontheir backgrounds and experience.As well as the maintenance training received inCanada, training for maintenance activities hasbeen an on-going process throughout thecommissioning phase. Usually this training hasbeen given by the expatriate staff and to date thishas been very satisfactory and most maintenanceactivities can now be competently carried outwith the Romanian field crews working with themaintenance procedures that are now in place. Insome cases the Romanian systems engineersprovide additional support to the maintenanceactivities on their own systems. Certain specialprocedures such as changing the fuelling machineram assembly, rebuilding the F/M snout assemblyand rebuilding a ram assembly do still requirespecial training which is scheduled to take placesoon.By the time the reactor will be ready for firstrefuelling this programme will have trained sixfuelling specialists ready for control paneloperation and eight fuel handling field specialiststo give support to the fuelling activities. The fieldspecialists will also be supported by thespecialists-in-training and this team will be wellable to operate and maintain the fuel handlingsystem at Cernavoda 1 IMPS

5.0 THE FUTUREThe immediate future for the fuel handling groupis basically to continue the training programme aswell as maintaining the system in readiness tosupport any on reactor operations that might berequired during phase B and C commissioning andto prepare for the commencement of reactorrefuelling. As the time for reactor refuellingapproaches the operations group will beorganised into two crews each working eighthour shifts to cover sixteen hours operation forfive days per week. Each shift will have two

49


senior fuelling specialists, two fuelling specialists,two fuel handling field specialists and one ormore fuelling specialists-in-training. The fuelhandling operations supervisor will not be on shiftand will not be directly involved with the day today operations of the crew, this will be theresponsibility of the two senior specialists. Thefuel handling control panel will be manned by asenior fuelling specialist assisted by a fuellingspecialist. From an operations point of view thesetwo men are working under the direction of theshift supervisor.Work protection will be provided through thework permit system and in the case ofmaintenance work on fuel handling systems thework protection and work permits will be issuedby the senior fuelling specialist at the controlpanel, after approval by the main control roomoperator. All electrical and mechanical isolationsin the field will be carried out by the field fuellingspecialists.The rest of the crew will be working in the fieldunder the direction of the second senior fuellingspecialist. Their task will be to support thefuelling operations and to complete themaintenance activities.The crews are scheduled to work the day shiftone week and the evening shift the followingweek. After this two week cycle the crewsrotate the job functions so that those working onthe control panel assume field responsibilitieswhile the senior and one fuelling specialist fromthe field take over the control panelresponsibilities. It can be seen that over a fewshift cycles all qualified personnel will spend timeon maintenance or field operations and on controlpanel operations thus forming a highlyexperienced and specialised team to operate andmaintain the fuel handling system. This teamapproach will add greatly to the system reliabilityas all qualified members of the crews have agreater understanding of the working of thesystems. Thus they can quickly respond tofailures, also they can more easily diagnoseproblems and often anticipate problems beforethey become more serious, which will ultimatelybe reflected in the overall efficiency of thestation.In the long term the group will be striving toimprove their skills in all areas and to improve

the maintenance operations through thedevelopment of special tools and fixtures. It ishoped that the inadequate fuel handlingmaintenance area will be replaced by a purposebuilt facility that will have sufficient space torebuild F/M rams and fuelling machines at thesame time, as well as providing space for specialtest facilities.The fuel handling technical unit as well assupporting the normal system operations will alsobe looking to the future to see how they canimprove the efficiency and reliability of theirsystems. Replacement of obsolete equipment isone of their priorities, this particularly applies tothe oil hydraulic valves most of which are alreadyout of production with spare parts no longeravailable. They also have to prepare for outagework and to support reactor maintenanceprogrammes such as CIGAR, scrape sampling andsingle fuel channel replacement. Looking furtherinto the future they must be giving thought toexpanding the irradiated fuel storage facilityincluding the possibility of dry fuel storage.Hopefully the philosophies presented here, whensuccessfully applied, will make the fuel handlinggroup at Cernavoda 1 NPS rank amongst the bestin the CANDU group of nuclear power stations.

Acknowledgement: The authors wish toacknowledge the significant contribution to thisproject provided by Mr. Jim Georgas of AECLwho was the fuel handling superintendentthrough most of this commissioning programme.

5 0


FUEL HANDLING SOLUTIONS TO POWER PULSE AT BRUCE NGS A

R. C. DAYONTARIO HYDRO

BRUCE NUCLEAR GENERATING STATION APROJECTS AND MODIFICATIONS GROUP

P.O. BOX 1000 • TIVERTON, ONTARIO, CANADA • NOG 2T0TEL. (519) 361-2673FAX (519) 361-5578

~ CA9700724

ABSTRACTIn response to the discovery of the power pulseproblem in March of 1993, Bruce A has installedflow straightening shield plugs in the inner zonechannels of all units to partially reduce the gapand gain an increase in reactor power to 75% FP.After review and evaluation of solutions tomanage the gap, including creep compensatorsand long fuel bundles, efforts have focused on adifferent solution involving reordering the fuelbundles to reverse the burnup profile. Thisconfiguration is maintained by fuelling with theflow and providing better support to the highlyirradiated downstream fuel bundles by changingthe design of the outlet shield plug. Engineeringchanges to the fuel handling control system andoutlet shield plug are planned to be implementedstarting in June 1996 thereby eliminating thepower pulse problem and restrictions on reactoroperating power.

1. INTRODUCTION1.1 Fuel String Relocation EffectFuel channels in CANDU 850 MW reactors eachhave a 13 bundle fuel string which is supportedagainst the flow at the downstream or channeloutlet by a latch. The latch consists of four spring

fingers which contact the ends of severalperipheral fuel bundle elements comprising thefuel bundle. On-power fuelling is achieved bypushing in fresh fuel bundles against the flowthrough this latch into the fuel channel,displacing spent fuel out into the upstreamfuelling machine.

Initially, the fuel bundle at the inlet end of thefuel channel is partially supported within the skirtof the inlet shield plug. In time, as the pressuretube creeps and elongates, the inlet fuel bundleeventually rests fully in the pressure tube,widening the gap between it and the inlet shieldplug. This gap can be as large a 15 cm (6 in).

One of the design basis safety analysis accidentswhich sets the most demanding requirements onspeed and reactivity of the shutdown systems isthe large break loss of coolant accident (LOCA).A LOCA can lead to a rapid power increase dueto positive reactivity arising from draining ofcoolant from the fuel channels. This is followedby a rapid power reduction once the trip initiatesand the shutoff rods and/or poison injection areactivated.

51


A postulated failure of a reactor inlet headercauses a reversal in the channel pressure drop,simultaneously driving the fuel bundles in a largenumber of channels against the inlet shield plugall in a fraction of a second. This relocation of thefuel bundles to the inlet end introduces positivereactivity as fresher more reactive fuel replacesmore irradiated fuel in the central high fluxregion. The reactivity is proportional to the size ofthe gap or the relocation distance. Therefore, inaddition to the positive reactivity introduced bycoolant voiding, the relocation affect causes the"power pulse" to increase in magnitude. Withoutfurther confirmation and modeling, theeffectiveness of the shutdown systems could onlybe guaranteed at lower operating power levels.

1.2 Discovery of Power PulseMethods of preventing pressure tube fretting atthe inlet rolled joint burnish mark was the subjectof investigation in the late 70's and early 80's.Studies were initiated to investigate theconsequences of removal of the 13th fuel bundle.Consideration was given to the higher reverseimpact velocity effect on the integrity of fuelbundles, fuel channels and the calandriastructure and limited reactor physics implications.Reverse flow bundle acceleration tests done in1989, confirmed the potential for rapid fuel stringmovement. A proposal was made in 1992-1993 toremove the 13th inlet fuel bundle at Darlington toaddress abnormal fuel support fretting. Inresponse to an AECB question regarding thereactivity effect, analysis of a large break LOCAcaused by a postulated inlet header failure forthis configuration revealed the significance of thepower pulse for a 12 bundle fuel channel.Subsequently, the analysis was extended toaddress current gaps and found to be a concernfor the operating reactors. Until further designreview and analysis was completed, Ontario

Hydro took a conservative approach where allCANDU 850 MW units, including Bruce A werederated to 60%FP to re-establish acceptableshutdown system margins. Differences in reactorcore physics in Bruce B, allowed an upratingback to 80%FP at that station. At Darlington, thereactivity increment was accommodated withinavailable safety margins and the derating wassubsequently removed. One contributing factorfor Darlington is the dual figure of eight loopprimary heat transport system configurationwhich limits the number of channels affected by apostulated inlet header failure to 120 channels asopposed to 240 channels at Bruce A and B.Another factor of significance is the small gap inDarlington fuel channels due to the short P/Tcreep life to date.

2. SOLUTIONS2.1 BackgroundTo eliminate the postulated "power pulse"concern and return the derated reactors back totheir high power levels of operation, amultidisciplinary team was setup to define thebest alternate design, safety and operationalsolutions.

Solutions fell into two basic categories; those thatserved to reduce and/or manage the size of thegap and those that eliminated the insertion ofpositive reactivity. After evaluation of a number ofpotential solutions, several options were pursuedin parallel. Gap management has been pursuedat all three stations, however because of itsuniqueness, Bruce A has adopted a separatelong term approach.

52


3 . GAP MANAGEMENTTo quickly reduce the gap and allow acommensurate increase in power, two gapmanagement solutions, creep compensators(CC) and flow straightening inlet shield plugs(FSISPs) were considered.

3.1 Creep Compensators3.1.1 DesignInstallation of a cylindrical spacer in thedownstream latch serves to relocate the fuelstring upstream and reduce the size of the gap atthe inlet. The spacer or "creep compensator"varies in length by 1.3 cm (0.5 in) increments,reducing the inlet gap from 3.8 cm (1.5 in) to asmuch as 15.2 cm (6 in) depending on the amountof creep in the particular fuel channel. Theincreasing gap size caused by pressure tubecreep could be compensated for by installing anincrementally larger size of creep compensator.The cylindrical CC is held in the downstreamlatch fingers by a shoulder on the outer diameter.The upstream end is shaped to match the outertwo rings of a fuel bundle end plate and reactedthe 6675 N (1500 Ibf) hydraulic load of the fuelstring through the downstream end plate of the1st fuel bundle. The coolant flow goes downthrough the center or through 10 angled ellipticalholes in the wall of the cylinder near theupstream end. The pressure drop across this endof the fuel string actually reduces slightly with theinstallation of a creep compensator because theflow is effectively redirected away from the latchfingers.

3.1.2 InstallationThe creep compensator installation tool occupiesa magazine position and functions to install andremove the CC from the downstream fuelchannel latch. It consists of a fuel carrier, which

is picked up by the fuelling machine ram andcharge tube assembly, and is capable of openingthe channel latch. Within the carrier is a springloaded plunger assembly which interfaces withthe ram. The front of the plunger is fitted withspring loaded fingers which engage with the CC.During removal of the CC, the ram and plungerare advanced, engaging the fingers with the CCand pushing the fuel string upstream off thelatch. The charge tube advances the carrier,opening the latch. The ram is then retracted to agiven position. The carrier is then retracted,closing the latch onto a taper on the upstreamend of the CC. Retracting the ram then allows thelatch to close and support the fuel string.Installing the CC, simply involves advancing theram and plunger with the CC attached, until thelatch fingers close into the groove on the outsideof the CC. The ram is then retracted, wherebythe spring loaded fingers engaged with the CCare pulled free from the CC by the larger springforce of the plunger. The removal and installationactivities requires approximately 10 minutes eachto complete.

3.1.3 OutcomeThe viability of this concept was jeopardized bythe additional 20 minutes of fuel handling timerequired to remove and reinstall the CC for eachchannel fuelled and the loss of one magazineposition. Fuel handling time was also required toreplace CC's with longer ones to account for P/Tcreep. This 24% reduction in capacity could notbe tolerated by the two trolley fuelling system atBruce A when operating at high reactor powers.The third, or south extension trolley has beenremoved from fuelling service and dedicated tothe fuel channel spacer location andrepositioning (SLAR) rehabilitation program.Consequently, this solution was abandoned atBruce A and at other stations.

53


3.2 Flow Straightening Inlet Shield Plugs3.2.1 DesignThe existing fuel channel inlet shield plug(MKIIIA) was modified to a MKIIID flowstraightening inlet shield plug by crimping a flowstraightener in the shield plug skirt. The primaryfunction of the shield plug is to provide radiationattenuation. It is removed during fuellingoperations and otherwise remains engaged withthe liner locking lug in the fuel channel. The flowstraightener is a solid disc made of low cobaltmodular ductile iron, identical to the shield plugmaterial, 10.3 cm (4 in) in diameter and 4 cm(1.6 in) thick, with 121 flow holes, 6.3 mm (0.25in) in diameter (figure 1). This insert provides aflow straightening function which has beenproven to reduce flow turbulence and fuel bundleinlet frett ing1. It also provides a one timereduction of 4 cm (1.6 in) in the size of the gap.

FLOWSTRAIGHTENER

Figure 1: Flow straightening Inlet shield plug.

3.2.2 Installation and CommissioningOperationsCrimping machines used to install and crimp theinserts into the shield plugs were installed andcommissioned on the fuel bay gantry.

The FSISP crimping campaign was initiated bytaking a dedicated trolley to the target unit,where each fuelling machine (F/M) head installed8 new inlet shield plugs complete with crimpedinsert and removed 8 regular inlet shield plugs,taking approximately 1 hour per channel visit.These 16 new FSISPs were removed at the endof the campaign and stored for use on thefollowing unit and finally sent for disposal. Onecontrol room fuel handling operator operated thededicated trolley. Two F/M heads were lockedonto the ends of the channel for each inlet shieldplug removed or installed in order to perform apressure drop measurement ensuring no flowblockage. Channel outlet temperatures were alsomonitored for the same purpose. The trolley wasthen taken to the central service area and theF/M heads locked onto the east and westancillary ports. On the bay side, one crimpingmachine was aligned with one of the ports andprepared to receive the shield plugs. On average,2 field operators worked in the fuel bay area,operating the crimping machines. The fuellingmachine ram and charge tube assembly pickedup a shield plug and pushed it through the port tothe crimping machine. An flow straightenerinsert, cooled in liquid nitrogen, was thenremotely installed into the skirt and 3 crimpswere applied to the outside of the skirt. The flowstraightener was therefore secured by thecombination of the shrink fit and the crimps. Theplug was then retrieved back into the fuellingmachine head and the process was repeated forthe next and remaining shield plugs in that head.

54


When all 8 plugs were modified, operations werecompleted for the shield plugs in the oppositeF/M head. The trolley then returned to the unitwith each of the 2 F/M heads carrying 8 modifiedshield plugs which were then swapped for regularshield plugs. The process was repeated until all280 inner zone fuel channel inlet shield plugswere modified.

3.2.3 OutcomeEarly in the installation campaign on the firsttargeted unit, after approximately 1 week ofoperations, the crimp depth on test coupons (1test coupon every 8 shield plugs) began towander out of specification. A delay of 2 monthswas incurred resolving this problem.

Typically, the process of installing 16 FSISPs andremoving 16 regular shield plugs took 16 hours.The fuel bay crimping operations for the 16shield plugs took approximately 4 hours. Onaverage, accounting for shift change and breaks,the rate of processing was 80 shield plugs perweek. The total effort (engineering andoperations) associated with conversion of 4 unitsrequired approximately 12,000 man-hours with adose expenditure of 1.97 rem.

Flow Straightening Inlet Shield Plugs installed in

the 280 inner zone channels of all 4 units atBruce A permitted an increase in reactor powerfrom 60% FP to 70% FP. Subsequent additionalsafety analysis permitted a further increase to75%FP. Table 1 shows the dates for completion ofinstallation of FSISPs and subsequent increasesin reactor power.

3.3 Long Fuel BundlesTo eliminate the "power pulse" concern, two longterm solutions; long fuel bundles and fuelling withthe flow (FWF) were devised. Bruce B opted forthe long fuel bundle solution, thereby controllingthe gap, while Bruce A chose to eliminate theproblem by adopting FWF. Lower channel flowrates made the FWF option undesirable atDarlington.

3.3.1 DesignLoading a fuel channel with a combination ofregular fuel bundles and bundles that are 1.3 cm(0.5 in) longer than regular fuel bundles (longfuel bundles) results in a fuel string length thatcan be managed to minimize the inlet gap andposition the 13th bundle bearing pads away fromthe burnish mark. This solution thereforeaddresses the power pulse problem as well aspreventing bearing pad interaction with thehighest stress area of the pressure tube. The

UnitFSISP Installation

Complete Date

Reactor Power to

70%FP

Reactor Power to

75%FP

3

January 1994

March 1994

October 1994

1

February 1994

March 1994

October 1994

2

March 1994

April 1994

October 1994

4

July 1994

August 1994

October 1994

Table 1 • Flow Straightening Inlet Shield Plug Installation and Power Increase Dates

55


design of the long fuel bundle is identical to the

regular fuel bundle except the distance between

the center bearing pad and the intermediate

bearing pad on each side of the bundle center

line is increased by 6.3 mm (0.25 in). This design

has been tested in-reactor and approved for use.

3.3.2 Installation

Although it is desirable to minimize the average

core gap from a power pulse perspective, it is

also necessary to ensure that the gap in any one

fuel channel does not get smaller than a

specified minimum gap size. A postulated large

break LOCA causes the fuel string to expand and

therefore reduce the gap. If the gap is too small

prior to the event, the fuel string expands into

compression which may lead to pressure tube

failure. The gap management system must

therefore continually function with high reliability

to ensure that the average core gap is

maintained low enough to prevent power pulse

yet not exceed the allowable minimum gap size in

any one fuel channel. Channels are fuelled with

sufficient long fuel bundles to prevent fretting on

the burnish mark and reduce the gap at the inlet.

During the normal course of fuelling, spent long

fuel bundles removed from the channel must be

replaced and additional long bundles installed to

account for pressure tube creep. However at the

same time, the inlet gap in each fuel channel

must be maintained greater than the minimum

allowable gap size. This management system

impacts a number of work group activities i.e.

new fuel loading, spent fuel storage, generation

of fuel change orders, fuel handling.

3.3.3 Outcome

Pressure tube fretting at Bruce A has been

identified as a concern, however, levels of fretting

are more severe at sister stations. The

installation of the FSISPs in the inner zones of all

units will help to minimize fretting damage.

The core physics at Bruce A is different than

sister stations. The unique peaked neutron flux

makes the response to fuelling operations more

difficult to simulate and reduces the ability to

accurately predict in advance, fuelling sequences

long enough to allow for the required control of

long fuel bundle fuelling.

Also, the requirement for a minimum gap has

great potential to limit the maximum achievable

power output from the Bruce A units. A minimum

gap size 3.2 cm (1.25 in) would be required to

achieve 92%RP which is the full electrical

equivalent of the Bruce A units. The currently

approved minimum gap size is 5.1 cm (2 in).

This gap management solution utilizing long fuel

bundles was not pursued at Bruce A. However,

Bruce A has pursued a contingency plan

involving the loading of one long fuel bundle at

one time into all inner zone channels to gain a

temporary reduction in average core gap without

violating the minimum gap requirement of 5.1 cm

(2 in). This would permit an increase in reactor

power to 80%FP for a period of time. Although

approval has been received to proceed with this

program, Bruce A has chosen not to implement it

at this time.

4. REVERSAL OF FUEL BURNUPPROFILE4.1 Fuelling With The Flow

Changing the direction of fuelling from against

the flow to with the flow, can be used to maintain

a reversal of the fuel burnup profile. The most

irradiated or spent fuel bundle would reside at

the outlet end of the fuel channel and the new

56


fuel at the inlet. Therefore, given the postulatedinlet header failure, the relocation of the fuelstring would introduce negative reactivity asirradiated fuel would replace newer, morereactive fuel in the central high flux region. Thesize of the inlet gap would no longer be ofconcern and the remaining power pulse causedby voiding could be negated by the shutdownsystems. There has also been some indicationthat having new fuel at the inlet of the channeldoes not cause fretting damage to the sameextent as highly irradiated fuel2. After reviewingthe alternatives and performing business caseevaluations, Bruce A has committed to resolvethe power pulse problem by pursuing reorderingthe fuel bundles and FWF.

4.1.1 DesignPrior to implementing FWF, it is necessary toreverse the order of the fuel bundles in the fuelchannels. Without doing this, the most irradiatedfuel would be pushed back through the core,causing a loss in reactivity and subsequentshutdown. In addition, new fuel string supportingshield plugs were designed to prevent fuelbundle end plate cracking during a hot shutdownonce the channels have been reordered. Theseare discussed in following sections.

Software design changes to the fuel handlingcontrol system, namely the controller software,protective software, operations and sequences,control console, termination cabinet and wiringare required to implement a fully automaticoperation of the FWF process.

Control software changes have been made topermit synchronization of the rams based onencoder inputs. A channel status map iscontained within the control software, to record

the state of various channel parameters. Thestatus for each fuel channel is updated aschanges are made such as new shield plugsinstalled, channel reordered and channel readyfor FWF. Also the map is automaticallyinterrogated prior to a change being made, inorder to verify that the change is appropriate.This verification step serves as one of at leasttwo barriers to performing erroneous operations.The fuelling machine ram encoder signals havebeen directly wired from one control 'quadrant' tothe controller of the other fuelling machine on thetrolley, to provide better ability to track the rammotion of the other machine for FWF.Additionally, signals have been wired betweenthe systems which indicate status of ram brakesand clutches and facilitate ram and fuel positionsensing. Power supply and brake release logicchanges have been made to ensure the brakeswill hold the fuel string during a Class III or IVpower failure.

New outputs have been assigned to commandthe redundant brakes added for fuelling machineram and charge tube axial and rotary drives.These F/M signals are used in the control andprotective logic to prevent applying excessiveforces to the fuel string. Protective softwarechanges include additional interlocks to ensurelatch operation and the fuel is not damaged. Aseparate set of protective logic has beenprovided for 'abnormal' or maintenanceoperations.

The operating data includes new sequences andoperations that define FWF operation. Thesequences for fuelling against flow have beenretained since during the transition to FWF, bothmodes of fuelling will be required.

57


Console panels have been modified to add theprotective normal and abnormal indications andthe redundant ram, charge tube axial, and chargetube rotary brakes. Interlock bypass switcheshave been allocated for FWF and charge tuberotary interlocks. All operations executed by thecontrol computer, as well as those executedmanually while the controller is active will belogged. This information will be used forconfirmation of proper completion of operation ifmanual intervention becomes necessary.

The fuelling carriers required modification topermit them to open the latches and a taper isrequired on the back of the ram head to allow itto return through the latches.

Basically all design work was done by GeneralElectric Canada and commissioning and testingdone by Bruce A Fuel Handling Operations.Bruce A Engineering Services Departmentfunctioned as the design authority.

4.1.2 Installation, Commissioning andTestingIn order to permit installation and commissioningof FWF without adversely affecting the north andsouth trolleys from normal fuelling or to adverselyaffect fuel handling outage support work severalinitial changes were required. The purpose ofthese changes was to have the trolley quadrantsconverted to a state where changing from FWFcommissioning mode over to regular fuellingmode on a frequent basis could be doneexpeditiously. These changes were declared in-service in May 1995.

Commissioning work plans were prepared basedon GE commissioning test specifications by FuelHandling Technical. The work plans were

implemented by the Fuel Handling Operationsshift crews. Completed commissioning workplans were reviewed by Fuel Handling Technicaland GE.

Due to the demands on Fuel Handling systemsand resources to provide regular fuelling for 4units at 75%FP with 2 trolleys, subsequentlyreduced to 3 units, support outage work such asdefuelling channels for CIGAR inspections andmini-SLAR operations, supply fuel handlingoperators to SLAR operations, defuel unit 2 andperform preventative and forced systemmaintenance, progress on commissioning FWFsoftware has been delayed. A total ofapproximately 13,000 man-hours of engineeringand operations labor have been used to date toperform FWF commissioning related activitieswhich commenced in late 1994.

4.1.3 Status and PlansA full set of interlock checks and follow-ups havebeen completed on the south trolley. The northtrolley system interlock checks are scheduled forMay 1996. Controller software checks, follow-upsand tests have been completed. North and southtrolley protective checks have also beencompleted.

The FWF control and hardware changes weretested on the Maintenance Area Test Facility andthe Service Area Rehearsal Facility. Thesefacilities provide the necessary mockups to testthe interface between the fuel handling systemand fuel channel components. An on-reactor testis also planned. Two outer zone channels will befuelled with new fuel (done to reducecontamination if fuel damage occurs) and FWFoperations will be fully tested.

58


Hardware modifications to the fuel handling

carriers and ram heads were completed at site

with components designed and supplied by GE.

Fuel Handling operations staff have been trained.Operating documentation has been revised. TheOperations ECN walkdown and design authorityapproval is expected to be completed in thesecond quarter. Submissions have been made tothe AECB and approval is expected prior to theputting FWF in-service by the end of the secondquarter.

4.2 ReorderingReordering reverses the burn-up profile in eachchannel so at the next fuelling visit, the channelcan be fuelled with the flow. Bruce A undertookextensive design and safety analysis to supportthe selected reorder scheme. In addition, areorder trial was performed to test the selectedreorder process.

4.2.1 DesignThe selected reorder scheme is a 12 bundlecascade which is done against the flow. Theprocess is started by displacing 12 fuel bundlesin the seed channel with new and depleted fuel.The fuel from the seed channel is then loadedinto a neighboring channel, in order, with leastirradiated first to most irradiated last. Thedisplaced fuel is then loaded into the nextchannel and the cascade continues forapproximately 14 channels. The final channelloaded is the original seed channel. Thedisplaced new and depleted bundles are thenused to fuel for reactivity gain as usual. TheBruce A Fuel and Physics section areresponsible to conduct SORO simulationspredicting maximum channel and bundle powersduring the reorder push of the selected channelsto be reordered.

4.2.2 TestingReordering of fuel bundles using the 12 bundlecascade scheme while at power was analyzed toshow that the degree of power ramping of highburnup bundles would not lead to increased fueldefects. However, a reorder test was done toverify the conditions and parameters under whichproduction reordering will be performed. InAugust of 1995, fuel bundles from 12 channelscontaining F3SPs were successfully reordered inunit 4. The conclusions were that the testconfirmed predictions and verified limitspertaining to the reactor regulating system,maximum channel power peaking factor,maximum channel power, neutron overpowerprotection and fuel performance. A FWF ReorderSpecification has been prepared incorporatingtest results which specifically describes theconditions and parameters under which reorderwill be safely performed.

4.2.3 PlansIt is planned to reorder the 280 inner zonechannels on each unit at a power level ofapproximately 70%FP before returning tocomplete the 200 outer zone channels for safetyand economic reasons. Based on the reorder trialon unit 4, it is estimated that the reordering ofone unit will take approximately 9 weeks. Adedicated fuel handling trolley will be used toreorder batches of approximately 14 channels ata time, with about 1 to 2 days of regular fuellingbetween batches. The fuelling for reactivity gainbetween reorder batches will be either fuellingagainst the flow or fuelling with the flow.Therefore, reordering will not commence untilFWF has been declared in-service and new fuelstring supporting shield plugs installed.

59


4.3 Fuel String Supporting Shield Plugs4.3.1 DesignThe most significant impact of reordering andFWF, is the relocation of the most irradiated fuelbundle from the inlet to the outlet. The outletbundle is supported by the latch fingers reactingagainst the outer elements of the mostdownstream fuel bundle and is subjected to thehighest hydraulic load (6675 N). Although fuelbundle end plate cracking due to PHT pumppressure oscillations had not been seen in BruceA reactors, a test was done to confirm theintegrity of highly irradiated bundle end plates inthis new configuration. In the fourth quarter of1993, 12 fuel bundles were positioned in thereverse flow direction into fuel channel position#1 after cycling through the reactor in Unit 2 andleft there for 7 weeks of operation. They wereremoved from the core after a hot shutdown.External visual inspection of the end platesshowed no through wall cracking. However, ashort time after the completion of the designreview for FWF, scientists at Ontario HydroTechnologies warned against the potential fordelayed hydride cracking in the highly stressedfuel bundle end plate during a hot shutdown. Atotal of 6 fuel bundles from the test and 2 controlbundles were sent to Chalk River Laboratoriesfor destructive examination. The 6 bundlesdisplayed numerous cracks, up to severalmillimeters into the spigot to end plate weld. As aresult, a change to the design of the outlet shieldplug was made to fully support the outlet fuelbundle end plate, reduce the stress and preventcracking.

A nosepiece was designed to be added to thefront of the body of the existing outlet shield plugto fully support the end plate of a fuel bundleagainst the hydraulic load on the fuel string

(figure 2 see next page). The 410 stainless steelnosepiece is a cylindrical arrangement which fitsover the end of the existing cast iron shield plugbody, increasing its length by approximately 7.6cm (3 in). It is attached to the body by a specialbolt which limits the nosepiece to rotation on athrust bearing. The back of the nosepiece iskeyed to the shield plug liner locking lug duringinstallation and removal thus preventing relativemotion between the nosepiece and the fuelbundle. A formed leaf spring in the back of thenosepiece provides a detent between thenosepiece and the shield plug body. This ensuresthat nosepiece orientation, relative to the shieldplug, will be correct during installation.

New shield plugs have been manufactured anddelivered to site. It is planned to replace theexisting outlet shield plugs with the new F3SPs.A shield plug handling system designed by GE,has been installed and commissioned in the fuelbay. This system is used to handle and load theF3SPs and old shield plugs. Modules, eachholding 8 F3SPs were designed fortransportation, storage and handling. The moduleconsists of 8 horizontally arranged tubes whichare welded to a handling frame. Locating lugs ineach tube serve to control the orientation of theF3SP which facilitates pickup by the fuellingmachine ram and charge tube assembly.

The modules are handled in the fuel bay by amodule handling tool which is suspended fromthe fuel bay crane. The handling tool providesshielding and functions to attach to the moduleframe and rotate in a horizontal plane. The fuelbay crane has been modified to permit remoteinfrared control to reduce radiation dose to theoperator. Modules are lifted by the handling toolfrom the 615' elevation up to positioning

60


SHIELD PLUG

NOSEPIECE

SHIELD PLUG

SHIELD PLUGLOCKING LUG

FUEL LATCH

FUEL BUNDLE

NOSEPIECI(PREVENTED FROM R .THE SHIELD PLUG LOCKING LUG)LPR_EyENTED FROM ROTATING BY CHANNEL LINER

SHIELD PLUGSHIELD PLUGLOCKING LUG

06 MIN. FUEL PUSH15 MAX. FUEL PUS

Figure 2: Fuel string supporting shield plug.

mechanisms, mounted on each end of the fuelbay gantry. The positioning mechanism supportsand shields a module and positions the moduletubes with respect to the ancillary port. Thepositioning mechanisms are also operated byremote infrared controllers.

The fuelling machine exchanges old shield plugsfor new F3SPs at the ancillary ports. Themodules containing the old shield plugs are

loaded into shielded storage containers locatedon the 615' elevation. Each container holds 14modules (112 shield plugs) and when full will betransported to the Radioactive Waste OperationsSite (RWOS) for storage in concrete trenches.Lifting beams have been provided to handle thecontainers by both the fuel bay crane and thecrane used at the RWOS. The module handlingtool is also used to handle the shielded lids onthe container. Specially designed portable steel

61


tanks ( 1.2 m wide by 0.8 m thick by 1.9 m high)containing water are used to provide additionalshielding for the storage containers and thecontrol center. A number of video cameras havealso been strategically located to facilitateoperations when handling the old radioactiveshield plugs. The control center is located at theextreme north end of the fuel bay so that both theactivities on the 615' elevation and at the fuel baylevel can be seen by the operators. Videomonitors, infrared controllers for the fuel baycrane and positioning mechanisms and electricalcontrols for the module handling tool are locatedat the control center.

4.3.2 Testing and InstallationAn extensive out-reactor testing program wasconducted to verify the performancecharacteristics of the new shield plug and theimpact on interfacing components such as thefuel channels components and fuel. Parameterstested included pressure drop, vibration, strengthand crudding. Functional tests were performedon MATF to confirm the fuelling machine and fuelchannel interface. A total of 14 prototype F3SPswere used in a delayed hydride cracking test in-reactor prior a unit 4 outage in March 1995. Eachprototype supported highly irradiated fuelbundles which were subject to a hot shutdown.These bundles were then defuelled andinspected by a specially designed ultrasonic endplate inspection tool developed at Ontario HydroTechnologies. Underwater fuel bay inspectionsusing this tool showed no cracking. The resultswere confirmed by sending a number of thesefuel bundle elements with controls to Chalk RiverNuclear Laboratories for destructive examination.Controls were selected from the previous unit 2test bundles which displayed cracking.

Changes to fuel handling control softwarenecessary for the installation and removal of theF3SP were included in the latest release of theFWF control software. Specific opdata testinghas also been completed. The channel statusmap will also be used to monitor the installationof the F3SPs and serve as a barrier to preventingerroneous operations.

The shield plug handling system has beeninstalled and commissioned. Training of fuelhandling operators is complete.

4.3.3 PlansIt is planned to start installation of F3SPs into the280 inner zone channels of the first unit thissummer and complete all three unit inner zonesby the end of the year. This is a prerequisite toreordering. A dedicated trolley will be used on acontinuous basis to remove the old shield plugsand install the F3SPs. Based on experience withthe FSISP program, it is estimated that the innerzone of one unit can be completed in 2 weeks.Once all inner zones are complete, the 200 outerzone channels will be converted by the firstquarter of 1998 for ease of configurationmanagement.

62


5.0 CONCLUSIONS 6.0 REFERENCES

The impact of the power pulse problem on theproduction of electricity at Bruce A over the past3 years has been significant. Although almost allstation work groups and several Ontario Hydroservice organizations have been affected tosome degree, the Fuel Handling section hasexperienced a extraordinary demand on humanand equipment resources. Not only weresolutions born, but developed and ultimately willbe implemented by them. The engineeringchanges to the fuel handling system to convert tofuelling with the flow were interrupted in August1994 with the discovery of the end plate crackingproblem. The need to replace outlet shield plugswith a new design has added to the work loadand resulted in delays to the schedule. It remainsas a challenge to Fuel Handling in next twoyears, to implement power pulse solutions byinstalling F3SPs, reordering fuel bundles andplacing FWF in-service, in addition to supportingSLAR outages and providing normal fuelling inorder eliminate reactor operating powerrestrictions imposed by the power pulse problem.

1 M. Trelinski, "Darlington NGS Unit 4, In-ServiceReport, DGS-IR-4-03641.3-03, May 1995 In-Service Inspection of Fuel Channel PressureTubes Using PIPE and OPIT", August 1, 1995.

2 D. Jarron, "Darlington NGS Unit 4, In-ServiceReport, D-IR-4-31110.01 Rev 0, August 1993In-Service Inspection of Fuel ChannelPressure Tubes Using PIPE", January 21,1994.

63

MEXTleft BLANK

1ST INTERNATIONAL CONFERENCE ON CANDU FUEL HANDLING SYSTEMS, MAY 1996

FUEL STRING SUPPORTING SHIELD PLUG (F3SP)FOR ONTARIO HYDRO - BRUCE NGSA

PT HenryNuclear Fuel Handling

GE CanadaPeterborough, Ontario, K9J 7B5

CA9700725

ABSTRACTA reactor "power pulse" problem was identifiedfor the Ontario Hydro Bruce generating stations.On a postulated inlet header break, the fuelstrings in a large number of channels couldrelocate toward the upstream end, resulting in apower pulse.

The solution adopted for Bruce GSA is tochange the direction of fuelling, from against theflow, to fuelling with the flow. In this revisedfuelling scheme, given a postulated inlet headerfailure, the fuel bundle with the highest bumupwould relocate into the reactor core andintroduce a negative reactivity during theaccident. However, this fuelling configurationresults in a highly irradiated fuel bundle residingin the most downstream position against thelatch. The latch supports only the outer ring ofelements, not the end plate. A resulting highstress on the end plate coupled with high levelsof hydrogen and deuterium may result in Zrhydride assisted cracking in the end plate duringhot shutdown conditions. (In fuelling againstflow, this is not a problem, since the latchsupported bundle is not irradiated and has onlylow levels of hydrogen and deuterium.)

A fuel string supporting shield plug (f3sp) whichsupports the bundle end plate has beendeveloped as a solution to the fuel bundle endplate cracking problem. It would replace theexisting outlet shield plug in all channels.

This paper will describe the f3sp design,associated fuel handling operation andqualification for reactor use.

1.0 DESIGN (SEE FIGS 1A and IB)

The Dsp uses the existing outlet shield plugdesign, modified to include a nosepiece that

protrudes through the latch to fully support theend plate of the most downstream fuel bundle.All other functions of the shield plug are retainedin the f3sp;eg, diverts the primary heat transportsystem flow from the pressure tube to the endfitting/liner annulus and provides shieldingagainst radiation streaming.

The downstream fuelling machine removes andinstalls the f3sp during each channel fuellingcycle (while the upstream fuelling machineremoves and installs the existing inlet shieldplug). When installed on the shield plug lockinglug, the nosepiece extends forward enough tosupport the fuel string, resulting in a minimumfuel string push of 0.06 inches off the latchsupported position.

Main components of the nosepiece assembly are;fuel support plate, nosepiece body, attachmentbolt, thrust bearing, and detent spring. Theassembly is installed complete on an outletshield plug (s/p) body with the specialattachment bolt. The s/p body requiresmachining to accommodate the new nosepiece,additional to the design of the standard outletshield plug.

A major design intent was to prevent relativerotary motion between the f3sp and the fuelduring s/p installation/removal operations usingthe f/m, because the fuel is unproven under theseconditions (see also section 2.0). Since theBruce style fuel channel and fuel handlingsystem was based on a rotary motion for shieldplug lock/unlock the design evolved as follows.Relative rotary motion between the fuel bundleand the nosepiece during s/p installation andremoval is prevented under normal operation bya thrust bearing between the nosepiece and theshield plug body. The effective friction torqueradius at the bushing is less than half that at thenosepiece/fuel bundle interface, biasing rotation

65


at the bearing rather than at the fuel. Thisbearing also limits the amount of torque that canbe transmitted to the fuel string. In the event ofbinding;eg, crud build up, relative motion islimited by keying the nosepiece to the shieldplug locking lug.

The fuel support plate and nosepiece body arejoined during assembly to form the subassemblythat rotates relative to the remainder of the f3sp.This subassembly is radially supported on theattachment bolt with axial contact against thethrust bearing. The bolt and bearing are lockedto the shield plug body at assembly. Furtherdetail component descriptions follow.

The fuel support plate is held captive on thebearing portion of the attachment bolt. The fuelsupport plate of type 410 stainless steel is 1.00inch thick with a flow hole pattern similar to afuel bundle end plate. The outer, intermediateand inner rings on the fuel support plate areraised to restrict contact with the bundle endplate to these areas giving full support to allthree rings of 37 element fuel. The inner ring ofthe fuel support plate is extended towards the s/pand provides the interface with the attachmentbolt. Only the center fuel element isunsupported. (The outlet shield plug for 500 -600 Mwe reactors only supports the outer tworings of the bundle end plate.) The fuel supportplate is attached rigidly to the nosepiece body byan interference fit and single plug weld whiletwo axial dowel pins are also placed at the jointdiameter to carry rotary loads.

The nosepiece body of type 410 stainless steel isa cylindrical arrangement supported off the outerdiameter of the fuel support plate. Two axialprojecting wings extend outboard from thenosepiece and mate with two sections milled inthe outer diameter of the s/p body. The wingskey on the shield plug locking lug when the s/pis being installed or removed. In the event ofbinding between the nosepiece and the thrustbearing due to crud build up or similarrestriction, this keying limits the potentialrelative rotation between the nosepiece and thefuel bundle to 3 degrees maximum.

The attachment bolt of type 410 stainless steelserves as the axle for the fuel support to turn onbut restricts axial movements of the nosepiecerelative to the s/p. The bolt shaft diameter, at theinterface with the fuel support is chrome platedfor wear and low friction reasons. It is shrunk

and threaded into the nodular cast iron s/p bodyand provides a cantilever support for the fuelsupport. The bolt is locked by two radial pins ofstainless steel. These pins are a press fit and arefurther retained by coining their access hole.

The type 410 stainless steel used in the majorityof the nosepiece assembly was chosen for itscorrorsion resistance, high strength and loweractivation properties compared to other highstrength stainless steels.

The small cylindrical thrust bearing of Nitronic60, having good ami galling properties, isinstalled on the attachment bolt and carries thefull axial load of the fuel string under staticconditions or during rotary motions when thef3sp is installed or removed.

A formed leaf spring of inconel 718 rivetted intoa slot in the larger of the nosepiece body wingsprovides a detent between the nosepiece and thes/p body. The purpose of the detent is to preventrotation of the nosepiece relative to the s/p whenthe s/p is not locked on the liner lug. Thisensures that nosepiece orientation, relative to thes/p, will be correct when the s/p is reinstalled.The spring is attached to the nosepiece by aclamp plate and inconel rivets.

To prepare the s/p body for installation of thenosepiece the following machining is requiredon the inboard end of the s/p (additional to thedesign of the standard outlet s/p); a 5/8 - 11UNC tapped hole in the center of the inboard endface for the attachment bolt, a 0.748 inchdiameter hole leading into the tapped hole forattachment bolt location, a 0.850 inch diametercounterbore leading into the 0.748 diameter forlocation of the thrust bushing, a spotface is alsoprovided to give a quality mating surface for thebushing, two sections are milled on the outsidediameter at the inboard end of the s/p in whichthe nosepiece wings sit - one section coversabout 94 degrees while the other section coversabout 136 degrees, in the larger of the two milledsections two grooves are machined to mate withthe detent leaf spring. The standard shield plugdesign of the bayonet slot and c-springarrangement is retained so that there is no changeto the way it is held in the fuel channel.

2.0 OPERATION

Compared with the existing s/p, operationsassociated with installation or removal of thef3sp are similar since the existing shield plug

66


bayonet slot and c-spring arrangement isretained. A design intent of the f3sp is tominimize any rotation against the fuel and theoperating sequence was developed accordingly:

• The f3sp is picked up by the f/m and isunlocked from its f/m magazine site. TheDsp is advanced into the fuel channel withthe nosepiece aligned to pass over the shieldplug locking lug. The advance is halted atthe position shown in figure 2-1.

• The f/m then rotates the f3sp clockwise tostall against the shield plug locking lug. Seefigure 2-2. This precludes any rotary motionof the f3sp against the fuel string duringinstallation. During removal, if the f3sprotary bushing is seized, there could be amaximum 3 degrees of rotary motion, in acounter-clockwise direction, against the fuelstring (this is the clearance between thenosepiece and the liner locking lug). Thispotential rotation of the downstream fuelbundle has been shown to be acceptable.(Immediately after removal, the fuel bundlepair interfacing the f3sp are removed fromthe channel and discharged to the irradiatedfuel bay.)

• The f/m then advances the f3sp at slow speedto stall axially in the fuel channel against thefuel latch cage. See figure 2-3. During thismotion the Dsp will pick up the fuel stringfrom the fuel latch and thus the Dsp and f/mare carrying the fuel string hydraulic load forthe remainder of the sequence.

• The f/m then retracts the Dsp a small amountfrom the hard stall position. See figure 2-4.

• The f/m then rotates the Dsp about 45degrees clockwise to stall, to lock on theshield plug locking lug,. See figure 2-5.

• Finally the f/m is disengaged from the s/pand the channel closures can be installedprior to the f/m's unclamping from the fuelchannel. Once the f/m unlocks from the Dspthe hydraulic load on the fuel string will pushthe Dsp back into hard contact with theshield plug locking lug but the fuel will stillbe a minimum of 0.06 inches off the fuellatch. See figure 2-6. The Dsp makes nocontact with the pivoting fuel latch segmentsand thus they remain closed under theirspring bias (see figure 1A).

The same basic cycle is repeated in reverse forremoval of the Dsp. The f/m stalls into the Dspto pick it up. The f/m is retracted a smallamount. The f/m rotates the Dsp about 45degrees counterclockwise to stall to unlock fromthe lug. The f/m retracts the Dsp from the fuelchannel to its temporary f/m magazine storagesite and the fuelling sequence continues (the fuelstring will be hydraulically pushed into contactwith the fuel latch until the Dsp is reinstalled atthe end of the fuelling sequence). Note thatbundle support by the latch is acceptable undernormal fuelling conditions. It is only if thiscondition exists during a hot shutdown processthat the potential for hydride assisted crackingoccurs.

3.0 QUALIFICATION TESTINGThe following qualifications were performed invarious test rig channels representing the Brucefuel channel geometry.

3.1 Pressure DropThis testing was carried out in a representativetest channel at GE Canada Laboratory inPeterborough, Ontario.

The Dsp pressure drop was compared to that ofthe standard outlet shield plug. The intent was tomake the pressure drop similar. The nosepiecegeometry is in part a result of this developmenttesting. The final test results showed that for thereactor operating conditions of 300 ° Ctemperature, 9.5 MPa pressure and 27 kg/s massflow, the Dsp pressure drop was within about+/- 7 kPa (+/- 1 psi) of the standard s/p,depending on the particular manufacturingtolerances of components used in the testcomparison.

PRESSURE DROP VS FLOW - REGULAR S/P AND F3SP

S 300

s

~ « — STA«3ABDS/P#1 [KP3) |

- • — STA.fCl*.flO S/P #2 (kfa) ;

-*—F3SP(W=1)

fLOW(k9»)

67


3.2 Fuel Latch Vibration

These tests were carried out using arepresentative test channel at Stern Laboratoriesin Hamilton, Ontario.

When the Dsp is installed it supports the fuelstring off the fuel latch, leaving the latchsegments unloaded except for a light springpreload. Fuel latch segment vibrations werechecked at flows up to 50 kg/s which verifiedthat the motions were insignificant (less than 0.8um).

Similarly, the f3sp nosepiece vibration waschecked in a defuelled channel at flows up to 50kg/s (this would be the case in a defuelledchannel). Again motions were verifiedinsignificant (less than 1.2 |xm).

3.3 Strength Testing

The f3sp was subjected to several abnormalloadings to verify its integrity. All testing wasdone using pertinent sections of fuel channel ofBruce design. The loads were applied using atensile test machine. All were completedacceptably without any loss of function orsignificant permanent deformation. These testswere carried out in the GE Canada laboratory.

The Dsp was pushed back against the shieldplug lug at loads up to 350001b to represent anoutlet LOCA condition where the full systempressure is applied to the f3sp.

The Bsp was stalled hard forward against thefuel latch cage at 10000 lb to represent theabnormal forward axial stall of the component inthe fuel channel during s/p installation using thef/m.

The Osp was stalled hard forward against a flatplate at 10000 lb to represent the abnormalforward axial stall of the component in theancillary port during f/m pick-up.

The Bsp nosepiece was purposely snagged in afuel channel endfitting liner tube while pullingon the outboard end to a load of 10000 lb torepresent the abnormal retract axial stall of thef/m in an abnormal load case were the Bspnosepiece seizes in the fuel channel during axialwithdrawal from the channel by the f/m.

The fisp nosepiece was purposely seized on itsaxle while torquing the nosepiece at loads over360 ftlb to represent the abnormal stall of the f/min an abnormal load case where the fispnosepiece seizes during f/m removal orinstallation.

The Bsp nosepiece was fully rotated back andforth on its bushing while under an axial load of2000 lb while immersed in water to represent thenormal loading seen by the thrust bushing whenthe fisp is installed or removed from the fuelchannel while supporting a fuel string. Threehundred rotation cycles were completed.

The above testing verified Bsp strength asdocumented in a stress report utilizing finiteelement analysis which also covered normal andabnormal load cases.

3.4 Long Term Flow Tests

These tests were carried out in the representativetest channels at GE Canada Laboratory inPeterborough, Ontario.

The purpose of these tests was to verify that theBsp lock/unlock mechanism remains operableduring the extended time period that the s/p sitsstatically in the channel between f/m visits. Thetest was also used to record any damage to fuel,fuel channel or Bsp during the test duration,such as from vibration. Two tests werecompleted, both for a 3 month duration, one at300 °C, 10 MPa and flow of 28 kg/s, the other at300 °C, 10 MPa and reduced flow of 0.5 kg/s.The higher flow conditions represent typical on-power conditions while the lower flowrepresents maintenance cooling flow. Neither ofthese extended tests resulted in the nosepieceseizing on its axle and Bsp/fuel/pressure tubeinspections before and half way through the highflow test indicated that wear of thesecomponents was virtually zero.

3.5 F/M Functional Tests

These tests were performed by Ontario Hydro atBruce A on their representative test channelslocated in the maintenance area test facility. Oneof the station production fuelling machines wasused.

The purpose of these tests was to verify thefuelling machine/fuel channel interface and thespecific f/m computer control. The fuel string

6 8


hydraulic load was simulated by spring loadingthe bundle residing against the fuel latch (againstthe f3sp when installed).

The tests were completed in a cutaway channelso that the operating motions could be verifiedcorrect.

Abnormal operating motions were also tested toverify f3sp integrity during forward axial stall inthe fuel channel with the f3sp purposelymisoriented to prevent passage over the liner lugand to verify that f/m rotary motions away fromthe shield plug lug can not detent the f3spnosepiece.

3.6 Vibration Testing

These tests were performed based on OntarioHydro test specifications using a representativetest channel at Stern Laboratories in Hamilton,Ontario.

The purpose of this testing was to show that thevibrational characteristics of the f3sp are notdetrimental to the fuel channel nor fuel nor itself.This was done by comparing fuel/fuel channelvibration in the existing design (latch supported)with the shield plug supported design.

3.7 Reactor Prototype Tests

Fourteen f3sp's were loaded by production f/m'sinto Bruce unit 4 to demonstrate that their fuelsupport function would prevent cracking of theend plate during a hot shutdown. This wasshown to be the case. Prior to their installationthe particular channels had been partiallyreordered to locate the 13th bundle (an irradiatedbundle) from the inlet position to the outletposition to be supported by a f3sp.

Additionally the f/m channel pressure drop datarecorded during operation on these channelsshowed no excessive differential pressuresacross the fuel string and shield plugs which canbe attributed to the Dsp and thus no measurableeffect on the hydraulic characteristics of the fuelchannel is expected.

4.0 CURRENT STATUSIt is planned to begin installing production Dsp'sin the reactors at Bruce NGSA in 1996.

5.0 ACKNOWLEDGEMENTSThe author wishes to thank Ontario HydroBAND and GE Canada for their approval andsupport in preparing this paper.

6 9

1 ST. INTERNATIONAL CONFERENCE ON CANDU FUEL HANDLING SYSTEMS; MAY 1996

FIGURE 1 A- FUEL STRING SUPPORTING SHIELD PLUG

SHIELD PLUG

NOSEPIECE

SHIELD PLUG


FUEL BUNDLE

FUEL LATCHNOSEPIECE

(PREVENTED FROM ROTATING BYTHE SHIELD PLUG LOCKING LUG)

minikin i mrn

CHANNEL LINER

.06 MIN. FUEL PUSH

.15MAX.FUELPU»J

NOSEPIECETHIBEARING

SHIELD PLUGSHIELD PLUGLOCKING LUG

ATTACHMENTBOLT

=3LtZKFUEL STRING SUPPORTING SHIELD PLUG ROTATED TO LOCK INTO LINER

SHOWN AT CTR 61.5° (1026B )

70


FIGURE IB - FUEL STRING SUPPORTING SHIELD PLUG

FUEL SUPPDRT PLATE

FLDV

3SP READY FDR INSTALLATION DN LINER LUG

NDSEPIECE BODY WINGS

C-SPRING

XBETENT SPRINGSECTIDN 1-1 (S/P BDDY DMITTED FDR CLARITY) SECTIDN 2-2

SECTIDN 3-3(S/P BDDY DNLY)


2.0 OPERATION: F3SP INSTALLATION SEQUENCE

FIGURE 2-1

F3SP STALLS AXIALLYAGAINST THIS SHDULDERX

FACE DF F3SPFUEL SUPPDRT

S/P 1.06 IN<tH FROM STALL

FIGURE 2-2

F3SP STALLS AGAINSTTHIS SHDULDER \

FACE DF F3SPFUEL SUPPORT

F/M RDTARYSTALL

CW TO STALL TO PREVENT ANY ROTATIONL DURING LOCK-IN (IF BEARING IS SEIZED)


FIGURE 2-3

F/M AXIAL STALL

X S / P LOCKING LUG.X)


F/M ADVANC

(DURINGUP THE F

ED TO STALL F3SP AGAINST LATCH CAGE

HIS MDTIDN THE F3SP PICKSUEL STRING)

FIGURE 2-4

FACE DF F3SPFUEL SUPPDRT DUE

RETRACTED CXO5 INCHES FROM STALL. S /P MOVES BACKTO HYDRAULld LOAD ON S/P FROM FUEL STRING


FIGURE 2-5


F/M RDTARY STALL

F/M R3TATED CW TO STALL (45 DEGREES)

FIGURE 2-6

HYDRAULIC AXIAL STALL

FACE DF F3SP IN THlS POSITION THFUEL SUPPORT INBOA?D OF THE LA

BY 0 . (6 TO 0.15 INC

I FUEL STRING HAS BEEN MOVEDCH SUPPORT POSITIONHES

FIRST INTERNATIONAL CNS CONFERENCE ON CANDU FUEL HANDLING SYSTEMS MAY, 1996

FUELLING WITH FLOW AT BRUCE A

M.G.Gray

GE Canada

Nuclear ProductsCA9700726

ABSTRACT

Fuelling with flow is the solution chosenby Bruce A to overcome the potentialpower pulse caused by a major inletheader failure. Fuelling with flow solvesthe problem by rearranging the core toplace new fuel at the channel inlet andirradiated fuel at the channel outlet. Thechange has a significant impact on theBruce A fuel handling system which wasdesigned primarily to do on powerfuelling in the against flow direction.

Mechanical changes to the fuellingmachine include a modification to theexisting ram head and the replacement ofstandard fuel carriers with new fuellingwith flow fuel carriers having thecapability of opening the channel latch.

Changes to the control system are moreinvolved. A new set of operationalsequences are required for both theupstream and downstream fuellingmachines to achieve the fuel change.Steps based on sensitive ram push areadded to reduce the risk of failing to closethe latch at the correct position toproperly support the fuel string. Changesare also required to the protectiveinterlocks to allow fuelling with flow andreduce risk.

A new fuel string supporting shield plugwas designed and tested to reduce therisk of endplate cracking that could occuron the irradiated bundle that would havebeen supported directly by the channellatch. Some operational changes havebeen incorporated to accommodate thisnew shield plug. Considerable testing hasbeen carried out on all aspects of fuelhandling where fuelling with flow differsfrom the reference fuelling against flow.

INTRODUCTION

Early in 1993 the potential for a powerpulse accident on a Bruce or Darlingtonreactor was identified. Following apostulated inlet header failure the fuel inhalf the reactor core channels would shifttoward the inlet end. Since the freshestfuel exists at the outlet end andapproximately half of the mostdownstream bundle has never beenirradiated, this fuel shift immediatelycauses a power spike. When this occurson 240 channels simultaneously, anuncontrolled power pulse results.

Many potential solutions to the problemwere put forth and following methodicalevaluation, two solutions appeared mostfeasible: long fuel bundles and fuellingwith flow. The long bundle conceptsolves the problem by minimizing thegap which exists between the mostupstream bundle and the inlet shieldplug, thereby limiting the potential fuelstring shift. This is achieved byintroducing fuel bundles that are longerthan standard and continually monitoringthe remaining gap. This may seemstraightforward but one must appreciatethat due to channel creep thecombination of normal and long bundlesbecomes unique for every channel. Thenumber and position of long bundlesmust always be tracked, the minimumgap size in any channel cannot beviolated, the fuelling machines must beprecisely loaded with the correctcombination in the correct order in thecorrect magazine positions and there canbe no deviation from the planned fuelchanging mission once the machine isloaded. Much effort has gone intodevelopment of the long bundle program,the details of which are outside the scope

75


of this paper. It has currently beenimplemented in limited form at Bruce Band Darlington.

Fuelling with flow solves the power pulseby reversing the core to place the freshestfuel at the inlet end and the irradiated fuelat the outlet end. The fuel string shiftfollowing an inlet header break thenresults in a decrease in reactivity as themost irradiated fuel enters the core andthe freshest fuel leaves. This is thesolution chosen by Bruce A. This paperdiscusses the changes required to thefuel handling system along with testingcompleted to support the fuelling withflow program.

BACKGROUND

Although the Bruce A reactors and fuelhandling system were designed to befuelled against the flow, there was alwaysa requirement to be able to hold a latchopen to allow fuel to pass back through.Initially this was intended to be done bymanual operations with special toolingunder shutdown conditions, but laterevolved into the flow defuelling techniquedeveloped as part of the breakdownchannel defuelling system. Thistechnique has been used successfully atall stations with latch endfittings.Experience with flow defuellingcontributed significantly to the fuellingwith flow program at Bruce A.

Fuelling with flow does not result in anymajor changes to the fuel handlingsystem. The new fuel is now inserted atthe channel upstream end and theirradiated fuel discharged at thedownstream end. New control systemOPs and Sequences have been written toachieve this. Also a new fuelling withflow fuel carrier has been provided in thefuelling machines, and a new fuel stringsupporting shield plug has been installedat the channel outlet end. The new fuelmechanism ram stroke has beenextended to accommodate the new fuelcarriers. Changes to some protectiveinterlocks have been provided and somechanges have been made to control

hardware. Testing of some criticalfuelling with flow parameters wasperformed in the GE Canada lab andChalk River Nuclear Labs. Full testing offuelling machine sequences are beingcarried out at Bruce A.

MECHANICAL HARDWARE

Fuelling with flow requires inserting newfuel at the channel upstream end whiledischarging irradiated fuel at thedownstream end. Although the originalfuelling machine fuel carriers are capableof the new operations at the upstreamend, they are unable to open the channellatch to allow passage of fuel out of thechannel at the downstream end. Newfuelling with flow fuel carriers weretherefore designed and supplied. Thenew design is shown in Figure 1. Foreconomic reasons the design reuses theoriginal fuel carrier body, truncated, withthe new nosepiece riveted on.

The fuelling with flow fuel carrier has aprotruding, thin walled nose which canopen the channel latch and hold it openduring the fuel change. The new carrieralso has a flat shoulder machined on it,specifically to allow stalling against theshield plug lug in the end fitting. Anoperation stalling the carrier against thelug is carried out during the fuel acceptsequence to verify the charge tube axialencoder prior to opening the latch. Thenew carrier incorporates inserts tooperate the existing breakdown channeldefuelling ram catch adapter. It is nowunnecessary to replace the fuelling withflow fuel carriers with defuelling carrierswhen grappling is required.

Testing indicated a potential for bundleendplate cracking of irradiated fuel heldby the channel latch. A new fuel stringsupporting shield plug ( F3SP ) has beendesigned and supplied, and is shown inFigures 2 and 3. This shield plug featuresa nosepiece which protrudes just throughthe channel latch and supports the fuelstring off the latch fingers. Relativerotation can take place between the shieldplug body and the nosepiece. In service

76


the nosepiece is prevented from turningby the end fitting shield plug lug whilethe shield plug body rotates to lock orunlock from the end fitting. This ensuresno relative rotation between thenosepiece and the fuel which otherwisemight result in fuel damage.

The stroke of the new fuel mechanism'sfuel transfer ram is too short to reliablypush bundles past the lip of a defuellingcarrier. Because of this shortfall,equipment was designed to retrofit thefour mechanisms with different ramposition detection components. The sameretrofit allows the ram to providesufficient additional stroke.

The chamfer has been increased on thefour cruciform sections on the back endof the fuelling machine ram head. Thisimproves the ability of the ram head toopen the channel latch in the retractdirection. Channel creep results in theneed for the downstream ram topenetrate through the latch and anychange in the orientation of the ram headduring retract will result in contactbetween the ram head and the latchfingers. The ability to open the latchsmoothly prevents the ram from hangingup or possibly causing damage.

OPS and SEQUENCES

New fuelling machine control operationsand sequences had to be written toachieve fuelling with flow. The fuelchange occurs while the channel latch isbeing held open by the fuel carrier at thedownstream end. Any operations whilethe latch is open are critical since anincident could result in the fuel stringpassing uncontrolled through the latch. Itis especially important that the fuel beheld in the correct position while thecarrier is withdrawn to close the latch toensure that the latch fingers dropbetween bundles correctly to properlysupport the fuel string. Consequently thefuelling with flow ops and sequencescontain more steps specifically added toverify correct position or operation.

New sequences were also written forinstallation and removal of the new fuelstring supporting shield plug at the outletendfitting. New Ops in the sequencescheck the type of channel end beingvisited to ensure that the appropriatesequence is being used.

PROTECTIVE INTERLOCKS andCONTROL

In the fuelling with flow mode ofoperation, the channel latch is opened toaccept fuel into the fuelling machine oneach fuel change. Holding the latch openwith the fuelling with flow fuel carrierincreases the hazard, since there is agreater period of time during whichsystem failure, either due to power failureor system error, could lead to fuel beingwashed back through the open latch intothe fuelling machine, leading to a difficultrecovery process. For this reason, anumber of interlock changes have beenmade to the system.

While the ram is supporting the fuelstring, as the string is retracted to receivethe irradiated fuel at the downstreamchannel end, at least one of the ramclutches must be engaged so that the ramis not pushed back freely by the fuelstring. Hardwired interlocks preventreleasing the ram brakes at thedownstream end unless one of theclutches has been engaged.

Similarly, while the charge tube hasadvanced to a position where the fuelcarrier is holding the latch open, if theaxial drive brake is released and theclutch is not engaged, the charge tubewould be pushed backward by the fuelstring, leaving fuel improperly positionedrelative to the channel latch. Hardwiredinterlocks prevent releasing the chargetube axial brake unless the clutch isengaged.

A number of protective system interlockshave been changed to ensure that the fuelstring is properly supported while thelatch is open, and to ensure that the fuelwill be left properly positioned after the

77


latch is closed. Another set of changesprevents advancing the fuel string againsta braked ram during the various sensitiveram push operations, and otherwisepinching the fuel string between tworams at high torque. Torque limitationson the fuel supporting shield plugrequired a limit on charge tube rotarytorque while installing or removing theshield plug.

Interlock changes are incorporated toboth ram and charge tube axial drives toensure proper positioning of fuel relativeto the latch. For the ram drive interlocks,one change specific to ram motionapplies. When retracting the fuel stringinto the downstream fuel carrier while itis holding the latch open, the ram can beretracted only to the position where thebearing pad at the upstream end of thesecond bundle is over the channel latchline, so that as the latch is closed itcannot catch the bearing pad. At thispoint, the third bundle is still upstream ofthe latch, and there is no danger of itcoming through the latch as the fuelcarrier is withdrawn. The charge tubeaxial interlocks are mainly related toensuring that the fuel is properlypositioned before the fuel carrier can beretracted to close the latch. Compensatedaxial motion is used when closing thelatch such that the charge tube and fuelcarrier are retracted while the ram isdriven forward relative to the charge tubeto hold the fuel string stationary withrespect to the latch. This forward motionof the ram inside a fuel carrier in a lipdown orientation was previouslyprevented due to the possibility ofdamaging a bundle against the lip, but isnow allowed over specific definedranges.

To ensure that the forces on the fuelstring are kept low enough to prevent fueldamage, a number of new protectiveinterlocks have been applied. Signals aregenerated from the upstream fuellingmachine to indicate that both of the rambrakes and clutches have been released,and therefore the ram is 'floating'. Thiscondition is required during ram positionsensing operations which are done more

often during fuelling with flow. Whiledischarging fuel from the upstream fuelcarrier, the ram must be advanced behindthe fuel to move it along within the carrieruntil it reaches the area of channel flowand is carried by the flow into thechannel. Since there is the possibility ofadvancing this ram to pinch the fuelstring between the two ram heads oragainst the latch at the downstream end,the speed and torque for this ramadvance motion are limited to lowervalues than when advancing the ramwithin a fuel carrier at the downstreamend.

Controller program functions have beenadded to accommodate the fuelling withflow mode of operation. The newfunctions ensure that fuelling with flow isbeing carried out at appropriate channelsand each fuelling machine is executingthe right sequence for the end at which itis operating. There is capability ofmonitoring the motion of the rams of thetwo fuelling machines to ensure that theyare properly synchronized during the fueltransfer. As the downstream ram retractsto accept fuel bundles into the carrier, theupstream ram advances to move fuelalong so that it is carried into the channelby the flow. The upstream ram followsalong behind the fuel to ensure that itcontinues moving into the channel.Maintaining a gap between the ram andthe fuel ensures that the fuel string is notpinched between the two fuellingmachine rams, while minimizing the gapreduces the impact as each bundle isswept by the flow into the slowly movingfuel string. This synchronization isachieved by having ram encoderfeedback from both fuelling machinesmonitored by each controllersimultaneously. The ram encoder signalis connected directly as an input to thecontroller of the other fuelling machineon the same trolley, thereby avoidingdelays inherent in communicating thedata over the interprocessorcommunication links.

During the changeover from fuellingagainst flow to fuelling with flow, it isimportant to keep track of the fuel

78


ordering in each channel since both typesof fuelling will be in service until thechangeover is complete. A channelinformation data file, CHANN.TD, hasbeen created to store the characteristicsof each channel of each reactor so thatthe control system can determinewhether the proper fuelling approach hasbeen selected for the channel. Otherpertinent data stored for each channelinclude: channel data updated, channelquarantined, shield plug not installed,channel defuelled, long bundle(s) inchannel, flow restricting outlet shield pluginstalled, and F3SP installed.

The operation of all of the fuellingmachine brakes has been changed to a'command to release1 from the previous'command to engage". In this way, onloss of computer control, the brakesrevert to the braked state, holding themachine components in their currentpositions.

For additional reliability, Class II powerhas been provided to both of the brakecircuits for the fuelling machine ram,charge tube axial and rotary drives. TheClass II power is switched within the fuelhandling power distribution cabinet sothat as long as the normal Class III supplyis available, it is used and the Class II isstandby. If the Class III is lost, Class II isimmediately switched in with only amomentary delay in the supply to thebrakes. This change will prevent the fuelstring passing through the latchuncontrolled in the event of a loss ofpower during the fuel change while thelatch is being held open.

TESTING

A series of tests were performed in theGE Canada lab to demonstrate thefeasibility of fuelling with flow at Bruce Aand establish the key flow rateparameters. These tests are summarizedas follows:

1. Check the release of fuel bundles pastthe lip of the upstream fuel carrier atnormal full flow conditions.

2. Establish the maximum flow at whichfuel bundles will successfully movepast the lip of the upstream carrier.

3. Determine the effect of a bundleimpacting the fuel string over adistance of 45 inches under full flowconditions.

4. Establish how far the fuel carrier canretract from full forward stall in thelatch before a bundle no longerpasses through.

5. At full flow conditions establish thepoint of relative movement duringrotation of the upstream carrier whenthe new fuel is in contact with the fuelstring.

6. Establish the minimum flowrequirements to consistently washtwo bundles past the lip of thedownstream fuel carrier.

The results indicated no reportabledamage to any of the fuel bundles, fuelcarriers or channel components duringany of the tests. The minimum flow rateto successfully receive two bundles overthe lip of the downstream carrier wasestablished to be 20.9 Kg/s at 300°C. Theseries of tests showed that pursuingfuelling with flow at Bruce A would be aviable option.

Testing at CRNL indicated a possibleproblem with cracking of endplates ofirradiated bundles supported on the latch.This concern led to the design andimplementation of the fuel stringsupporting shield plugs in the outletendfittings.

Testing of the new Ops and Sequences,and the new fuel handling protectiveprograms are being performed at Bruce Aby Ontario Hydro with support from GECanada Fuel Handling personnel.

CONCLUSIONS

With appropriate changes to the fuelhandling equipment and control system,the reactors at Bruce A can be refuelled inthe fuelling with flow direction. Thecurrent test program together with

79


previous flow defuelling experienceestablished confidence in this fuellingmethod. The replacement of outlet shieldplugs with the new fuel string supportingshield plugs will eliminate the risk ofbundle endplate cracking at the latch.

out as part of the Power Pulse programfunded and managed by Ontario Hydro.

ACKNOWLEDGEMENTS

The author would like to acknowledge thecontributions of Ontario Hydro, AECL andthe GE Canada fuel handling team to theoverall program. The work was carried

FUEL CARRIER(MODIFIED)

INSERTS

NOSEPIECE

PINS

FLAT FOR STALLING LINERAGAINST SHIELD PLUG LOCKING LUG

\

INSERT

HUM \FUEL CARRIER

FUEL CARRIER SHOWN WITH FLAT STALLEDAGAINST SHIELD PLUG LOCKING LUG .

SECTION AT INSERT(INSERTS ALLOW USE OF THIS

CARRIER FOR GRAPPLING)

Figurei - Fuelling With Flow Fuel Carrier

80


FUELB NDLE

SHIELD PLUG

CHANNEL LINER


FUEL LATCH

NOSEPIECE

FUEL STRING SUPPORTING SHIELD PLUG ADVANCED INTO LINER

0.19 MINI. FUEL PUSH

FUEL BUNDLE

CHANNEL LINERSHIELD PLUGLOCKING LUG

Figure 2 - Fuel String Supporting Shield Plug - Unlocked

81


SHIELD PLUG

NOSEPIECE

SHIELD PLUG

FUEL BUNDLE

FUEL LATCHNOSEPIECE

(PREVENTED FROM ROTATING BYTHE SHIELD PLUG LOCKING LUG)

SHIELD PLUG

CHANNEL LINERSIHIELD PLUGLOCKING LUG

.06 MIN. FUEL PUSH ..15 MAX. FUEL PUSH 11

FUEL STRING SUPPORTING SHIELD PLUG ROTATED TO LOCK INTO LINER

Figure 3 - Fuel String Supporting Shield Plug - Locked

82

FIRST INTERNATIONAL CNS CONFERENCE ON CANDU FUEL HANDLING SYSTEMS MAY. 1996

BUNDLE 13 POSITION VERIFICATION TOOLDESCRIPTION AND ON-REACTOR USE

byT.G. Onderwater

GE CanadaNuclear Fuel Handling

107 Park St. N., Peterborough, Ontario, K9J ,~~

CA9700727

ABSTRACT

To address the Power Pulse problem,Bruce B uses Gap: a comprehensivemonitoring program by the station tomaintain the gap between the fuel stringand the upstream shield plug. The gapmust be maintained within a band. Thegap must not be so large as to allowexcessive reactivity increases or causehigh impact forces during reverse flowevents. It should also not be so small asto cause crushed fuel during rapid,differential reactor/fuel string cool downs.Rapid cool downs are infrequent.

The Bundle 13 Position Verification Tool(BPV tool) role is to independentlymeasure the position of the upstreambundle of the fuel string. Themeasurements are made on-reactor, on-power and will allow verification of theGap Management system's calculated fuelstring position.

This paper reviews the reasons fordeveloping the BPV tool. Design issuesrelevant to safe operation in the fuellingmachine, fuel channel and fuel handlingequipment are also reviewed. Testsensuring no adverse effects on channelpressure losses are described and actualon-reactor, on-power results arediscussed.

1. INTRODUCTION

Gap Management is a comprehensivemonitoring program to maintain the gapbetween the fuel string and the upstreamshield plug {see Figure 1). The programdefines a safe range of gap between the

upstream shield plug and the thirteenthbundle. The gap is to be small enough tolessen the reverse flow impacts and limitincreased reactivity. As well, the gap isto be large enough to ensure that the fuelstring is not compressed due todifferential cooling during rapid reactorcool down conditions.

When channel creep lengthens the fuelchannel, the gap between the thirteenthbundle and the upstream shield plugincreases. As the channel creeps,increasing numbers of special longbundles are used in the fuel string tomaintain the proper gap.

The BPV tool assists in the process byobtaining independent measurements ofthe gap. These measurements are usedto determine correction factors for thegap calculation.

2. GAP MANAGEMENT REQUIREMENTSFOR BPV TOOL

The Gap Management program containsalgorithms and data bases to continuallytrack the position of bundles. Therecorded bundle lengths in the fuel stringsare combined with knowledge of fuelchannel lengths to calculate the existinggaps. The program requires verificationof this gap by means independent of theprogram inputs (bundle records, fuellingmachine end fitting positionmeasurements, fuelling machine fuelstring position measurements). The gapcalculations must also account for varyingreactor conditions. Again, the need foran independent tool capable of operating

83


in reactor conditions ledrequirement for the BPV tool.

to the

The gap is maintained through knowledgeof the gap and the use of two sizes offuel bundle (19 inch standard and 19.5inch 'long', 483 and 495 mmrespectively). The fuel string is composedof all standard length bundles when thefuel channel is in its early years of use.As the radiation induced creep lengthensthe pressure tube, the two end fittingsgrow apart. This causes the gap toincrease. When the gap is determined tobe approaching the upper tolerance, oneof the new bundles installed will be a longbundle. The gap is thereby reduced byhalf an inch. The long bundle is advancedupstream with each subsequent fuellingoperation until the it eventually reachesthe discharge position. At this time thenew fuel charge must have the samecombinations of bundle types to maintainthe fuel string length and the gap.

The gap management algorithm, utilizedby the Bruce B fuel handling controlsoftware, incorporates a basic equation todefine the position of upstream bundle13. During each channel fuellingoperation, upstream and downstreamfuelling machines have their rams incontact with the fuel string. The ramencoders supply two of the variablesneeded to measure the gap. The allowedgap, the accuracy of the fuel handlingsystem, and the fuel channel currentlength are constants that have beendetermined. Other variables aredependent on what combinations of fuelbundle types are required to be insertedand discharged. These fuelling operationvariables are specified when the operatordetermines the bundle combinations used.

The basic equation is affected by thepower level, the specific station, reactortemperature, and the use of flowinjection. These effects are grouped intothe single correction factor Kp. Thisfactor can only be determined by actualmeasurements in representativeconditions. The BPV tool was developedto assist in this role. It is designed to be

installed in any channel of an operatingBruce B reactor. Measurements retrievedfrom the tool can be used to assignvalues of Kp for each combination ofpower level, channel position, etc.

An additional role of the BPV tool is toassist in recovery from some accidentcases due to bundle combination errors.If the gap management system's gapalgorithm detects an error, the BPV toolcan be used to measure the actual bundleposition independently.

3. BUNDLE 13 POSITION VERIFICATIONTOOL DESCRIPTION

The BPV tool obtains measurementscompletely independent of the channeland fuelling machine components becauseit incorporates a mechanical recordingdevice. This device is basically a slidingdepth gauge and takes its reading directlyfrom the fuel bundle relative to the endfitting.

The BPV tool (see Figure 2) is composedof three major components:• tool body• fuel sensor• spring plungerAll three parts are made of the same typestainless steel.

3.1 Tool BodyThe tool body (see Figure 2) is a modifiedfuelling machine fuel carrier made ofstainless steel with chrome plating of itssliding surfaces. The outboard end(relative to the reactor centre) has slotsand recesses to facilitate connection tothe fuelling machine charge tube. Alongthe length are holes for the primary heattransport coolant flow. A relief of theouter diameter for most of the length onone side provides clearance when thebody is warped by the stratifiedtemperatures encountered during service.A slot with gauge markings, inconjunction with a pin mounted in thefuel sensor, indicates the measuredposition of the fuel bundle. A bulkhead ispinned within the bore of the body. Thebulkhead restrains the spring of the spring

84


plunger. A tubular extension of thebulkhead restricts the fully extendedreach of the fuel sensor to no more than1.5 inches (38 mm) past the inboard endof the body. The inboard end enters thechamfer at the end fitting spacer. Thespacer line is a fixed plane through theend fitting at the end of the channel linertube. The shield plug, fuel carrier andBPV tool all abut the chamfer at this line(see Figures 1, 2 & 4).

3.2 Fuel SensorThis part is contained with the springplunger and body (see Figures 2 and 3).It functions as a depth gauge probe. Achromed, segmented flange providessliding contact with the spring plungerbore. The flange is designed to yield ifthe sliding surfaces jam. The axialposition of the sensor within the plungeris held by friction pads. The springswithin the pads are shimmed to providesufficient friction to resist the drag of thecoolant flow yet slide with an acceptableforce when the fuel bundle isencountered.

A set of concentric rings at the inboardend distributes any forces against the fuelbundle. The outermost ring has threepads to support the inboard end and allowthe sensor to slide along the bores of thepressure tube, liner spacer and tool body.Radial holes allow coolant flow to reachthe centre of the fuel bundle.

3.3 Spring PlungerThe spring plunger is tubular. The borecarries the fuel sensor. The outerdiameter has chromed bearing rings toslide within the body. The axial limits ofthe plunger stroke are set by the bulkheadand a step in the body's bore near itsoutboard end. Normally the parts arekept in the retracted outboard position bya large spring between the plunger andbulkhead. During measurementoperations the fuelling machine ramcompresses the spring and advances theplunger and fuel sensor.

4. BPV TOOL OPERATION

The BPV tool is brought out of storageand placed in a trough attached to theancillary port (a containment penetrationaccessible to the fuelling machine withincontainment and by personnel in theirradiated fuel bay room). A hand tool issupplied by which the indicator pin can bepre-set. When the BPV tool is ready, thefuelling machine engages the tool andstores it in its magazine.

After locking onto the target end fitting,the fuelling machine engages with theBPV tool and advances it into the endfitting to stall, positioning the tool nose atthe spacer line. Fuel coolant flow will beconstantly flowing through the tool.

The plunger, with its sensor having beenpre-set in its fully extended position, isadvanced to the measuring location bythe fuelling machine ram (see Figure 4).When the sensor encounters thethirteenth bundle before the plunger hasbeen fully advanced, relative slidingbetween the stalled sensor and theadvancing plunger causes the indicatorpin to slide back along the groove.Retracting the ram allows the gauge to befully stored within the BPV tool's body.When the plunger has returned to itshome position, the pin lines up with thescale inscribed on the body and allowsthe detected position of the bundle to beread.

The scale is set up to read positivenumbers if the thirteenth bundle isoutboard of the spacer line. Thus areading of positive 2.5 inches wouldindicate that the bundle might encounterthe flow straightener if the shield plugwere re-installed. A zero readingindicates that the bundle end plate is inthe plane of the spacer line. If the bundleis farther inboard than the spacer line,negative measures are recorded.

It should be noted that the second fuellingmachine will lock onto the other end ofthe fuel channel for the duration of theBPV tool visit. With two fuelling

85


machines in use, the ch?rr^« >>»essuredrop is monitored for any excursionsbeyond what is normal during fuellingoperations.

The BPV tool is then retracted into thefuelling machine and returned to theancillary port. The fuelling machine lockson the port, the port is opened manuallyand the trough installed. The tool is thenbe slid onto the trough. The reading onthe tool's scale is recorded as thethirteenth bundle position.

The tool could either be removed andstored or reset for further measurements.

5. DESIGN ASPECTS FOR RELIABLE IN-REACTOR USE

5.1 Accommodating Flow InjectionThe Bruce B fuelling machines incorporatea flow injection system. While thefuelling machines are locked onto a fuelchannel, cooling water is circulatedthrough the machine's magazine. Toguard against the ingress of hot waterfrom the channel and any debris or crudthat may come with it, sufficient coolingwater is injected to create a slow flowout of the fuelling machine. • Whenconsidering tools or components thatleave the relatively cool fuelling machineinterior (between 27 and 93 Celsius, or80 to 200 Fahrenheit) and enter the hotfuel coolant flow (up to 305 Celsius, or580 Fahrenheit), thermal effects must beconsidered.

The tool was designed with similarmaterials throughout such that thermalexpansion has virtually no effect on themeasurement. However, a small error isunavoidable based on the differentialmovement of the expanded parts. Theexpansion/contraction applies only to therange of sensor shift (-1.5 to +4 inches= 6 inches). The net effect onmeasurement is well below the desired+ /- 1/16 inch accuracy required.

A less obvious thermal effect is due tothe fact the flow injection water does notmix for some length down the end fitting.

The stratification causes a considerabletemperature difference along the top andbottom of any component installed. Thedifferential thermal expansion causescomponents to bow, sufficiently in somecases to cause interference within theend fitting liner bore.

The BPV tool accommodates the bowingeffect by being designed with largeclearances and ensuring long componentswere not over constrained.

Commissioning tests included differentialheating of the tool's body to duplicate theworst case bowing. Stroking the tool'scomponents in this condition ensured thatno jamming would occur in the reactorthat could cause excessive forces againstthe fuel bundles.

5.2 Coefficient of FrictionThe sensor relies on its friction pads tolimit the force against the fuel bundlesduring initial installation and themeasurement operation. Extensivecommissioning tests were performed toprove that the design force was notexceeded regardless of whether the toolwas heating up, cooling down, wet ordry.

Tests uncovered one unique condition inwhich the friction pads locked. Thesequence of events was:• the tool was exposed to simulated

coolant water,• then allowed to dry, and• bowed by applying heat.Previous tests had shown the coefficientof friction between certain components tobe within the expected range foraluminum bronze against stainless steel.Within this range, the geometry is suchthat window jamming should not occur.

An interesting trend developed afterinvestigating the possible combination ofapplied forces and spring forces withvarious coefficients of friction. As thefriction coefficient increased the staticmodel could be shown to jam in onestate. However, at higher coefficients thejammed case would develop a different

86


set of equilibrium forces and theoreticallynot jam or be statically indeterminate. Inthat condition, another equilibrium statewould exist, jamming and locking at othercoefficients of friction.

Tests on the tool were able to duplicatethe coefficient of friction and show that itcorresponded to the lowest valueresponsible for the geometry to lock.This low value was twice as high asnormal for the combination of materials.The high coefficient of friction wasthought to be due to the extremely cleancondition of the tool after being exposedto the hot, basic pH, coolant. It was alsothought that the lithium hydroxide used toadjust the water's pH may have formed athin deposit of crystals. Stroking thecomponents once or twice alwaysreduced the coefficient of friction towithin the expected range

It is noteworthy that the jamming casewas not classical window jamming, but infact, the friction pad pistons stalling all onone side of their bore hole. The uniqueequilibrium cases were caused by therelative movement of internal componentsas the tool body bowed. By changing thefriction pad tip profile and by movingthem to a plane not sensitive to bodybow, the jamming is avoided, even at thehigh coefficients of friction.

5.3 Tool Failure SafetyAs the tool is to be pushed against thefuel string through commands of the fuelhandling control software, several toolfailure modes were investigated.

The first to be investigated was thefailure of an extended sensor to retract(due to spring failure or particles jammingin the sliding surfaces of the sensor). Insuch a case, the tool would not fit withinthe fuelling machine magazine. Thefuelling machine magazine would becommanded to rotate regardless. Thesensor is strong enough to stall themagazine drive without itself beingdeformed. This would allow for acontingency manoeuvre in which the fuelstring would be advanced by the opposite

fuelling machine to push the sensor backinto the tool body.

To guard against particles preventingretraction of the sensor, two designfeatures were incorporated. On theoutboard end of the sensor, bearingsurfaces were designed with the ability tofold away from jammed particles. At theother end, generous clearances and smallfootprint pads allowed jammed particlesto be dislodged by jogging the sensor.

The folding bearing pads also guardagainst excessive forces being applied tothe fuel string when the fuelling machineram extended the plunger.

5.4 Fuel Channel Pressure DropThe BPV is an intrusive tool in that itmust place a stiff sensor within thecoolant flow up against the upstreambundle. If the flow were to be blocked,fuel bundle temperatures would rise.Normal fuel handling control softwaremonitors the pressure differential. If aflow blockage is deduced, the operatorsare warned and the BPV tool would needto be removed.

The sensor head was designed to avoidblocking flow to the bundle. Ample flowareas are machined through the head andthe central tube is cross drilled to deliverwater flow to the bundle centre. Upstream of the bundle, coolant flow entersthe tool body radially through holes thatare in the same area of the radial holes inthe liner. All plunger, spring and bulkheadcomponents are placed behind the area ofthe entering flow.

Tests were performed in the GE Canadafuel channel rig. The sensor was placedin various positions and pressure tapswere used to obtain pressure differentialsalong the BPV tool. The pressure dropsrecorded were well below the designlimits. Drag on the sensor was deducedfrom some of the pressure differentials.This allowed the establishment of frictionpad settings.

87


6. SAFETY ASPECTS OF BPV TOOLMAINTENANCE

The safety concerns of tool maintenancefall into two categories. First, ensuringthe tool operation is reliable and will notdamage fuel channel components.Second, ensuring that handling the highlypreloaded springs is safe for themaintainers.

Prior to being sent on its mission, thesensor friction is checked. The hand toolsupplied to pre-set the indicator pin isused to jack the sensor back and forth.The capacity of the hand tool is lowerthan the friction loads that would damagethe fuel bundles or latch. If the sensor isdifficult to move, a test is performed toapply loads to the sensor and documentthe maximum force required to push thesensor. The test results confirm that theforce limiting friction pads are performingwithin the acceptable limits.

Visual inspection prior to the missionwould confirm that the sliding surfacesare intact.

If the plunger is suspected of beingjammed, tests and test fixtures areprovided to check the plunger movement.Stacked weights added to the test fixturedepress the plunger and allowconfirmation of the spring preload andstiffness.

The tool's highly preloaded internal springpresents a potential hazard to themaintainers. A set of maintenance toolsand maintenance instructions areprovided. Using these, the spring can besafely locked during maintenance andoverhaul.

7. SITE MEASUREMENTS

The BPV tool has been commissioned andused successfully at site. Fullcommissioning was performed on theBruce B fuel channel test facility. There afuelling machine head installed the toolinto a test channel with full reactoroperating temperature, pressures and

flows. Indicator pin readings were foundto duplicate the known position of thethirteenth bundle in the test channel.

Subsequent use on a number of reactorchannels has provided fuel bundlepositions under operating conditions.These readings have confirmed theestimated Kp correction factor for one setof reactor conditions.

ACKNOWLEDGMENTS

The author would like to thank OntarioHydro and GE Canada for the use of theirdata and to acknowledge Patrick Henry ofGE Canada as the originator of the BPVtool concept.

8 8

FLOW STRAIGHTENER,THIRTEENTH

BUNDLE

COOLANTFLOW FLOW INLET

END FITTING

FEEDER

oo

CHANNELCLOSURE

FLOW OUTLETEND FITTING

IPRESSURE

TUBE

H3

ino

Io2:

2

r52:o

o(/J

5


SCALE

*-— BODY (fuel carrier)

FLOW HOLES

FUELLING MACHINERAM INTERFACE

"RELIEFSTROKE

INBOARD / OUTBOARD

FIGURE 2 - BUNDLE THIRTEEN POSITION VERIFICATION TOOL

90


INBOARD BEARING PADS

INBOARD END

FLOW HOLES

INBOARD BEARING PADS

HOLES FOR FRICTION PADS

PINOUTBOARD / IN8OARD

OUTBOARD BEARING SURFACES

FIGURE 3 - FUEL SENSOR

91


Spacer Line Within }End Fitting r

Bulkhead

Body

^Indicator pin set at -1.5 inch( for possible maximum measurement

Spring Cavity-v \

AS PREPARED AT ANCILLARY PORT

Fuel Bundle AS INSTALLED IN END FITTING

Gap closed by F/M r a m ^ ^ ..j.XI.,.f.,.?...t...

-b .F /M

DURING MEASUREMENT STROKE-1.0 inch indicated (underhang)

mmmmmmAS RETRACTED Plunger returned by spring

FIGURE 4 - BUNDLE THIRTEEN POSITION VERIFICATION TOOL OPERATION

92


THE OPERATION AND MAINTENANCE OF THE SLAR SYSTEM AT BRUCE A

by

S. Ahuja, M. Eng, P. EngSenior Technical Engineer/Officer

Ontario HydroBruce A Nuclear Generating Station

Fuel HandlingP.O. Box 3000

Tiverton, Ontario, Canada, NOG 2T0

CA9700728

AbstractThe SLAR (Spacer Location And Repositioning)system at Bruce A consists of two (2) DeliveryMachines and a Fuelling Machine Trolleyequipped with the D2O and Air auxiliarysystems. The Delivery Machines are designedto perform all the Fuelling Machine operationsand have the capability to rapidly defuel/refuela reactor channel and traverse the SLAR tool tolocate and reposition the spacers in thechannel. The number of functions that aDelivery Machine must perform makes it morecomplex as compared to the operations of aFuelling Machine.

The paper discusses the operation of the SLARDelivery Machines and the problemsencountered with the operation andmaintenance of this system at Bruce A.

1.0 IntroductionSLAR is a maintenance program developed forthe rehabilitation of the fuel channels in theCANDU reactors. It was determined that theP/T (Pressure Tube) to C/T (Calandria Tube)contact can be detrimental to the safe and longterm operation of a reactor. The P/T to C/Tcontact can be caused by the displacement ofthe fuel channel spacers. The contact area canlead to the formation of Hydride blisters leadingto a failure of the P/T over a period of time.Hence, Bruce A implemented the SLARprogram to reposition the spacers in Units 3and 4 reactor channels.

The SLAR program at Bruce A has beensuccessful in completing Three (3) campaignssince 1993 on Units 3 and 4. A total of FourHundred and Forty Nine (449) channels havebeen SLAR'd in approximately Three Hundred(300) days of critical and non-critical reactoroutage time. The SLAR Channel processingtimes have varied depending on the size of thespacer and the relative position of the channel

in the reactor. The 'Fat' spacers in the Unit 3channels were found closer to their designposition than the 'Thin' spacers in the Unit 4channels because of their tighter fit. Thespacers in the top row reactor channels weredisplaced more from their design location thanthe channels closer to the reactor core.

2.0 The Operation of the SLAR SystemThe SLAR system at Bruce A consists of two(2) D/Ms' (Delivery Machines, Fig. 1) and aFuelling Trolley (Fig. 2) complete with its D2Oand Air auxiliaries. The D/Ms' are designed tooperate within the Fuelling Machine (F/M)envelope. This has made the design of thesemachines more complex, as they must performmore operations, compared to a F/M.

After the D/Ms' remove the channel plugs, thehydraulic ram within the M/T (Module Tube,Fig. 3) is used to defuel the channel into theF/T (Fuel Tube, Fig. 4) of the D/M on theopposite side. The hydraulic ram also traversesand rotates the SLAR tool during its operationwithin the channel. The fuel is returned to thechannel by a hydraulic piston in the F/T and thechannel is closed.

2.1 The SLAR Delivery MachinesThe basic design of the SLAR D/Ms' is modifiedby separating a F/M at the flat head and leavingthe front end unchanged (Fig. 1). Theremaining assembly consists of an indexingturret in place of a flat head on which three(3)tubes are mounted:

1. A Mechanical Ram Tube, containing theexisting charge tube and ram ballscrews andinput drives

2. A Fuel Tube

3. A Module Tube containing the SLAR tooland the hydraulic ram.

93


By rotating the turret, each of the tubes can bealigned with the snout of the head. The F/Tand the M/T are longer than the mechanicalram tube and are retracted when the headenters or leaves the reactor vault.

2.1.2 Fuelling TrolleyThe South Extension Trolley has beendedicated for the SLAR equipment. The trolleyis parked in the ESA (East Service Area) wherethe testing and maintenance on the SLARequipment is performed.

The South Extension trolley (Fig. 2) is modifiedby the installation of two (2) D20 hydraulicdrive units, connected to the existingauxiliaries. These units are used to power thehydraulic devices on the new tubes. Thehydraulic power supply (Charging Pump) ofeach hydraulic drive unit are interconnected toprovide backup.

SLAR tool calibration devices are provided ineach F/M head (Calibration sleeve) and on thetransport trolley (Calibration channel betweenthe two (2) D/Ms').

The trolley catenaries and power track remainunchanged.

2.1.3 Delivery Machines DrivesThe SLAR head can handle the normal D/Mhead operations to automatically manipulatethe channel components. The extrarequirements of the SLAR head has led to theaddition of seven (7) new drives in addition tothe seven (7) F/M drives for a total of fourteen(14) drives per side. The new drives are:

1.2.3.4 .5.6.7.

Turret RotaryTelescopicUmbilical Cable DrumRam TetherPiston TetherM/T rotaryF/T rotary

The controls for the original seven (7) F/Ms'drives have been replaced with controls forfourteen (14) DC servo drives complete withbrakes and encoders except for the umbilicalcable drum motor which has no brake. Sincespare conductors are not available in thecatenary or power track for the drives,multiplexing is used where drives share theexisting conductors. The maximum number ofdrives required for any SLAR operation is three(3).

The system is normally operated in a semi-automated mode with the operator initiating theoperating sequence stored in the F/Mcomputer.

2.1.4 SlarToolThe SLAR tool (Fig. 5) used to locate andreposition the Spacers consists of a hydraulicbending tool to raise the P/T, eddy currentsystem for locating the spacers and measuringthe gap between the P/T and CAT, ultrasonicprobes to detect cracked hydride blisters on thelower section of the P/T, and LIMs' (LinearInduction Motor) to reposition the spring fromwithin the P/T.

The operation of the SLAR tool in the channelis controlled by the SCADA (SupervisoryComputer And Data Acquisition) computerwhich is slaved to the D/M control computer.The LIM power, control and inspection signalfrom the tool use the umbilical cable connectedto the tool. The one hundred and ten (110) ft.of umbilical is wrapped around the cable drumas the tool is extended and retracted in thechannel and is connected by slip rings at theCable Drum.

2.1.5 Operation ControlThe SLAR equipment is maintained and testedin ESA prior to going to the unit for the SLARcampaign by the F/H and SLAR personnel. TheSLAR campaign at the unit is an involvement ofthe various work groups, required frompreparing the Reactor Vault to the operation ofthe SLAR equipment.

The operation of the SLAR equipment isperformed by F/H SLAR trained operators whoare responsible for all D/M, trolley and bridgeoperations. After the D/Ms' have been homedon the channel, its C/Ps' and S/Ps' removedand the channel defuelled the control of theSLAR tool is transferred to the inspectionpersonnel on the SCADA system to process thechannel. The SCADA system however remainsslaved to the F/H control computer. When theprocessing of the channel has been completed,the control is returned back to the SLARoperator to return the fuel to the channel andclose it up.

2.1.6 Reactor PHT systemDuring SLARing, the PHT (Primary HeatTransport) System is put in a low level drainstate as the D/Ms' can only operate at amaximum channel pressure of thirty two (32)psig. The limit is set because of the pressure

94


force exerted on the turret seal.

2.1.7 System SurveillanceTo monitor the D/M D2O supply pressures andrecord the performance of the various drives, adatalogger is used with each machine. This isin addition to the information available from theF/H computer and Operator logs. Thedatalogger has fourteen (14) input channels formonitoring the analog signals from the D/Ms',the data from which is graphically displayed onits computer monitor and is stored to its harddrive (Fig. 6 & 7 show the graph of some ofhard drive data). The collection capability hasproved to be a very useful tool in monitoringand diagnosing the problems with themachines.

2.2 Operation Problems

2.2.1 Temperature Transient on ChannelWhen SLARing, on three (3) occasions thechannel temperature went up from thirty (30)degree Celsius to sixty eight (68) degreeCelsius after the fuel was returned to thechannel. The channel temperature wasmonitored for five (5) hrs. until it returned backto normal. Inspection of the fuel from one ofthe channels showed no visible heatdeformation or damage, and all bundlesinspected had normal wear patterns on thebearing pads and endplates.

The possible cause of the transient wasattributed to the stagnated flow in the channel.With the PHT system in the low level drainstate for SLAR, the flow in the channels isreduced. The SLAR tool, when inserted in thechannel restricts the flow. When the fuel isreturned to the channel the flow stagnates.The heat is eventually removed from thechannel via IBIF (Intermittent Buoyancy InducedFlow) and rejected to the header/boilers. Thiscauses a gradual rise in the channeltemperature over a short period of time till theflow in the channel gets established.

As a result of these temperature excursionincidents, flow switches are being installed tomonitor the cooling flow to the F/T. Also theCOT (Channel Outlet Temperature) is monitoredfor a period of time three (3) hrs. after the fuelhas been returned back to the channel.

2.2.2 Crud in the ChannelThe D2O for the operation of the F/T and M/Tis taken from the header supplying theCirculating Pumps which in turn is supplied by

95

the D/M heads. This supply is unfiltered andcontains the crud that gets transferred to thehead from the E/F (End Fitting) when the C/Ps'(Closure Plugs) are removed. The crud in thesystem effects the PRVs' (Pressure ReducingValve) pressure settings controlling the M/Tand F/T operations. The F/T side sees morecrud as it gets removed from the channel bythe fuel bundles in the defuelling operation.The crud problem gets worse the longer theunit is down.

To handle the crud problem a spare D2O filterassembly has been installed on the D/M trolleyof the F/T side handling the most F/Toperations. This has reduced the problemswith the PRVs'.

2.2.3 Slar Drive SystemThe SLAR drive system on each Deliverymachine consists of fourteen (14) individualpermanent magnet DC motors that can beoperated with a variable speed and torque.Each motor has a tachometer and a brake(Except the Cable Reel). The motors are fedfrom three (3) DC buses using a multiplexsystem. Eight (8) of the motors are Contravesand six (6) are PMI. A number of problemshave occurred with the PMI amplifiers andmotors.

AmplifiersThe PMI Model #CX amplifiers outputTransistors (PTC7001) and Optocoupler hadfailed on numerous occasions. The PMIamplifiers were obsolete and could not berepaired. The PMI amplifiers have beenreplaced with standard off the shelf CopleyControl amplifiers. These amplifiers required anedge filter to increase the inductance in themotor circuit and an electronic potentiometermodule to replace the built in peak currentpotentiometer.

The control OP's had to be revised to limit thetorque because these amplifiers are capable ofsustaining higher current. The total cost of thismodification has been less than the originalprice of the PMI amplifier.

SLAR Umbilical Cable Drum Drive MotorsThe cable drum (Fig. 8) must be controlled insynchronous with many other drives in order tokeep the umbilical cable under a relativelyconstant tension while releasing or taking incable. The control signal is derived by a singleloop micro-controller fed with appropriate inputsignals. An LVDT (Linear Voltage Displacement


Transducer) is used to monitor the cabletension and its conditioned signal is fed to oneinput of the loop controller. During turret,telescopic or Z motions, the controller uses thefeedback from the LVDT to readjust the cabletension. In M/T axial motion the controller willratio the tachometer feedback from the moduleaxial drive and use the LVDT signal to trim themotor speed to achieve the correct tension.

This control system places a heavy demand onthe cable drum drive motor. The motor mustachieve high acceleration rates and outputtorques as well as rapid changes in direction.Monitoring the tachometer feedback has shownthe motor operating at two thousand (2000)RPM in one direction and two (2) seconds laterrunning at two thousand (2000) RPM. in theopposite direction.

Numerous failures of the original cable drummotors led to a detailed analysis whichrevealed the following deficiencies:

• The motor power rating of 1.41 HPwas less than the expected load of 1.6HP.

• The motors were overheating due tothe light thermal mass of the pancakestyle armatures with printed circuitwindings and the high peak current.

• High pulsed current which wereconstantly reversing were suspected ofpartially demagnetizing the motorpermanent magnets.

The motors were replaced with 1.66 HP unitswith conventional copper wound armatures anda higher current rating. These new motorshave performed well with no failures to date.

2.2.6 Multiplexing SystemThe problem also existed with the design of themultiplexing system, the amplifiers were beingpowered before the motors contactor waspicked up. This was suspected of causingsome of the PMI motor failures. To correct theproblem, timers have been installed that delaythe power up of the drive amplifiers until themultiplexed motor contactor relay is picked up.

2.2.7 Vault Modifications for the DeliveryMachinesPrior to operating the D/Ms' in the unit vault, ithas to be modified to extend the machines.This requires removal of rails, catwalk and

96

ACUs' (Air Conditioning Units) in the vault.The work requires about ten (10) days tocomplete during which the inspectionequipment for the SLARing is also setup insidethe vault.

2.3 Maintenance Problems

2.3.1 SLAR Tool InstallationThe maintenance of any equipment insidecontainment poses a problem. The SLARsystem has been no exception to this. Thecomplexity of the system requires a number ofelectrical, mechanical and control systems towork together that complicates the diagnosingand maintenance of this equipment even more.

SLAR Tool ChangesThe SLAR tool and umbilical cable must bereplaced with a new unit several times in eachoutage. The SLAR tool is changed in aV-trough mounted on the trolley between thetwo (2) delivery machines. To avoid problems,tool changes must be completed by trainedpersonnel following a complex and detailedprocedure. Some of the problems encounteredwith tool changes at Bruce A are:

• The umbilical cable wrapped around thetether cable as it was fished throughthe M/T. This eventually led to thetether cable cutting through theumbilical cable.

• The coaxial and twisted pair signalcables from the eddy current andultrasonic probes in the tool must beterminated using small, fragileconnectors in an awkward locationbeneath the cable drum. Poorconnections have led to erratic signalsfrom the SLAR tool probes causingdelays in the channel processing.

• The stainless steel sleeve and pottedsection of the umbilical cable are a verytight fit inside the M/T ram # 1 . Slightlyoversize sleeves and out of roundpotting have made connecting the toolto the ram very difficult.

• The SLAR tool and umbilical cable areheavy and awkward to handle in thelimited space available on the trolley.At least one back injury has occurredwhile attempting to manoeuvre the toolinto position for installation in thedelivery machine. To correct this, the


working platforms on either side of theCalibration channel had to be loweredto provide better access to the V-trough to install the tool. Also, a jib hasbeen installed to handle the tool.

The sag/droop developed in the M/T and F/Twhen left in the extended positions gaveproblems with sticky M/T and F/T operations.This was resolved by leaving the machines inthe parked or collapsed stage. Minor problemsresulting from these were difficult to diagnoseat the time again because of the complexity ofthe machines.

The F/H SLAR personnel have been used towork on the equipment. The maintenanceprocedures had to be tested and revised in thefield when the work was done as there was nomockup to train the personnel at the expenseof time and dose.

SLAR was supposed to be a short term project,but has extended longer than expected. Thisposes problems with the availability of thespares such as amplifiers and motors for thedrives and wearing of the equipment, as it wasnot designed to operate for this extended SLARoperations.

3.0 ConclusionThe SLAR equipment has performed well incarrying out the three (3) Slar campaigns onUnit 3 & 4. The system is scheduled for theSLAR campaign on Unit 3 in May of 1996 andUnit 4 sometimes in 1 997. The duration of theSLAR project has created some problems withassigning permanent SLAR personnel to operateand maintain the equipment but because thepersonnel assigned for SLAR are from the F/Hgroup the experience remains in the group.

SKOUT HOMrNG AMGlOMMGUICIMIIISM

1URMTHOUIWC

SHOUT lociOIIVE

OnivtHOUSING

K«H HPIItDRIVE

MM limitDmvc

COIINTfRWIIGK]'1SSFMBIV

M M I t BESV'PODt

FIGURE 1. SLAR DELIVERY MACHINE HEAD

97

FIRSTINTERNATIONAL CONFERENCE ON CANDU FUEL HANDLING SYSTEMS. MAY 1996

jcl'5 SU»n ItfAVY WATjH AI

FIGURE 2. 8LAR SYSTEM INSTALLED ON THE SOUTH TUEttlKO TROLtEY

FUELBL'KDUS |

_ \ _ '

EMOflTDNC

\SHum10CUKC 1UD

c • \ — *

SS 13 JR LOST MOTIONl l l fE* UIEVE _ ."<WH.IU.tOMS-o.MO

MAB Sl"VE r

rrr.| ._

C UMODULE TVSE DPIMT CUHPDRIVE SMC

U1CKinosEoi

HEAD LOCKED ONTO END FITTING.MODULE TUBE FULLV TELESCOPED OUT

uitno

tl iTETHER C»8lf DAIVE

simnc i « i f

'. rrorCSLUII

SUEVETELCSCOPED WJ TO POSITIONSLEEVE HEAD AT OE-LATCH POSITION

^fasco*E ofimtuvmc mn

musMc

16 I-OM W FUCL LAICM CM)• •1 m M. SPACER END

imEU M U TETHEHCAIIE

ocMil

IRCTRACTfOI

7TSII9IN MAX iLAtt 1OCH. STHC«C

s rr

) flue sun RAM10CKI1C IUC 1001 HUH* ADVANCIOI

MUttnniu

UMtlllCUMILE

LINER f ULir AOVANCEO AND LOCKEDINTO FUEL CHANNEL END FITTING

sC\

UICH tcieuo*mssuu win

MODULE TUIE LATCH OPENED.RAM ANO SLAB TOOL f UUY AOVANCED

IIKiflsror

3. MODULE TDBE OPERATION

98


FUEL TUBE FULLY TELESCOPED IN TO TRAVERSE FROMTROLLEY TO REACTOR FUEL CHANNEL ICOUPIEO).

rtuc[aeonsluo

iJ"&tf8HrBlli /

TELESCOPE DRIVE-UOmNtKlTT STOP

fUEL TUBS FULLY TELESCOPED OUT FOB TUHRET ANOMAGAZINE DOTATION. HEAD LOCKED ONTO FUEL

CHANNEL3 11 V* AT 0 4« lK. AT

Sf *C£» END LATCH END

COUPLING U>EH \{L aimSLEEVE CCILM HEAD,

r

I1-FLTELTUBE SLEEVE TELESCOPED IN (COUPLEO) UNTIL

SLEEVE HEAD IS AT DE-LATCH POSITION. UNEHADVANCED UNTIL LINER HEAD STALLS AGAINST

SLEEVE HEAO SLEEVE AND LINER ROTATED TO LOCKLINER INTO FUEL CHANNEL. FU£L BUNDLES PUSHEDINTO LINER BY SLAR TOOL AND RAM OF OPPOSITE

SLAR FUELLINO MACHINE.MTCK FIHMftS

(OPENED)

sunmot

LOWER THREE ROWS OF rUCL CHANNCtj OMIT. THE FUELruarn umouruo ANO IHISOOVED •< TO THE ot LATCH

rosmoN usuro ixt t r a m SLKVI

LATCH FINGERS OPENCD TO RELEASE PISTON. PISTONADVANCED TO PUSH FUEL BUNDLES INTO FUEL

CHANNEL.

PIGDRE 4. TOEL TDBB OPERATION

OVERALL DIAGRAM OF MARK III SLAR TOOL AND COMPONENTS

ISH PLATE

^-BLISTER I

• f ;

:REE END BEARNG PAD1 rFREE END SPHERICAL JOINT

/ /-COIL FORMER

RAM END BEARING PAI

RAM EM) SPHERICAL JOINT

CENTRE BEARING PADS

/-CENTRE LM

:REE END UM

MOLXLEOL FORMERS

IVERSAL JONTS

5 . 8IAR TOOL

99


WIIST M. T OP CHECK FILE 1449IM - RID - 17 APR 95

t,.5 !_ . ._ .18:4 J:58.GG ] 8:4:1:35 1 1

18:42:0(5.73 lfl:42:22.ri 18:42:38.7 1B:42:55I2 18:43:10.94 18:13:27.03

MOA TAtril ' L'llA TACII o M/T TORQUE a M T PRESS x Si'EliU COM

W I)-BEND

FIGURE S. MODULI TDBE GRAPH (DATALOGGER)

10 WEST FUEL PUSH PILE 14-50U4 17 APR 95

- I

f

19:11 :394

¥

19:12

!

59.04

aSKB0

19:14

T

19.39

JfI

f

19:15

rLyKm

39.09

• i

f

19:16 : 59.34

*****

19:13:

aw**

18.76

1

1 9 : 1 2 : 1 8 7 2 1 9 : 1 3 : 4 0 1 2 1 9 : 1 4 : 5 8 . 6 7 1 9 : 1 6 : 1 9 . 2 4 1 9 : 1 7 3 9 . 3 8 1 9 : 1 8 : 5 9 . 1 0

MDA TACH ^ CRA TACI1 O M/T TORQUE a M/T PRESS x SPEED COM

? U-BEND

FIODRE 7. FUEI. TUBE GRAPH (DATALOGGER)

100


MTMMY SUPPORT(MFEHENCt ONIYI

UMMJUL CULE

EA81Í OMIU SUPPORTаМСХЕТ KFTMiD

FIQORB t . CXBLE DRUM

SËXT PACKS)toft

101

CA9700729


MINI-SLAR DELIVERY SYSTEM

by

D. AlsteinOntario Hydro, Bruce AP.O. Box 3000Tiverton, ON NOG 2T0

and

K. DaltonSpectrum EngineeringP.O. Box 687Peterborough, ON K9J 6Z8

ABSTRACT

In the Spring of 1993, a need to completeSpacer Location and Repositioning (SLARI onthe Bruce 'A', Unit 1 Reactor was identified.An alternate SLAR delivery system wasrequired due to conversion constraints thatprevented the existing Bruce SLAR Systemfrom being used in Unit 1.

A Portable SLAR Delivery System called MINI-SLAR Delivery System was developed,designed and fabricated in a 14 month period,then used to successfully SLAR 109 channels.

The system is a portable remotely operatedNuclear Class I registered fitting that isindependent of the Fuelling Machine, allowingthe station to continue normal Fuelling andMaintenance activities. It is designed to aLevel "D" faulted condition of HPECI Pressurethus minimizing PHT Heat Sink configurationrequirements and minimizing outage set-uptimes. The system is based on a modulardesign allowing for easy fabrication, assemblyand repair. It consists of a Snout Assembly, aClosure Plug Assembly, Shield Plug Assembly,SLAR Ram assembly. Work Table Assemblyand Control Panel. Controls are through aProgrammable Logic Controller with softwaretested and certified to a Software QualityAssurance of Level III.

INTRODUCTION

In the spring of 1993, the Canadian Nuclearindustry was rocked by Ontario Hydro'sdecision to cancel the retubing of Units 1 & 2

Reactors at Bruce 'A'. It was decided that Unit2 would be "laid up" beginning in September of1995 and Unit 1 would be run until it met itsFuel Channel Life Limits.

In developing a "Life Cycle Management"Strategy for Unit 1 Fuel Channels, the stationquickly realized that it must address the graveissue of potential pressure tube to calandriatube contact that results from channel annulusspacers moving from the design location, orface a shutdown of that unit as early as theSpring of 1997. At Bruce, Units 3 & 4, SpacerLocating and Repositioning (SLAR) has beenquite successful and is performed using twospecially converted fuelling machines. Thissystem could not be used, however, as theFuelling Machine Trolley could not reach Unit 1and the equipment could not be converted toanother trolley in time to meet projectcommitments. Thus was born the need for analternate delivery system.

Although the need for SLAR of Bruce Unit 1fuel channels was first identified following thecancellation of the proposed retubing of Unit 1in the spring of 1993 work on a deliverysystem did not begin until the summer of 1994.In spite of a late start to the design process andthe complex challenge of designing, buildingand commissioning a replacement for the SLARdelivery system in just 14 months, the projectsucceeded in completing SLAR of 109 channelsin unit 1 in 31 days of SLAR processing at avery substantial saving to Bruce A. Thepurpose of this paper is to describe the processfollowed to develop the mini-SLAR system, todiscuss the distinct aspects of this approach

103

FIRST INTERNATIONAL CONFERENCE ON CANDU FUEL HANDUNG SYSTEMS. MAY 1996

and examine what made this projectsuccessful. The technical description of themini-SLAR mechanical design has beendocumented by others [1.] and this paperconcentrates on the organizational aspects ofthe project.

The mini-SLAR delivery system is a small scalefuelling machine-like device. It was used in1995 to deliver a SLAR tool into a fuel channelfrom tooling mounted on a fuel channelmaintenance platform. The designrequirements of the delivery system were tomanually home and clamp onto a fuel channelunder shutdown conditions, extending the heattransport pressure boundary, remove theclosure and the shield plug and install a linersleeve into the end fitting from either the inletor outlet ends of a fuel channel. The targetchannel processing rate for the system was 2channels per day. Both the SLAR processingand the delivery machine systems performedextremely well overall. The average timerequired to perform SLAR operations once onthe channel was 1.8 hours, and the timerequired to move between sites was 1.3 hours,with an average 4.3 hours required for toolcalibration, equipment maintenance, toolreplacement and other support activities [2.].SLAR processing rate was slow initially, andgradually improved as the crews became moreexperienced with the equipment and resolvedsome initial operational problems.Ultimately the SLAR processing of Unit 1 wascompleted in less time than originallyscheduled, and processed 25 percent morechannels than originally planned. This hasmade it possible for Unit 1 fuel channels tocontinue to operate safely until the year 2000.

What distinguished the Mini-SLAR project fromother similar tooling development jobs pursuedat Bruce, such as West Shift, SLAR andRetube, was that the project managementgroup was a site based team, highly accessibleto and in regular contact with station personnelas well as the design organizations. Theproject team strongly influenced the design, butmore importantly ensured that the designadapted to the needs of the site organization.The availability of existing retube tooling as abasis for some of the design and developmentgave the project a running start in the testingarea, heading off many potential problems andguiding the direction of the project based on

solid practical experience. The opportunity toassemble and play with simple and sometimescrude prototype equipment may not have been"doing it right the first time" but it led to somevery interesting results.

- "Americans don't do it right the first time,they just don't. But when that second timecomes around, not even the stars are out ofreach."

"Doing it right the first time and zero defectsmay be expectations, but as messages toAmericans they're more debilitating thanmotivating; they make Americans feelcontrolled and restricted."

from the book INCREDIBLY AMERICAN,RELEASING THE HEART OF QUALITY

CONCEPT DEVELOPMENT

Upon determining the need for a DeliverySystem, a "Skunk Works" Team of Engineers,Technical Staff and Trades was assembled.The team developed an approach of mountinga number of existing Retube Tools on a slidingplate. A fixed plate with a snout that attachedto the End Fitting would remain in a permanentposition with the sliding mechanism behind toalign individual tools to an access port in thefixed plate. An ' 0 ' Ring Seal would form thepressure boundary between boundary and thetwo plates. With little more than sketches, thetwo plates were fabricated and rudimentarycycle tests and pressure tests were undertaken.

FIGURE 1SLIDING PLATE

TEST ARRANGEMENT

104


Having passed these tests, the team now hada concept tool, a reference that was physical,against which improvements could bemeasured and could be used to validate thefeasibility of the project. The team could nowproceed with confidence in scoping out theproject, seeking out proper approval forfunding, then move on to designing andbuilding a prototype.

PROJECT TEAM

The project team members had four distinctroles, to manage, design, build and implement.While one person was primarily responsible foreach of these areas, the entire project teamwas collectively involved in key decisions andmet frequently to review progress. The teamconsisted of staff with a varied backgroundfrom fuel handling, operations, retubing, SLAR,fuel channel design, and mechanicalmaintenance.

Members of the project team brought widelyvarying individual view points and experienceand as a result contributed to a broader internalreview of the design which was constantlyevaluated by the team in addition to moreformal review by staff in the customer'sorganization. In addition, the team membershad a wide and varied range of personalcontacts. This provided a source of informalcommunication that was important to anunderstanding of the customer's need at alllevels.

CUSTOMER CONTACT

The close working relationship betweenmembers of the project team and station staffresulted in ongoing formal and informaldiscussion of design requirements and designpreferences. The articulation of designrequirements required frequent meetings anddiscussion with station staff at every stage ofthe project.

Numerous design ideas resulted fromdiscussion with station staff during designreview, information presentations and training.More importantly, many of the staff who wereeventually involved in the end use of thesystem, were also involved and hadopportunities for input during the design andtesting process. Frequently, the groups of staff

visiting the training and mock up facility towitness prototype system operation, gained animpression that the design had already beencompleted and was not subject to change.This impression was discouraged. Suggestionsfor improvement were encouraged and actedupon whenever possible. The site organizationwas encouraged to develop a sense of designownership and involvement.

USE OF EXISTING FACILITIES ANDCOMPONENTS

Site partnership in the design process waspossible because of the existence of manytools and components that were used in thedesign and testing of the delivery system. Keysite individuals with direct experience indeveloping tools of this type were involved intesting and experimentation with variousequipment already available to the site. As thedesign progressed off site, it was possible fora parallel development effort to proceed at thestation.

DESIGN FLEXIBILITY

It was necessary for the design to adapt tochanging design requirements resulting fromfield development feedback and evolvingcustomer requirements. Examples of this arethe addition of shield plug and liner handling tothe basic function of the system, the increaseof design pressure from 5 to 8 MPa, shorteningof the system to fit into airlock 2, as well ascomplete redesign of the support structure, thesnout locking mechanism, and the sliding platedrive system to resolve performance problems.The use of a PLC control system to automatesteps in the process allowed for the closureplug removal sequence to be modified andtested during the final weeks before theequipment was put into service. Byimplementing a modular design of hardwareand software, it was possible for hardware andsoftware development and testing to proceed instages. The modular nature of the deliverysystem is illustrated in the following overview.This shows how the shield plug, closure plugand SLAR ram modules are independentlymounted stand alone devices built into amodular support frame.

105


INFLUENCE OF PROTOTYPE TESTING

The use of a complete prototype system fortesting and training was an invaluable aid toimplementation as it allowed many potentiallyserious problems to be resolved ahead of time.Examples of this were the need for a Z drivebrake on the shield plug removal tool, the needfor rotary stops on the shield plug drive toprevent over torquing of the latch mechanism,and the identification of various designimprovements to provide smoother, morereliable operation.

The control system was extensively prooftested on the prototype system to establish itsperformance characteristics. This proved to bevital to fine tuning the control software forsmooth and consistent operation of the system.Although apparently straight forward, theprotective stops and interlocks built into thecontrols for equipment protection createdpractical performance problems which wereresolved through a process of testing andmodification. A list of hardware and softwaredeficiencies were gradually resolved through aprocess of realistic full scale simulation of thefield operation of the equipment.

Training of 40 fuel handling operators andmechanics on the operation of the system waspossible using the realistic rehearsal facilitiesused for prototype testing. These included afull scale mockup of the reactor vault,maintenance platform, and F/M bridge fulllength pressurized channels as well as drycutaway channels for viewing the internaloperation of the equipment, and a remotecontrol console, complete with viewing andcommunication systems.

Integration testing using the actual SLARinspection system was made possible by theparallel development of a transportable SLARcontrol trailer. The interface with the SLARcontrol system was designed to emulate theexisting fuelling machine interface to providethe least possible redesign of the existing SLARsystem.

Bruce A has been historically dependent onoutside sources for design support and has notdeveloped a strong design capability or culture.Site is more comfortable with practical resultsrather than a dependence on analysis and

procedures of design. From the outset, the siteteam based many decisions on actualperformance of field equipment, and on theintuition and judgement of field personnelrather than accepting the paradigms of atraditional design approach.

The sliding plate mechanism is a case in point.This is an unconventional design, notrecommended by various designers, butsupported by the site team as a simple solutionto the problem of tool indexing. In spite ofconcerns raised by experts to the advisability ofthis approach, the site team went ahead tofabricate and demonstrate the design. In amore traditional conservative design culture thissimple approach would have been rejectedalthough in practice it has performed reliablywith minimal development. This points outhow a risk averse tendency in our traditionaldesign approach can result in missedopportunities. When necessity is the mother ofinvention, some interesting and usefuldiscoveries can be made.

Is this just luck? Is it equally likely that in afuture design project, this design approachwould fail to produce acceptable results? Theanswer to this question may be yes, but if welook at the track record of the "conventional"design approach we see that there are anynumber of unsuccessful designs that havefailed to produce results in spite of the bestavailable design talent being brought to bear.

We believe that in the field of nuclear stationmaintenance equipment design, the mostdemanding problems are not technical ones.The technology associated with what we needto do is not only readily available, but in mostcases it has become relatively commonplace.The most important issue is design acceptanceby field personnel and the field implementationorganization. As designers, by focusing onmore elegant technical solutions we riskalienating the people who have to use ourproducts at the end of the day.

The Mini-SLAR project is an illustration of analternative organizational approach intended toaddress field involvement and acceptanceissues. There were both positive and negativeaspects to the project, and some importantlessons to be learned.

106


XWEH

CM OS W< ><

W J tO« CO3 I ><O M OSM Z Wh H>

S

107


The project team was proud of the fact that inspite of a short implementation schedule, itwas able to deliver a system capable ofreplacing the existing fuelling machine basedSLAR delivery system, with a projectjustification, design build and test duration ofapproximately 14 months.

The business case for Mini-SLAR was based ona need to SLAR an estimated 110 fuelchannels in Unit 1 by spring of 1995 or facethe need to perform a series of CIGARinspections, scrape samples and fuel channelreplacement of 27 channels at risk. Based ona realistic estimation of the cost of channelinspections and replacements, the mini-SLARproject was expected to save Bruce A asignificant capital cost associated with channelreplacement, and 100 outage days based onthen current multi fuel channel replacementtechniques. The actual performance of thesystem was somewhat better than expected.A total of 109 channels were SLARed in 31days of SLAR processing time, exceeding thetarget rate of 2 channels per day by asignificant margin. All but one channel requiredspacer repositioning to extend its operating life[2]. An estimated 27 channel replacementswere avoided, and work was completed withminimal impact on the critical path duration ofa planned 103 day boiler chemical cleanoutage. Total direct and indirect savings toOntario Hydro are considerable and Hydro hasgained some very valuable experience.

Like any successful job there were manyhurdles to overcome, the most demanding onewas project schedule. Many activities had tobe performed in parallel in order to meet therequired completion date. The resultingprocess operated in the reverse of the usualdesign process. Fabrication and testingpreceded the formal design activity, and the GECanada and NTS SLAR design teams had tofrequently adjust their priorities to respond tothe practical requirements of the fabricationand testing schedule. Equipment wasassembled and tested in an emerging prototypeform and the design organization, while tryingto analyze and formally document the designwas at the same time responding to frequentrequests for changes from the fieldorganization. The flexibility demonstrated byour key designer, GE Canada, while no doubtfrustrating for them, was key to ensuring the

final product was accepted in the plant. Inspite of the effort made to maintain frequentcontact with the design group, it was apparentthat the design process could have been moreefficient if the designers had been workingtogether with the fabrication and testing teamat the station. The circumstances of thisproject allowed the design and fieldorganizations to be on a more equal basis thanmost design build projects. From thebeginning, the field organization had a roughprototype to work with using various retubeequipment on hand. Whenever possibleexisting components were used to reduce thedesign and test time. Significant upgrades ofthe equipment were required but many of thebasic mechanical concepts were available fortesting before the design process started. Thisallowed the field team to begin hands onexperimentation and to provide early feedbackto the designers to correct various deficienciesin the equipment and to refine theunderstanding of the operating characteristicsof the system. Valuable design input wasprovided by the field organization based on theexperience with the mechanical prototypesystem that contributed to a more useablepiece of equipment when it went into service.The control system, and the SLAR inspectionsystem were dovetailed into the design later inthe project schedule, and required an intensededicated effort by site staff as well as by NTSSLAR site team and GE Canada designers.Because of the modular nature of the design itwas possible to marry the control systems forthe delivery system and the SLAR inspectionsystem with a minimum of testing problems.Much of the credit for this accomplishment isdue to the effective communication betweenGE Canada and NTS SLAR designers whodeveloped a detailed interface specification atan early stage in the design.

SITE FABRICATION AND TESTING

Fabrication and testing of the prototypedelivery system at site resulted in a muchgreater site awareness of the project althoughnot everyone was comfortable with the waythe work was allocated. Half way through thefabrication and testing process, the team hadto negotiate a revised work assignment toreflect the emerging trend toward greateroperations staff participation in projects of thistype. This was disruptive to progress, but in

108


the end the staff who were involved in thedevelopment and testing were able to beclosely involved in the field support of theequipment and were largely responsible forfinding and efficiently solving problems duringimplementation. The construction andoperations trades staff involved in this jobworked closely and effectively togetherexhibiting a high standard of professionalismand dedication.

At the time the Mini-SLAR system was built theBruce site organization was not authorized tofabricate this type of nuclear class 1equipment. However, the MCCR granted aconcession for the assembly and testing of thepressure boundary, and the fabrication of theNF support structure for this project. In future,a site fabrication certification will make theassembly of this kind of equipment morepractical. The advantages of site fabricationand assembly were the responsiveness toschedule requirements, direct supervision andhigh standard of workmanship. The fabricationof the equipment on site contributed to a senseof design ownership and a highly effectivereview of manufacturability of the design. Thehands on involvement of field personnel led toa detailed understanding of the maintenancerequirements for the equipment.

INDEPENDENT OPERATION

The use of the station fuel handling systems forreactor maintenance and inspection has been acommon practice to the extent that the fuelhandling system availability for fuelling isbecoming restricted. The original strategyconsidered for Unit 1 SLAR was based onextensive use of the fuel handling system forshield plug and liner sleeve handling. This wasreviewed with station fuel handling staff at theoutset of the project and it became immediatelyapparent that a design strategy with the leastinvolvement of station fuel handling systemswas a basic design objective. Ultimately themini SLAR system was able to replace thefuelling machines for all aspects of the projectexcept fuel removal and replacement.Defueliing was completed as a batch processusing flow assisted defueliing methods in aperiod of 17 days following shutdown. Fuelreplacement was accomplished by transferringfuel from Unit 2 during the Unit 2 layupdefueliing at a substantial fuel cost saving.

Independence from station systems wasassured by providing independent powersupplies, and the capability of operating onchannel independent of the state of the heattransport system. The system was designed towithstand fuel channel pressures up to 8.3 MPaand was capable of operation of up to 1.38MPa. This made it possible to perform SLARoperations during all phases of boiler chemicalcleaning. Defueliing at the start of the outagealso made the vault accessible throughout thebalance of the outage up to the start ofrefuelling. This provided unrestricted accessfor other outage work to proceed in parallel.

FUTURE PROJECTS

The mini SLAR system has numerous potentialfuture uses. In the immediate future it is beingadapted for mounting in the fuelling machinecarriage with a remote homing and lockingcapability, for use in draining Unit 2 during layup. In addition to improving drawing efficiencythis equipment is also useful in providingpressure relief during fuel channel isolation.This alleviates the need for the F/M duringchannel isolation. A second mini SLAR systemwill be built and used to perform SLAR on Unit4 in 1997, using two SLAR systems in parallel.As an upgrade for this project, the Unit 4system will be designed to defuel the Unit 4channels by pushing fuel into a fuellingmachine at the opposite end of the channel.This method of defueliing allows for economicalfuel shuffling, avoiding the disposal of partiallyutilized fuel. Defueliing during a shutdown withthis technique is much more efficient than thefuel grappling alternative. It avoids the need tooperate the heat transport pumps to permitflow assisted defuelling, and can be performedfrom either the upstream or the downstreamend of a channel.

The use of a mini-SLAR type device for generalfuel channel maintenance and inspection hasmany advantages over current practice. Theequipment can be picked up by the fuellingmachine bridge, and requires a very little vaulttime to install. Once in place, the equipmentcan deliver a variety of special purpose tools toaid in inspection and maintenance activitieswhile placing few demands on the existing fuelhandling system. Since the delivery system ismodular in design, it can be easily reconfiguredto accommodate special purpose devices in

109


place of or in addition to the existing modules.Examples of some potential future applicationsare the delivery of tools for channel inspectionand gauging, isolation, closure seal repairs,pressure tube wet scrape, fret blending andreplication, and visual inspection. In the longerterm similar devices may be used to assist infuel channel replacement.

CONCLUSION

Site based project teams are assuming animportant role in providing specialized designand equipment needs for OHN. The mini-SLARproject demonstrated how a site projectmanagement team can be more responsive tothe specific needs of an individual project, andprovide greater levels of site involvement in thedesign while maintaining a high standard ofquality. Emphasis on using available resourcesto speed the development cycle, keeping thedesign simple and flexible and encouraging verybroad participation by the user organizationwere key elements of the success of thisproject. Future projects should address theneed for qualified site fabrication capability,and a closer working relationship between thedesign organization and the fabrication andtesting group. These factors would improvethe efficiency of the design feedback processand ensure an effective site fabrication andtesting process

REFERENCES

1. Design of the Mini-SLAR System for BruceA, M. Gray, GE Canada Nuclear Products,Candu Maintenance Conference Oct., 1995.

2. Mini-SLAR of Bruce NGS A Unit 1 PressureTubes in 1995, S. Motomura, Ontario HydroNTS SLAR, ISMP-BGA-00773.1-14, Dec 3 1 ,1995

110


CANDU 9 FUELLING MACHINE CARRIAGE

D.J. Ullrich and J.F. Slavik

AECL CANDUCANDU 9 Reactor and Fuel HandlingSaskatoon, Saskatchewan, Canada

CA9700730

ABSTRACT

Continuous, on-power refuelling is a key featureof all CANDU reactor designs and is essential tomaintaining high station capacity factors.

The concept of a fuelling machine carriage can betraced to the early CANDU designs, such as theDouglas Point Nuclear Generating Station. In theCANDU 9 480NU unit, the combination of a mobilecarriage and a proven fuelling machine headdesign comprises an effective means oftransporting fuel between the reactor and the fueltransfer ports. It is a suitable alternative to thefuelling machine bridge system that has beenutilized in the CANDU 6 reactor units.

The CANDU 9 480NU fuel handling systemsuccessfully combines features that meet theproject requirements with respect to fuellingperformance, functionality, seismic qualificationand the use of proven components. The designincorporates improvements based on experienceand applicable current technologies.

1. INTRODUCTION

CANDU reactor operation is based on the conceptof continuous, on-power fuelling, which is carriedout by the remotely controlled Fuel HandlingSystem. The fuelling machine head, a pressurevessel which contains new or irradiated fuel duringits transport between the reactor and the transferports, can be supported either by a bridge system,such as that used in the CANDU 6, or by a mobilecarriage such as in Douglas Point NGS. For theCANDU 9 480NU design, a carriage system hasbeen adopted.

An important consideration in the selection of thecarriage system for CANDU 9 was having a designthat would lend itself to seismic qualification forfuture sites with potentially higher seismicities thanthose of the previously constructed CANDUstations. Other design goals included reduced

construction time and improved on-powermaintainability of the carriage and the fuellingmachine head. Simplicity and the maximum use ofstandard, commercially available componentswere also a part of the design intent. Supplier andoperator input was incorporated.

In the following sections, an outline of therequirements applied to the fuelling machinecarriage concept is provided, along with a detaileddescription of the carriage system functional andstructural design. A summary of the benefits of thecarriage system is included.

2. DESIGN REQUIREMENTS

The key CANDU 9 480NU project requirements,related to the fuelling machine carriage systemperformance and functionality, included:

i) meeting the fuelling demands of the CANDU 9480NU reactor in support of the target lifetimecapacity factor of 90%;

ii) access, with the required positioning accuracy,to all fuel channels, the new and irradiated fueltransfer ports, the rehearsal facility and theancillary port;

iii) drive mechanisms compatible with computercontrol in the fully automatic, semi automatic andmanual modes;

Design requirements related to safety andlicensing were:

iv) meeting the requirements of the CAN/CSAN285 series of standards for Class 1C pressureboundary component supports, with references tothe ASME Boiler and Pressure Vessel Code,Section III, Subsection NF;

v) application of quality assurance programsduring design and construction, based on the CSACAN3-N286 and Z299 standards;

vi) carriage design licensable in Canada;

111


vii) seismic qualification to ensure structuralintegrity during a 0.2 g Design Base Earthquake(DBE);

viii) providing safe and reliable transport of thefuelling machine head, thus ensuring the integrityof the contained new and irradiated fuel and of theinterfacing components;

ix) meeting the environmental qualificationrequirements for normal operating conditions andpostulated accident conditions;

x) satisfying the current human factorsrequirements with respect to the system layout,equipment maintenance and overall minimizationof dose expenditures;

Other requirements:

xi) minimizing construction time.

3. CARRIAGE SYSTEM DESIGN

The design of the CANDU 9 480NU carriage isfunctionally based upon the Douglas Point NGScarriage. There are two dedicated fuelling machineand carriage assemblies per reactor that operate inconcert to fuel a single channel. Correspondingly,there are two independent fuel transfer systems tohandle new and irradiated fuel.

Each CANDU 9 480NU carriage (Figure 1)operates between the fuel transfer ports inside themaintenance lock and the fuelling machine vault.Once in the vault, the carriage turntable rotates 90°so that the fuelling machine faces the reactor. Thecarriage positions the fuelling machine head at theselected lattice location in the X (lateral) and Y(vertical) directions, engages the seismic lockingmechanisms, and then advances the fuellingmachine head in the Z direction. Once the headarrives at the pre-stop position, X and Y correctionmeasurements are established. If necessary, thehead is retracted clear of the end fitting and X andY homing corrections are carried out by the fine Xand carriage Y drives, respectively. Therepositioned fuelling machine is then advanced tocontact the end fitting with sufficient force to stallthe Z drive. At this point, a standard CANDUfuelling sequence commences.

The carriage is designed to the requirements of theCAN/CSA N285.0, General Requirements ForPressure Retaining Systems and Components inCANDU Nuclear Power Plants and CAN/CSAN285.2, Requirements for Class 1C. 2C. and 3CPressure Retaining Components and Supports inCANDU Nuclear Power Plants. These standardsrecognize the unique design of CANDU fuel

handling systems and allow pressure vesselsupports such as the carriage to be designed inaccordance with the ASME Boiler and PressureVessel Code. Section III. Subsection NF. Throughconformance to these design standards andcodes, the carriage system becomes eligible forlicensing in Canada.

The carriage assembly structure consists of sixmain sub-assemblies: the base, turntable,columns, outer elevator, inner elevator, and guideplate (Figure 1). The movements of the carriagesub-assemblies relative to one another and withrespect to fixed supports facilitate the fuellingmachine positioning at the reactor. The fixedsupports consist of a floor rail and an upper guidebeam. The carriage rail runs parallel to the reactorface and into the fuelling machine maintenancelock. The upper guide beam is centered over therail, spanning its full length. The guide beamsupports the carriage structure in the Z direction bytrapping the guide plate within a slot in the bottomof the guide beam.

The base is a structure on two wheels that providesthe X motion and half of the rotation freedom of thecarriage. It is a welded assembly consisting of twobeams, internal bracing and a circular skirt. Thebeams run parallel to the rail with cross membersspanning the rail. Additional bracing, perpendicularto the main members, supports the skirt. Onewheel of the base is driven by an electric motorthrough a reduction gear train, while the secondwheel is an idler that, when required, can be drivenby the backup X drive. The base is guided alongthe rail by guide rollers that contact the outsideedges of the rail. Hold down hooks, that run in slotsin the sides of the rail and are bolted to the mainbeams, prevent lifting of the carriage during anearthquake. The skirt supports a mounting ring atthe top and four outriggers at the bottom. Theoutriggers are provided for construction andmaintenance purposes and are capable ofsupporting the carriage assembly independentlyof the wheels. Finally, the inner race of a slewingring is bolted to the mounting ring at the top of theskirt.

On top of the base, the outer race of the slewingring is bolted to the underside of the turntable. Theouter race is manufactured with an external gearcut in its outside perimeter. For carriage rotation,this gear meshes with a pinion fixed to the basethrough another reducer and an electric motor.The turntable structure is of a welded, sandwichstyle construction. It has internal cross-bracingthat runs diagonally between the main membersand it is enclosed on top and bottom with steel

112


plates. Access openings are provided for theconstruction of the turntable, and for inspectionand replacement of the bolts between the innerslewing ring cage and the base.

The column sub-assembly is bolted to the top ofthe turntable and it consists of two vertical columnsand two horizontal crossbeams (Figure 2). Thevertical columns are internally braced and have Yguide tracks along their front and back sides. Theinternal bracing at the bottom of the columns is inan X pattern facing upwards, providing hightorsional rigidity. The bracing in the active portionof the columns is arranged in a zig-zag pattern withthe edges of each brace corresponding to areactor lattice position. These angled bracessupport loads imparted by the outer elevator Yguide bearings to the columns through the Y guidetracks. The top outside stub of each columnhouses a radial thrust bearing which holds theballscrew. The lower stub on each columnsupports the ballscrew tensioner, while the Y driveseismic brake and gear reducer are bolted to theoutsides of the columns below the stubs. The Ydrive electric motors and drive trains are mountedon the turntable for improved access. The lowerhorizontal crossbeam between the two columnsserves as a mechanical lower stop position for theouter elevator. Permanent shielding is bolted to thelower crossbeam that will shield the underside ofthe fuelling machine head magazine duringmaintenance. The upper crossbeam houses thebearings which in turn hold the shaft of the upperguide plate. These bearings provide a pivot pointto allow rotation of the turntable and the columnsrelative to the base and the upper guide beam. Themating surfaces between the upper crossbeamand the columns are provided with keys that limitseismic shear loads through the bolts.

The outer elevator is positioned between thecolumns and allows the vertical movement of thefuelling machine. The outer elevator consists of aframe, four pedestals, outer crossbeams, a Z drivemounting beam and Z drive tracks (Figure 3). Theframe fits between the columns with extensions onthe sides of the frame protruding ahead of andbehind each column. The pedestals are bolted tothese extensions and they house the Y guidebearings. The Y guide bearings provide support tothe outer elevator in both the X and Z directions byriding along both the face and sides of the Y guidetracks mounted on the columns. An outercrossbeam spans between the front and backpedestals along the outside of each column,thereby surrounding each column. This outercrossbeam is the mounting point for the Y drive

ballnut. The Z drive mounting beam bolts acrossthe back two pedestals and serves to transfer Zdrive reaction and seismic forces from the fuellingmachine to the columns through the pedestals andY guide bearings. The Z drive tracks are mountedon the bed of the outer elevator and provide a trackfor the inner elevator to ride along. The Z drive is anAcme screw which is turned by an electric motorthrough a reducer. The screw, while rotating, drivesan Acme nut along its length. The nut itself is fixedto one end of a seismic spring pack which in turnis fixed to the inner elevator. The spring pack iscomprised of a tuned stack of hard Belleville springwashers. Its purpose is to allow relative movementof the fuelling machine with respect to the carriage,thereby limiting the carriage seismic inertial forcesimparted to the end fitting should the fuellingmachine be clamped on to an end fitting during anearthquake.

The inner elevator moves relative to the outerelevator to provide the Z motion of the carriage.The inner elevator consists of a base, a yawturntable, and a cradle support. The base haslinear guide bearings along the bottom that run onthe top and sides of the Z tracks. The yaw turntableis a smaller slewing ring that has its inner racebolted to the top of the inner elevator base, and itsouter race bolted into the bottom plate adjoiningthe cradle support pillars. The support pillarshouse the pitch pivot trunnions of the fuellingmachine cradle and the cradle moves laterallywithin the support pillars to provide the fine Xmotion. Spring centering mechanisms betweenthe fuelling machine cradle and the cradle supportpillars, and between the cradle support and theinner elevator base provide the pitch and yawfreedom necessary to allow a fuelling machine tocenter over an end fitting.

4. SEISMIC DESIGN FEATURES

Undesirable motion of the CANDU 9 480NUcarriage under seismic conditions is preventedthrough several means. The first restraintmechanism is an arrangement of wedge and pinstyle seismic locks (Figure 4). When activated,these mechanical devices prevent rotation andtranslation by locking the carriage sub-assembliesto one another or to fixed supports. The seismiclocks are engaged during every channel refuellingsequence, prior to the fuelling machine beingadvanced over the end fitting. Both types of locksemploy electric motors and Acme screw drives,and readily accessible manual drives for recoveryoperations are available.

113


The wedge style seismic locks accommodatethermal expansion between interfacingsub-assemblies and/or fixed supports. The locksemploy force control feedback through the Acmescrew drives to limit the depth of lock engagement.The force control establishes a positive contactbetween the locking wedge face and the lockingstriker plate but it limits the force in order to preventany tendency of the lock system to move thecarriage. The lower arrangement consists of twolocks bolted back to back and mounted at thereactor side of the carriage base. These two locksthen engage locking holes in a floor embedment.At the top of the carriage, one lock is bolted to theback of each column and it engages acorresponding hole in the upper guide beam. Thelocking holes in the upper guide beam and floorplate are spaced one lattice pitch apart. Due to thevarious angles at which the carriage must operatein the fuelling machine maintenance lock, a radialpattern of locking holes exists in that section of theupper guide beam. Because the base does notrotate, only a single embedment with two lockingholes is sufficient to lock the base to the floor whenthe carriage is in the maintenance lock.

The locking of the turntable to the carriage base isaccomplished with a straight pin style seismic lock.The straight pin style is used only in this locationbecause the two sub-assemblies, given theirphysical proximity, will be subject to the samethermal transients, thus leading to minimalmisalignment. One lock is located at the base ofeach column. The pin extends through the bottomof the turntable and engages a mating hole in alocking ring welded to the base structure within theslewing ring.

In addition to the wedge and pin type seismiclocks, all electric drive motors have integral brakeswhich are automatically applied when the drive isde-energized, making them fail-safe. For furthersafety, the ballscrews are equipped withseismically qualified fail-safe brakes and arepre-tensioned with spring packs that preventcompression of the ballscrew under seismicallyinduced lifting loads.

In order to protect the fuelling machine carriageassembly from an earthquake occurring while thecarriage is in transit, a seismically activated andqualified switch will cut the power to all drives,thereby activating their brakes. As a backup to themotor brakes, seismically qualified snubbers andmechanical stops prevent overtravel and providefor controlled deceleration of any componentshould the motor brakes slip.

5. DESIGN BENEFITS

The benefits of the carriage design range fromsimplified construction to improved reliability andreduced time and radiation dose expendituresduring operation and maintenance activities. Thenew design incorporates feedback from operatorsand suppliers. The CANDU 9 480NU design is alsoexpected to be adaptable for CANDU sites withseismic levels higher than the 0.2 g DBE currentlybeing considered.

Manufacturing tolerances are eased by the use ofshims between sub-assemblies becausecustomized shims bring critical sub-assembliesinto alignment. Along with these shims, simplebolted connections between sub-assemblies allowfor a shorter carriage set up time at site.Procurement, manufacturing and construction ofthe carriage are all simplified by maximizing theuse of standard commercial parts. Additionally, theselection of proven electric motors as the primemovers has led to a significant reduction in thenumber of mechanical parts required for theCANDU 9 carriage in comparison to the CANDU 6carriage, largely due to the elimination of thecarriage oil hydraulic system.

The reliability of the CANDU 9 carriage is expectedto be improved over that of the CANDU 6 carriage.The structure is inherently robust, with all carriagecomponents that support the fuelling machinebeing designed to accommodate loads resultingfrom a DBE. Given that the normal operating loadsare, typically, an order of magnitude lower thanseismic loads, the mechanical equipment of thecarriage is expected to have considerableoperating lifespan. In order to further improve thesystem reliability, a backup X drive is provided anda single Y drive unit can safely lower the elevatorand fuelling machine head assembly. As well, oneseismic brake and one ballscrew can support theentire suspended mass. Other features thatenhance safety and reliability include an increasedZ drive range for recovery purposes; if required,the full Z drive range will allow the recovery of afuelling machine with a guide sleeve stuck in thesnout, even with a crept pressure tube.

Maintenance activities, and the anticipatedradiation dose expenditures attributable to theseactivities, were a significant consideration in theselection of a carriage system. The CANDU 9480NU mobile carriage system allows for carriagemaintenance to be carried out inside the shieldedand environmentally isolated maintenance locks,rather than in more active locations such as thefuelling machine vaults. On-power maintenance

114


of the carriage is possible, a feature that positivelycontributes to the station capacity factor ascarriage maintenance does not require a reactorshutdown. Other improvements are obtainedthrough ready access to carriage components andthe use of bolted component connections. Boltedconnections facilitate the removal ofsub-assemblies that can then be taken to lowactivity locations for maintenance. For field work,strategically placed shielding for the fuellingmachine head will be incorporated into thecarriage structure so that fuelling machine headmaintenance can be carried out with less dosecommitment than in the previous CANDU designs.Finally, through the use of commercially availableparts, carriage maintenance can be performed byless specialized work groups, thus enabling doseequalization programs to succeed.

6. SUMMARYThe CANDU 9 480NU fuelling machine carriagedesign represents a departure from the bridgesystem utilized in the successful CANDU 6 reactordesign. The carriage concept is not new, however,as its origins can be traced to the earlier CANDUreactor designs such as that at Douglas PointNuclear Generating Station.

The two wheel fuelling machine carriage design,described in the foregoing sections, meets thefunctional, performance, safety and constructabilityrequirements of the CANDU 9 480NU project. Thecombination of a robust CANDU 9 carriage, fullyintegrated with the proven CANDU 6 fuellingmachine head, has the capacity to support thestation lifetime target capacity factor of 90%.

The carriage design has the potential to satisfy therequirements of future CANDU sites with higherthan 0.2 g DBE levels. Other significant benefitsinclude on-power maintenance access to boththe carriage and the fuelling machine head, andthe use of commercially available, standardizedcomponents. These features are intended tominimize dose expenditures and the costsassociated with fuel handling systemmaintenance. In addition, a high degree ofpre-assembly is planned in order to shortenin-situ construction and the commissioning timerequirements.

The carriage system, as a pressure boundarycomponent support, meets the requirements ofthe applicable CSA standards and ASME Coderequirements. It is designed to be licensable inCanada, as well as other comparable internationaljurisdictions.

ACRONYMS

CANDU - Canada Deuterium Ujanium

ASME - the American Society of MechanicalEngineers

CAN/CSA - Canada / Canadian StandardsAssociation

DBE - Design Basis Earthquake

115


GUOEBEAM

M E PLATE

COLUMN

FUELLNG MACHNECRADLE

OUTERELEVATOR

NNERELEVATOR

TURNTABLE

BASE

FIGURE I CARRIAGE GENERAL ARRANGEMENT

116


UPPER SOSMCLOCKS (2)

Y - G U O ETRACKS W

LOWERCROSSBEAM

TURNTABLESBSME LOOS (2)

UPPEREROSSBEAM

LOWER SEISMCLOCKS Gl

FIGURE 2 CARRIAGE COMPONENTS

117


NNERCROSSBEAMS C l

Z - QRIVEMOUNTING BEAM

OUTERCROSSBEAMS 12)

CRADLEPILLARS

YAWTURNTABLE

BALLNUTS (2)

Y - GUDE BEARINGHOLDERS

PEDESTALS IM TRACKS 0

FIGURE 3 ELEVATOR COMPONENTS

118

VO

WEDGE - STYLESEISMIC LOCK

PIN - STYLESEISMIC LOCK

FIGURE 4 SEISMIC LOCKING MECHANISMS

oozm3D

mOmOz

oc•n

m

>

oZ

a

m

w

I0)


THE CANDU 9 FUEL TRANSFER SYSTEM

Z.G. KeszthelyiGE Canada, Nuclear Products

107 Park Street NorthPeterborough, Ontario, K9J 7B5

D.T. MorikawaAECL

2251 Speakman DriveMississauga, Ontario, L5K 1B2

CA9700731

ABSTRACT

The CANDU 9 fuel transfer system isbased on the CANDU 6 and the OntarioHydro Darlington NGD designs, modifiedto suit the CANDU 9 requirements. TheCANDU 9 new fuel transfer system isvery similar to the CANDU 6, withmodifications to allow new fuel loadingfrom outside containment, similar toDarlington. The CANDU 9 irradiated fueltransfer system is based on the Darlingtonirradiated fuel transfer system, withmodifications to meet the more stringentcontainment requirements, improveperformance, and match station layout.

1. INTRODUCTION

The CANDU system is based on on-powerfuelling. An important aspect of thesystem is 'fuel transfer' which includes ailthe equipment and operations required toload new fuel into a fuelling machine fordelivery to the reactor and later totransfer the irradiated fuel between afuelling machine and the storage bay.

This paper describes an advanced fueltransfer system being designed for theCANDU 9.

CANDU 9 is the latest version of thePressurized Heavy Water Reactor (PHWR)system developed in Canada by AECL. Ithas a net electrical output in the range of870 MW.

The seemingly simple fuel transfer systemmust satisfy many stringent requirements.The basic ground rules for this equipment

are safety and reliability. The equipmentmust be designed and built, for instance,in such a way that at no time will theoperator be confronted with a situationwhich leaves him without any means forsolving a break-down problem involving astuck, uncooled fuel bundle in theirradiated fuel port.

The paper provides a brief description andillustration of the new and irradiated fueltransfer systems at both the CANDU 6and Darlington stations. The papercontinues with a more detailed descriptionof the proposed CANDU 9 fuel transfersystem which is being designed by a jointteam of AECL and GE Canada designers.

2. DESCRIPTION OF NEW FUELTRANSFER SYSTEMS

2.1 CANDU 6

In the CANDU 6 stations, new fuel isstored outside the reactor building anddelivered into the reactor building whenneeded. Once in the reactor building,new fuel is loaded manually into the newfuel transfer mechanism. It is thentransferred into the fuelling machine headunder remote control from the maincontrol room. (See Figure 1).

Two new fuel transfer mechanisms areprovided; one to supply each fuellingmachine head.

The new fuel transfer mechanisms arelocated in the reactor building, withincontainment, between the two fuellingmachine locks.

121


The new fuel transfer mechanism consistsof a loading trough and ram, a magazineand indexing drive, a single-stage transferram, and a port/magazine transition piece.

The magazine assembly consists of ahousing and rotor, with capacity for 12fuel bundles and the new fuel port shieldplug. The magazine is indexed by anelectric motor and a mechanical indexingunit.

An airlock gate valve is installed betweenthe new fuel magazine and the loadingtrough to seal off the magazine wheneverfuel is not being loaded into themagazine. This valve reduces the spreadof contamination from the fuellingmachine maintenance lock into the newfuel room. Contamination is also reducedthrough a duct connected to the vapourrecovery system which maintains a slightnegative pressure in the new fuelmagazine.

The loading ram is an oil/air driven ram.The transfer ram is driven by an electricmotor and chain drive.

The port/magazine adapter connects themagazine and the new fuel port. Thisassembly locks the new fuel port shieldplug in place and releases the shield plugfrom the transfer ram when needed.

The new fuel ports are mounted inembedments in the maintenance lockwalls. A shield plug is normally latchedin the new fuel port to provide shieldingfor the new fuel room.

2.2 DARLINGTON

There are four new fuel transfermechanisms in each of the two loadingareas. These mechanisms are locateddirectly above the irradiated fuel receptionbay in the fuelling facilities auxiliary areas.(See Figure 2).

The new fuel port penetrates thecontainment wall between the new fuelloading area and the fuelling machineroom. The containment integrity is

maintained by two gate valves betweenthe new fuel port and the magazineassembly. A third gate valve is locatedbetween the loading trough and themagazine and is interlocked with theother gate valves to prevent simultaneousopening. The magazine is maintained at aslightly negative pressure through anactive vent to prevent any flow of tritiuminto the new fuel loading area.

The fuel bundles are transferred from theloading trough into the magazine by aram, the force of which is restricted tolimit the force applied to the bundles.

The fuel bundles are transferred from themagazine into a transporter located in thenew fuel port and then into the fuellingmachine head by a two-stage ram. Thetransporter is moved from the new fuelport into the head by one ram while thesecond ram transfers the bundles fromthe transporter into the fuel carriers in thefuelling machine head magazine.

2.3 CANDU 9

The CANDU 9 new fuel transfermechanism is based on both the CANDU6 and Darlington designs. The designpermits new fuel to be loaded into thefuelling machine head from outsidecontainment.

The CANDU 9 design uses the standard37-element fuel bundle, but will also beable to accommodate the 43-elementCANFLEX fuel bundles in the future. Thefuel handling sequence for CANDU 9 isillustrated in Figure 4.

2.3.1 Description

Two new fuel transfer mechanisms areprovided; one to supply each fuellingmachine head.

The mechanisms, shown in Figure 5, arelocated outside containment in the reactorauxiliary building on either side of theairlock. New fuel is therefore loaded intothe new fuel transfer mechanisms fromoutside the reactor building.

122


The new fuel transfer mechanism consistsof a loading trough and ram, a magazineand indexing drive, a single-stage transferram, a port and two port containmentvalves.

The new fuel magazine, similar to theCANDU 6 magazine, consists of acompletely enclosed steel housing with arotor. As in CANDU 6, there are six fuelstations and a station for the port shieldplug. The magazine rotor is driven from anelectric motor through a pinion and bevelgear. Indexing is controlled by amechanical cam and roller drive indexingunit.

The fuel loading ram consists of astainless steel rack with a ram headassembly. The ram is driven by a pinionand brushless servo motor.

An airlock valve, installed between thenew fuel magazine and the loadingtrough, seals off the magazine wheneverfuel is not being loaded into themagazine. This valve reduces the spreadof tritium through the new fuel magazineinto the new fuel room and enables thetransfer mechanism and fuelling machinehead to be pressurized during fuel transferto the fuelling machine head.

The transfer ram is similar to the new fuelloading ram and consists of a ram tubewith a latching ram head and a stainlesssteel rack. The entire ram is contained ina housing with a seal at the magazinehousing. The ram is driven by a pinionwith a sealed shaft.

The new fuel port is similar to the CANDU6 component. This assembly locks thenew fuel port shield plug in place andreleases the transfer ram from the shieldplug when needed.

The shield plug is normally latched in thenew fuel port and performs the samefunction as on CANDU 6.

The new fuel port is mounted in anembedment in the reactor building wall

between the new fuel transfer room andthe fuelling machine lock. Because thenew fuel transfer mechanism is fixed tothe reactor auxiliary building, and theports to the reactor building, the transfermechanism is designed to accommodatedifferential settlement and building shiftsdue to seismic events.

The port containment valves are locatedbetween the port and the magazinehousing. These valves are normallyclosed to maintain the containmentboundary.

2.3.2 Operation

New fuel pallets are delivered from thenew fuel storage room to the airlocklevel, and from there to the new fueltransfer room.

The new fuel bundles are inspectedbefore loading into the transfermechanism. The fuel bundles are liftedfrom the pallet using a fuel bundle liftingtool, suspended from an air balance hoistand a monorail. The bundles are placedon an inspection table and checked forinterlocked element spacers, using agauge, and vacuumed/air blown ifneeded.

Two bundles are loaded into the loadingtrough and the trough lid is closed. Theairlock valve is opened and the bundlesare pushed into the magazine by the newfuel loading ram. The ram is retractedand the magazine is indexed to the nextchannel position and two more bundlesare loaded. This is repeated until themagazine contains the required number ofbundles.

During new fuel loading, the magazine isopened to the ventilation - system tomaintain air flow from the new fuelloading room into the magazine.

To transfer fuel from the new fuel transfermechanism into the fuelling machine, thefuelling machine clamps onto the new fuelport in the fuelling machine lock, withincontainment in the reactor building. The

123


water level in the fuelling machine islowered and the fuelling machine preparesthe port for fuel transfer by installing theguide sleeve.

The new fuel transfer mechanism and thefuelling machine head are pressurized andthe two port containment valves areopened. During the remainder of thetransfer sequence, the pressure in thetransfer mechanism and port is controlledby the new fuel transfer mechanismauxiliaries.

Until the containment valves are opened,the port and containment valves are thecontainment boundary and rest of thenew fuel transfer mechanism is outsidecontainment. After the containmentvalves are opened, the fuelling machinehead becomes the containment boundary.The pressure in the transfer mechanismand head allow containment integrity tobe confirmed and monitored.

To meet the regulatory requirements, thenew fuel port and containment valves areClass 2, and the remainder of the transfermechanism is Class 6.

Fuel bundles are transferred from thetransfer mechanism into the fuellingmachine head by the transfer ram. Thetransfer ram first removes the port shieldplug and stores it in the new fuelmagazine. The ram then pushes the fuelbundles in pairs from the new fuelmagazine into the fuelling machine head.This operation is controlled remotely fromthe fuel handling console in the maincontrol room.

3. DESCRIPTION OF IRRADIATED FUELTRANSFER SYSTEMS

A paper presented at an earlierconference (1) defined the major designrequirements for irradiated fuel transfersystems.

3.1 CANDU 6

The CANDU 6 irradiated fuel handlingsystem consists of discharge and transferequipment in the reactor building, andirradiated fuel reception and storage baysand equipment in the service building.(See Figure 1).

Irradiated fuel is transferred from thefuelling machine head to the dischargebay in air. The fuelling machine clampsonto the irradiated fuel port, prepares theport for transfer, and lowers the waterlevel in the head. The two ball valves onthe irradiated fuel port are opened andbundle pairs are pushed by the fuellingmachine ram through the port and ontoan elevator.

The elevator lowers the bundles into thedischarge bay onto a conveyor cart, thenreturns to the port to receive the nextbundle pair. This is repeated until all fuelhas been discharged from the fuellingmachine. There is one discharge bayconveyor serving both fuelling machines.The discharge conveyor transfers the cartto the reception bay conveyor.

During irradiated fuel transfer from thefuelling machine head, the containmentboundary is at a containment gate at theend of the discharge bay canal. EarlierCANDU 6 units were not equipped with acontainment gate and rely on the fuellingmachine auxiliaries and the head of waterin the discharge canal for the containmentboundary. At all other times, the portvalves are closed.

After fuel reaches the reception bay, thefuel rack is removed from the conveyorcart and placed in the reception bay. Fuelbundles are transferred to the storage baytrays using manual tools. Once the traysare full (24 bundles/tray), they aretransferred to the storage bay and placedon storage tray supports.

The storage bay trays are stacked 19 highand sets of four stacks are secured with acovering frame. This assembly is sealedwith an IAEA device.

124


For interim dry storage, bundles aretransferred from the trays to cylindricalbaskets. The baskets are then placed indry storage concrete canisters.

3.2 DARLINGTON

The major differences between theCANDU 6 and Darlington systems are asfollows:

The CANDU 6 irradiated fuel transfersystem is 'above water' meaning that theirradiated fuel port is located above thewater level in the bay. Darlington uses an'underwater' concept wherein the port islocated below the bay water elevation.An air-filled transfer chamber ('airchamber') is used. The bundles aretransferred from D20 in the fuellingmachine head, to the air chamber, andfinally into H20 in the bay. In a CANDU 6storage bay, irradiated fuel is stored insingle-layer trays. In Darlington, it isstored in multi-layer modules. Thisintroduces an additional indexingoperation.

A description of the Darlington systemfollows. Figure 3 illustrates the systemfeatures.

The irradiated fuel port penetrates thecontainment wall 4.4 m below the surfaceof the bay water. An underwater airchamber is attached to the port from thebay side. Flooding the air chamber withbay water provides emergency cooling incase a bundle becomes stuck in the port.Defected fuel in the underwater chambercan be identified by drawing air from thechamber through a detector ('drysniffing').Irradiated fuel is discharged from thefuelling machine onto a shuttle in theirradiated fuel port. The shuttle isretracted through the port into the airchamber where it is supported on a ladle.The ladle and shuttle are lowered totransfer the shuttle onto a verticallytravelling elevator. The elevator islowered to align the shuttle with a chosentube of the module, and the fuel ispushed out of the shuttle into the module.

Horizontal indexing of the modulecombined with vertical indexing of theelevator make it possible to fill all tubes inthe module.

Accurate indexing of the module conveyoris obtained by a Geneva mechanism. Atelescopic water hydraulic cylinderprovides smooth motion of the shuttlethrough the port. Shuttle positiondetection is by ferromagnetic sensors.Underwater equipment is driven, as far aspossible, from above water by motorizeddrives. All motorized equipment has thecapability of being driven manually.

The initial concept for handling defectedfuel at Darlington was to separatedefected fuel from sound fuel and, afterinspection in the reception bay, place it incans. Later experience indicated that theisolation of defected fuel in cans andcontainers was not necessary. A decisionwas made to place defected fuel directlyinto modules, with provisions for laterremoval of defected fuel from themodules for inspection.

3.3 CANDU 9

The fuel handling sequence for CANDU 9is illustrated in Figure 4. The systemminimizes the fuelling machine refuellingcycle time. This is achieved bytemporarily storing the irradiated fuel inthe port. Loading of the fuel into thestorage bay takes place later, after thefuelling machine departs from theirradiated fuel port.

3.3.1 Description

The irradiated fuel transfer system, shownin Figure 6, is located primarily in thereactor auxiliary building. It consists of:

a) An irradiated fuel port mounted in thereactor building containment wall.

b) Two port containment valves.c) A flexible port/magazine transition

piece.d) A totally enclosed irradiated fuel

magazine and its drive assemblywhich is located in the storage bay

125


wall within the reactor auxiliarybuilding.

e) An irradiated fuel transfer ram anddrive.

f) A bay valve.g) Associated air and water auxiliary

systems.h) Storage bay equipment which

includes the indexing tray, storageram, bundle tilter, 'basket modules',stacking frames and handling tools.

A paper published earlier (2) describes the'basket module' and future considerationsfor the dry storage of CANDU fuel.

The irradiated fuel port assemblyconforms to Class 2, Class 4 and Class 6as appropriate. It has the flexibility andstrength requirements to permit thehoming and locking of the fuellingmachine and to withstand high seismicloads. The port is sloped down towardsthe port magazine. D2O carried from thehead on the fuel bundles is collected toprevent its loss to the bay.

The irradiated fuel transfer mechanismspans two separate buildings; the reactorbuilding and the reactor auxiliary building.These two buildings can moveindependently of one another undersettlement and seismic actions.

The fuelling machine unloads irradiatedfuel through its associated port whichleads to one of the irradiated fuel storagebays located outside the reactor buildingcontainment wall. Each face of thereactor has one storage bay dedicated toit. The two storage bays are notinterconnected.

The port/magazine transition pieceprovides the required flexibility betweenthe buildings to allow these movementsto be accommodated without loss ofcontainment. This is achieved using twospecially designed ball joints toaccommodate the vertical, horizontal andaxial displacements. The transition piecehas a D2O recovery circuit to recover D2Ovapour that evaporates or boils off thefuel bundle surface during fuel transfer in

air. This D2O recovery loop also containsprovisions for 'dry sniffing' to detectdefected fuel as it passes through thetransition piece.

The irradiated fuel magazine (or portmagazine) is located in a sleeve in thestorage bay wall within the reactorauxiliary building. The end plate of themagazine housing is removable to enablemaintenance to be carried out on therotor assembly. Mounted on the endplate are the transition piece, themagazine drive assembly, the irradiatedfuel transfer ram, and cooling waterconnections. The heat generated by theirradiated fuel bundles within themagazine is removed through a heatexchanger fitted internally in the magazinehousing.

The magazine rotor consists of astationary shaft, two end plates and sixmagazine channels. There are sevenchannel positions with one position usedduring maintenance only. This position isfitted with a hinged blanking plate toallow space for special tools (e.g.cleaning tools). The capacity of themagazine rotor is therefore six pairs offuel bundles.

The CANDU 9 irradiated fuel transfersystem is being designed to accommodateemergency recovery for all foreseeablefailure scenarios in a manner thatminimizes radiation and conventional riskto station staff, the public and theenvironment.

3.3.2 Operation

To prepare for fuel transfer, a pump isstarted to establish water circulation flowthrough the port magazine. The waterflows over a weir for level control in themagazine. The irradiated fuel magazineand transition piece are pressurized withair. The fuelling machine homes andlocks onto the irradiated fuel port and ispressurized. The two port containmentvalves are then opened and the port airpressure system regulates the pressure inboth the port and the head.

126


The fuelling machine magazine channel,containing the fuel bundles to betransferred, is aligned with the port. Therams of the fuelling machine are used topush the irradiated fuel from the fuellingmachine magazine into the port magazine,two bundles at a time. The rams are thenretracted into the fuelling machine andboth magazines index to the nextposition. This procedure is repeated untilall the fuel has been transferred.

The irradiated fuel transfer takes place inair from the time that the fuelling machinemagazine is indexed to align the fuel fortransfer until the port magazine is indexedfrom the load position and the fuel issubmerged below the water level in theport magazine. Fuel is then cooled in thecirculating water in the magazine.

The control systems associated with thedesign ensure that the proper sequencingof the operation is achieved. Once thefuel transfer is complete, the two portcontainment valves are closed. The portis then flooded. In an emergency, theport and magazine can be flooded from adump tank by gravity.

Irradiated fuel is now transferred from theport magazine to the storage bay byopening the bay valve. This operationequalizes the pressure in the port with thebay pressure. The bundles are pushed bythe transfer ram, two at a time, from themagazine onto an indexing tray as it ismoved automatically across the face ofthe irradiated fuel port. The tray has sixpositions for fuel bundles. Uponcompletion of transfer, the bay valve isclosed.

A storage ram is then used to push thebundles, one at a time, into a bundle tilterdesigned to rotate the bundle into thevertical position. The tilter restrains thebundle during this operation and is fittedwith a shock absorber to reduce impacton the bundle during tilting. A toolsupported from the manbridge is used totransfer the irradiated fuel bundles fromthe tilter to the 'basket module1.

4. CONCLUSION

Major features of the CANDU 9 irradiatedfuel transfer system design includecommonality of equipment between thenew fuel and irradiated fuel mechanismsand reuse or adaptation of technologyused and proven at other CANDU plants.The transfer system design providesadaptability to a wide range of siteconditions, low capital cost, shortconstruction schedule, minimum systemcomplexity, high system availability, lowoperating and maintenance cost, and ageneric approach where many of thefeatures can be applied to future CANDUdesigns.

ACKNOWLEDGMENTS

The authors would like to thank OntarioHydro, AECL and GE Canada for the useof their data.

REFERENCES

(1) KES2THELYI Z.G., DIDOMIZIO F.,GAERTNER B.R., "The Evolution ofCANDU Irradiated Fuel TransferSystems", A Paper Presented at theThird International Conference onCANDU Fuel, Chalk River, Ontario,Canada, 1992 October 4-8.

(2) NAKAGAWA R.K., "StandardizedApproach to the Design of CANDUUsed Fuel Receptacles", A PaperPresented at the Third InternationalConference on CANDU Fuel, ChalkRiver, Ontario, Canada, 1992October 4-8.

127

4ACCEPT

MACHINE

IRRADIATED FUEL1

STORAGE BAY

CANOU 6FUEL HANDLING SEQUENCE

NEW FUELSTORAGE ROOM

REACTOR SUILO3NG

IRRADIATED FUELDISCHARGE ROOM

. DEFECTED" FUEL BAY

CANNEDDEFECTED FUEL

STORAGETRAYS

CQ FIGURE 1

LOADINGTRANSFER T R ° V G H FUEL AIRLOCK

RAM \ BUNDLES VALVE MAGAZINE

FUELTRANSPORTER

TRANSPORTERRAM

LOADINGRAM

FUELLINGMACHINE

ROOM

FUELLINGMACHINE

PORTVALVES

NEW FUELPORT

DARLINGTONNEW FUEL TRANSFER SYSTEM

CQ FIGURE 2

WATER LEVEL

rARECEPTION

BAYELEVATOR"

FUELLINGMACHINE

ROOM

CARRIER

• ; : -•*• ;v ' * = • • > • - L

LADLEDRIVE TUBE-

MAGAZINE

ELEVATOR COLUMNAND TRACK

>i nit

ELEVATOR

DARLINGTONIRRADIATED FUEL TRANSFER SYSTEM

LADLE'

. AIR/ XHAMBER

-SHUTTLE

LADLEARM

CQ FIGURE 3 SECTION A-A

NEW FUELLOADING AREA

REACTOR BUILDINGCONTAINMENT WALL

REACTOR

NEW FUEL INEW FUEL B " N D L E

PALLET VL

A

NEW FUELTRANSFER

MECHANISM\

IE

INTERIM DRYSTORAGECANISTER

IRRADIATED FUEL ^STORAGE SAY t

BASKET MODULESTACKING FRAME

BASKETMODULE

V

\

• •ID

BASKETMODULE

IN BUFFERZONE

»'••. . •

? « • • '

CANDU 9FUEL HANDLING SEQUENCE

CG

FUELLINGMACHINE AND

" CARRIAGE

EL. 100.0T-.••"•'>;V.: '*.

IRRADIATEDFUEL BUNDLES

IRRADIATEDFUEL FROMREACTOR

EL. 100.0

FUELTRANSFERSYSTEM 'A'

B EL REACTORBUILDING

FIGURE 4

FUELTRANSFERSYSTEM'C

TWO BAY LAYOUT

REACTORAUXBLtARY BUILDING

NEW FUELPALLET

BUNDLEHOIST LOADING

TROUGHACTIVE VENT

VALVE

LOADINGRAM

NEW FUELBUNDLE

EL. 104.0

AIRLOCKVALVE

TRANSFERRAM

NEW FUELTRANSFER

MECHANISMCONTAINMENT

VALVES


NEW FUELPORT

CAMDU9NEW FUEL TRANSFER SYSTEM

CG FIGURE 5

FUELLING MACHINE, AND CARRIAGE

-OTO REACTOR

EL. 100 0


W. L. EL. 109.0

IRRADIATED FUELSTORAGE BAY

EL. 104.014

PORT/MAGAZINETRANSITION PIECE

CANOU9IRRADIATED FUEL TRANSFER SYSTEM

CGFIGURE 6

CA9700732


OVERVIEW OF THE CANDU FUEL HANDLING SYSTEMFOR ADVANCED FUEL CYCLES

by

D.J. Koivisto, D.R. Brown, and V.K. Bajaj

Atomic Energy of Canada LimitedSheridan Park, Ontario, Canada

L5K 1B2

ABSTRACT

There has been much interest around theworld in the development of advanced fuelcycles for the purposes of resourceutilisation and the disposition of reactorwastes or weapons material. The CANDU1

system is an ideal system for implementingadvanced fuel cycles because it can beadapted to burn these alternative fuelswithout major changes to the reactor.

The fuel handling system is adaptable toimplement advanced fuel cycles with someminor changes. Each individual advancedfuel cycle imposes some new set of specialrequirements on the fuel handling systemthat is different from the requirementsusually encountered in handling thetraditional natural uranium fuel. Thesechanges are minor from an overall plantpoint of view but will require some interestingdesign and operating changes to the fuelhandling system.

Preliminary conceptual design work hasbeen done on the fuel handling system insupport of some of these fuel cycles. Somefuel handling details were studied in depthfor some of the advanced fuel cycles. Thispaper provides an overview of the conceptsand design challenges.

1 CANDU9 - Canada Deuterium Uranium Registeredtrademark of AECL HBXT

lef I BLASliK

135


DEMONSTRATING THE COMPATIBILITY OF CANFLEX FUEL BUNDLESWITH A CANDU 6 FUELLING MACHINE

by

P. Alavi and LE. OldakerFuel Design Branch

Atomic Energy of Canada Limited2251 Speakman DriveMississauga, Ontario

L5K1B2Canada

CA9700733

and

H.CSuk and C.B. ChoiPHWR Fuel Development Technology Team

Korea Atomic Energy Research Institute150 Dukjin-dong, Yusong-ku

Taejon, 305-600,Korea

ABSTRACT

CANFLEX1 is a new 43-element fuel bundle,designed for high operating margins. It has manysmall-diameter elements in its two outer rings, andlarge-diameter elements in its centre rings. By thismeans, the linear heat ratings are lower than those ofstandard 37-element bundles for similar poweroutputs. A necessary part of the out-reactorqualification program for the CANFLEX fuel bundledesign, is a demonstration of the bundle'scompatibility with the mechanical components in aCANDU2 6 Fuelling Machine (FM) under typicalconditions of pressure, flow and temperature.

The diameter of the CANFLEX bundle is the sameas that of a 37-element bundle, but the smaller-diameter elements in the outer ring result in aslightly larger end-plate diameter. Therefore, tominimize any risk of unanticipated damage to theCANDU 6 FM sidestops, a series of measurements

1 CANFLEX - CANDU FLEXible fuelling

2 CANDU - CANada Deuterium Uranium is aregistered trademark of AECL

and static laboratory tests were undertaken prior tothe fuelling machine tests.

The tests and measurements showed that; a) theCANFLEX bundle end plate is compatible with theFM sidestops, b) all the dimensions of theCANFLEX fuel bundle are within the specifiedlimits.

INTRODUCTION

A demonstration of the CANFLEX fuel bundle in anactual CANDU 6 FM is necessary, as a final step, toqualify the bundle for use in CANDU 6 reactors. Tominimize the risk of unexpected damage to the FM,the dimensional compatibility of CANFLEX fuelbundles with the components of the CANDU 6 FMhead was studied prior to the out-reactor testing. Thestudy included a series of bench-type tests andmeasurements to check if the bundles met theCANFLEX bundle design dimensional requirements.

First, the CANFLEX fuel bundles were passedthrough a "bent tube gauge" to check the bundles'ability to pass through the rolled joint sections of thefuel channels.

137


The maximum diameters of the CANFLEX bundleswere measured. These measurements were comparedwith the maximum diameter specified for theCANFLEX bundle. The diametral measurementswere repeated while the bundles were squeezedinwards at the mid-plane.

Other dimensions of the CANFLEX bundles weremeasured using a co-ordinate measurement machine(CMM) to ensure that all the bundles' dimensionswere within the specified limits.

As part of this investigation, measurements weretaken in a special CANDU 6 "sidestep fixture" todetermine the clearances between the end plate ofthe CANFLEX bundle and the sidesteps. Thesidestop fixture is a simple test fixture designed andfabricated with two sidestop pieces that can bepositioned so as to simulate different 'penetrations1

of the CANDU 6 FM sidesteps.

The sidestops are part of the separators assemblies toprevent the column of fuel from moving into thedownstream FM head under the influence of thehydraulic drag force created by the coolant flow.

The CANFLEX bundle end elevation measurementswere also conducted in the CANDU 6sidestop fixture.

Finally, in this study, grappling tests of theCANFLEX fuel bundles were carried out manuallyon a bench to ensure that in an emergency situation,the new design of fuel bundle can be withdrawn froma fuel channel using a standard CANDU 6grappling tool.

Good quality assurance practice was followed for thetests and measurements presented in this paper, toensure that they met the quality assurance programfor design verification of nuclear componentsaccording to Canadian Standard N286.2 andcomplied with the practice adopted for the jointAECL/ KAERI3 CANFLEX program.

FM SEPARATOR ASSEMBLIES

There are two separator assemblies mountedsymmetrically on either side of the FM magazine

end cover. Both separator units are identical in allrespects, except that they are handed left and rightEach separator assembly consists of a casing thathouses hydraulic pistons actuated by heavy water.The pistons actuate the sensor, the pusher and thesidestop through connecting linkages (see Figure 1).

The sensor is lightly loaded in a downward directionto ride on the circumference of the fuel bundles. Thisdownward loading allows the sensor finger to sense agap between two fuel bundles. When the sensorfinger drops into the gap, it generates a signal toindicate that the gap is in the proper position toallow for the insertion of the sidestops. During theinsertion operation the sidestop is restrained to stayin alignment with the sensor finger.

For the tests described here, the sidestop fixture,shown in cross-section in Figure 2, was fitted withtwo "sidestops". A gap at the centre of each sidestoppiece represents the sensor finger.

BENT TUBE GAUGE MEASUREMENTS

The "bent tube gauge" is a simple tube with a veryprecise diameter and with an accurate bend angle,which is used by the fuel manufacturers to monitorthe maximum bundle diameter and its ability to passthrough the rolled joint sections of the fuel channels.

The bent tube gauge measurements were carried outto investigate if the CANFLEX bundles could passunder their own weight through the existing designof the bent tube gauge. Each bundle was placedinside the bent tube gauge horizontally. Then, oneend of the bundle/tube arrangement was lifted so thatthe bundle passed through the tube under itsown weight

A total of eight CANFLEX bundles were passedthrough the bent tube gauge, and all met thisacceptance criterion.

DIAMETER MEASUREMENTS

The diameters of the bundles were measured toensure that they met the requirements for themaximum diameter specified for the CANFLEXbundle. These measurements were repeated while thebundles were under compression at the mid-plane.

' KAERI - Korea Atomic Energy Research Institute

L38


Using a granite surface plate for a base, V-bk>ckswere placed for support at either end of the bundles.The circumferences of each bundle over the mid-plane and end-planes at the sheath and at the bearingpad locations were measured. Then, the bundle wascompressed at the bundle mid-plane, and similarmeasurements were taken.

BUNDLE DIMENSIONAL MEASUREMENTS

Dimensions of the CANFLEX fuel bundles were alsomeasured using a Co-ordinate MeasurementMachine (CMM).

The CMM is a programmable machine consisting ofa bridge and a carriage assembly that may be drivenin the x-y-z axes. The bridge and carriage assemblyare attached to a granite table. In this system, allthree axes are driven and controlled by dc frictiondrive motors and drive bars. The system is mountedon a series of air pads that serve as "anti vibration"stabilizers.

For the purpose of these measurements, the CMMwas programmed to measure the Pitch Circle Radius(PCR), end plate profile (waviness), bearing padheights, element bow, bundle radii, and bundledroop (i.e., end-cap-to-pressure tube clearance) ofthe CANFLEX fuel bundle.

SIDESTOP MEASUREMENTS

The sidestop fixture is a simple test fixture with twosidestop pieces that can be positioned so as tosimulate different penetrations of the CANDU 6 EMsidesteps (see Figure 2). This fixture was made toallow easy comparison between bundle types, fortypical sidestop insertion distances. The sidestopmeasurements were carried out in two stages:clearance measurements, and bundle endmeasurements.

Although this fixture was designed as closely as possibleto reproduce the sidestop configuration, it is recognizedthat there may be differences in dimensioning an actualFM because of tolerancing considerations. Nevertheless,we compared the CANFLEX measurements also withtypical CANDU 6, 37-element bundle measurements,and we concluded that this is a valid method for testingfor bundle compatibility in a typical FM, prior to actualhot tests in a machine.

Clearance With Sidesteps

Measurements were taken to determine theclearances between the bundle end plate and thesidesteps set at the maximum penetration orinsertion distance into the FM liner tube. Theclearances were measured for CANFLEX fuelbundles, and were compared with those of standardCANDU 6 production bundles.

To set up the sidesteps with the maximumpenetrations, first the sidestop fixture was set in V-blocks on a granite surface plate. Then, one of thesidestop pieces was accurately located at the so thatthe flat edges of the sidestop piece (see Figure 2)were perpendicular to the surface plate. The bottomcorner of the sidestop piece was set flush with thetube inside diameter, the height gauge was set atzero on the sidestop centre line. Then the sidestoppiece was moved down to the desired penetration.The setup was repeated for the second sidestop piece.

The clearance measurements were repeated for thestandard CANDU 6 bundles for comparison. Figure2 depicts the overall clearances between the insertedsidesteps and the end plates on the CANFLEX andstandard CANDU 6, 37-element bundles. The actualmeasured clearances are compared in Table 1.

Table 1: Typical Clearance Measurements

SidestopCorner

TopLeftTopRightBottomLeftBottomRight

Clearance (mm)

Bundle1

0.84

0.84

0.38

0.45

2

0.73

0.79

0.20

0.40

3

0.73

0.73

0.38

0.30

4*

1.02

1.14

0.48

0.79

* CANDU 6,37-element fuel bundle

139


Bundle End Measurements

Very accurate bundle end measurements wereconducted on CANFLEX bundles for directcomparison with CANDU 6, 37-element bundles.These include: dimension "B" which is theperpendicular distance from the bottom of the FMliner, or sidestop tube, to the top surface of theelement end cap at the 12 o'clock position, anddimension "E" which is the perpendicular distancefrom the bottom of the liner to the top of the endplate". These are shown in Figure 3.

To prepare for bundle end measurements, the jig wasplaced on two precision parallel blocks, which wereset on a granite surface plate. This was done toenable an indicator on the height gauge to be used tomeasure the top elements and end plate. Theindicator was set at zero representing the bottom ofthe radius surface of the sidestop fixture. Using theheight set at the appropriate zero, measurements "B"and "E" were taken on both ends of the bundles.Precise measurements were obtained and recorded asthe indicator reached the top of the end cap and endplate. Table 2 tabulates typical bundle endmeasurement

Table 2: Typical Bundle End Measurements

Bundle

1234

Average Bundle End Measurements(mm)

B100.52100.4910035100.47

E96.7696.7197.1796.97

6. GRAPPLING TESTS

The fuel grappling tool is designed to defuel a fuelchannel using a single fuelling machine when theother fuelling machine is disabled or when a channelis to be de-fuelled for routine maintenance.

When the CANDU 6 grappling tool approaches afuel bundle, a plunger contacts the end plate. Theplunger will depress, releasing the ten extractingfingers behind the end plate to latch onto the fuel-bundle end plate.

The objective of these grappling tests on theCANFLEX fuel bundles was to determine howeffectively the CANDU 6 grappling tool couldgrapple the CANFLEX bundle. To achieve this, theCANFLEX bundle was placed horizontally inside apressure tube of sufficient length with a cut out at itstop portion for observation. Each CANFLEX bundlewas grappled a number of times by the grapplingtool. Two bundle orientations were tried. In the firstone the lowennost element occupied the 6 o'clockposition, whereas in the second the midpointbetween the two lowermost elements was at thatposition. A minimum of 10 trials for each orientationwas made. In each of these orientations thegrappling tool was oriented randomly and thenumber of fingers grappling the endplate wasrecorded. Typical results from these tests aretabulated in Table 3.

Table 3: Typical Grappling Test Results

Trial no.

12345678910

Number of Fingers Engagedouter element at

6 o'clock

Bundlel7543355555

Bundle23543445644

gap betweenouter elements at

6 o'clockBundlel

4453545533

Bundle24454566545

7. DISCUSSION

The CANFLEX bundles passed through the benttube gauge under their own weight This simple testshowed that the CANFLEX bundle diameter iscompatible with the rolled joint sections of the fuelchannels.

The diameter measurements of the CANFLEXbundles snowed that the maximum measureddiameter of 102.28 mm was less than the maximumdiameter specified for the CANFLEX bundle.Although this dimension was less than that of thestandard 37-element fuel bundle, it was consistent

140


with the smaller diameter of the CANFLEX bundleouter elements. The measurements were alsoconsistent with the results of the bent tube-gaugemeasurements.

The diametrical measurements on the CANFLEXbundles were repeated while the bundles were undercompression at the mid-plane. These measurementsshowed that the flexibility of the bundles was suchthat when bundles were clamped, the maximumdiameter, measured at the end bearing pads, wasreduced slightly. The outside diameter of the endplate at the end cap welds and between the weldswas also measured.

The CANFLEX bundles' dimensions were measuredbefore the compatibility test in CANDU 6 FM usingthe CMM. These measurements showed that allbundle dimensional specifications, including endplate waviness, met the relevant requirements. Themeasurements will be repeated after the FM test toensure that the bundles' dimensions will stay withinthe specified limits.

When considering how the CANFLEX bundlesinterface with the CANDU 6 FM, there are severalfeatures of particular interest During refuelling, it isdesirable that the sidesteps engage as fully aspossible with the endcaps of the fuel elements. If thebundle is low in the liner tube, the engagement willbe too small, particularly at the top of the bundle,and the sidesteps at the bottom corners might touchthe endplates. If the bundle is high, the sidestops atthe top corners might touch the endplate. This willtend to push down on the bundle. Also, when thesidestops enter the FM liner tube, it is equallyimportant that there be no interference across thehorizontal diameter. In this case, if the end plate istoo large in diameter, the sidestops may squeeze anddamage the end plate.

The sidestep measurement test was designed toinvestigate these effects and to use the results toevaluate the adequacy of the interface between thebundles and the FM sidestops.

These sidestep "clearance" measurements with theCANFLEX bundles show that, even for themaximum sidestop penetration, there are clearancesbetween the end plate and the bottom corners of thesidestop: a range from 0.20 mm to 0.45 mm.Although these clearances are less than those we

measured for a typical standard CANDU 6 bundlei.e., minimum of 0.48 mm, the measurementsshowed that there will be no interference at thebottom corners of the sidestops. Thus the CANFLEXbundle's end plate will not be squeezed as thesidestops enter the liner tube.

The bundle end measurements showed that theminimum "B" and maximum "E" dimensions werewithin specified limits for the operationalrequirements of the CANDU 6 FMs. Themeasurements snowed that there will be an adequateengagement with the fuel endcaps, especially at thetop corners of the sidestops. The bundle endmeasurements also showed that there areapproximately 20% more elements engaged for theCANFLEX fuel bundle than that for a typical 37-element bundle (see Figure 2). This will ensure thatthe fuel will always be held well against the flow,hydraulic drag force; and there will be a minimalchance that a bundle will be pushed by the coolantflow past the sidestop, particularly in the case offailure of a single sidestop.

The grappling tests of the CANFLEX bundles, witha CANDU 6 grappling tool head, showed that aminimum of three fingers always engaged on the endplate, which is sufficient to pull out the bundle inemergency situations. This was more than theminimum criterion of two tool fingers that wasrequired to engage the bundle endplate at each trial.

8. CONCLUSIONS

The measurements taken with the CANFLEX fuelbundles confirmed the following:

a) Each CANFLEX bundle passed through the baittube gauge under its own weight

b) The bundles' dimensions were all within thespecified limits.

e) There were always clearances between the FMsidestops and the CANFLEX bundle end plate whichensured no interference.

d) The bundle end dimensions "B"within the specified limits.

and "E" were

e) The CANDU 6 grappling tool engaged on thebundle with a minimum of three fingers.

141


Therefore, it was concluded that the CANFLEX fuelbundles are compatible with the CANDU 6 FM and arenot expected to interfere with the separator/sidestopassemblies. This will be verified in the actual CANDU 6FM compatibility test which is scheduled forApril 1996.

9. ACKNOWLEDGMENTS

The authors would like to thank the CANFLEX projectfor financial support and for permission to publish thispaper. We would also like to thank A.M. Burns andT. White and their co-workers, involved in the bundlemeasurements used in this paper, and D.J. Koivisto andG.A. Corbett from the Fuel Handling Section forconsultations during the work. Finally, we thankA.D. Lane and P.G. Boczar for their comments.

142

Pistons

Pusher

/

Sensor

Sidestep

Figure 1: Fuelling Machine Separator Assembly

143

Parallel Flat Edges,typical .

a.CANFUEX b.CANDU6,37-Element

Figure 2: Clearance Between Sidesteps and End Plate

(Top of end CapE ° toLinerTube)

/ / / / / / . / / / / / / / / / / A

Figure 3: Bundle End Dimensions

144

FIRST INTERNATIONAL CONFERENCE ON CANDU FUEL HANDLING SYSTEMS, MAY-1996

DESIGN OF FUELLING MACHINE BRIDGE AND CARRIAGETO MEET SEISMIC QUALIFICATION REQUIREMENTS

by

A.B. GHARE, A.G. CHHATRE, A.K. W A SAND H.S. BHAMBRA

Nuclear Power CorporationVikram Sarabhai Bhavan, Anushakti Nagar

Bombay - 400 094, India

CA9700734

ABSTRACT

During each refuelling operation, the boundary ofprimary heat transport system is extended up toFuelling Machines. A breach in the pressureboundary of Fuelling Machine in this condition wouldcause a loss of coolant accident. Fuelling Machinesare also used for transit storage of spent fuelbundles till discharged to fuel transfer system.Therefore fuelling machine including its supportstructures is required to be seismically qualified forboth on-reactor ( coupled ) mode and off-reactor(uncoupled) mode.

The fuelling machine carriage used in the firstgeneration of Indian PHWRs is a mobile equipmenton wheels moving over fixed rails. As thisconfiguration was found unsuitable for withstandingstrong seismic disturbances, bridge type design withfixed columns was evolved for the next generationof reactors.

Initially, the seismic analysis of the fuelling machinebridge and carriage was done using static structuralanalysis and values of natural frequencies forvarious structures were computed. The structureswere suitably modified based on the results of thisanalysis. Subsequently, a detailed dynamic seismicanalysis using finite element model has beencompleted for both coupled and uncoupledconditions. The qualification of the structure hasbeen carried out as per ASME section III Division 1,sub section NF.Details of the significant design features, static anddynamic analysis, results and conclusions are givenin the presentation.

1.0 INTRODUCTION

Fuelling machines of Indian PHWRs are designed tofacilitate on power loading and unloading of fuelinto/from the reactor core. Two identical fuellingmachines are clamped on to a coolant channel andthe sealing and shielding plugs are removed prior tomovement of the fuel. The boundary of the primaryheat transport system is thus extended to includethe fuelling machines during every refuellingoperation. A breach in the pressure boundary of thefuelling machines in this condition would cause aloss of cooling accident. Therefore, fuelling machineincluding the support structures are required to beseismicalfy qualified, for on-reactor ( or coupled)mode. The fuelling machines, after receiving thespent fuel are moved to fuel transfer port, todischarge the spent fuel to the fuel transfer system.Fuelling machines thus act as a transit storage forspent fuel and provide cooling to the bundles duringthis period. The fuelling machines including thesupport structures, thus are also required to beseismically qualified for off- reactor ( or uncoupled)mode.

The fuelling machines used in the first generation ofIndian PHWR's (RAPS and MAPS) consisted of twocolumns mounted a small carriage base structurefitted with wheels and it moves horizontally on railsembedded in the floor (see fig.1). As thisconfiguration was found to be unsuitable forwithstanding strong seismic disturbances anticipatedat the being considered sites for nuclear reactors,the fuelling machine carriage was redesigned asbridge type with fixed columns for future generationof standardised design of 220 MWe PHWRs,beginning with Narora (see fig.2). The same bridge

145


concept has been adopted for the 500 MWe PHWRreactors.

In this presentation, first the significant designfeatures, considerations for design and methodologyfor static seismic analysis of fuelling machine bridgeand carriage are discussed. In the latter part of thepresentation the dynamic seismic analysis offuelling machines and its support structures incoupled and uncoupled modes is discussed.

2.0 DESCRIPTION

The fuelling machine (FM) (See Fig.2 and 3)consists of FM head which contains complicatedmechanisms for manipulating the fuel, plugs andother accessories and FM bridge and carriage whichsupports and positions the FM head to any desiredreactor channel or FT port.

The head consists of front snout, pressure housingwhich houses magazine and ram housing at therear. The head is supported on the support frame.The support frame, in turn is supported in the lowergimbals through horizontal trunnion pins and is freeto tilt in a vertical plane. A spring loaded levellingmechanism keeps FM head horizontal and alsolimits the tilt. The lower gimbals is bolted with uppergimbals, to facilitate the removal of FM head forservicing. The upper gimbals is supported on 4linear ball bearings fitted on the top beam. Therelative motion between upper gimbals with respectto top beam provide Z-motion for FM head. The topbeam is suspended from the drive plate by a verticalspindle. The vertical spindle through a set of gearspermits 90 deg. rotation of FM head to facilitate itspassage through hatchway into the service area forconnection to FT port or for maintenance of fuellingmachine. A small amount of tilt about the verticalspindle in horizontal plane is allowed at the extremepositions of 90 deg. rotation. A spring loadedcentralising mechanism keeps the head incentralised position and also limits the tilt in thehorizontal plane.

The drive plate is bolted to the trolley by 4 supportstuds. The trolley moves horizontally over theguides fixed to the underside of the bridge structure.The trolley is moved by a rack and pinion drive.

The bridge structure is a long welded beam, which issupported on the bridge supports. Bridge supportsare guided vertically on guides fitted to twocolumns. One end of the bridge is bolted to the

bridge support, the other end is freely supported onlinear bearings to facilitate thermal expansion of thelong bridge structure. The bridge supports aresupported on 4 ball screws which are suspendedfrom the top of two columns. Each column is a builtup l-section which is supported on a base plateembedded in concrete and is tied with the nearestwall with help of the tie members.

All structural members like the column, the bridgestructure, the bridge supports, the trolley, the driveplate, the gimbals and the support frame aredesigned to be fabricated as welded structures fromASME A 515 gr. 70 material.

3.0 CONSIDERATIONS FOR DESIGNAND METHODOLOGY FOR STATICANALYSIS

Fuelling machine bridge and carriage has beendesigned as per ASME section III subsection NF asapplicable to class I components. However, theprimary considerations of rigidity of structures hasbeen incorporated to limit the deflections due tonormal operating loads. For example, the cross-section of bridge structure has been so sized thatthe deflection of bridge for the worst combination ofoperating loads should not exceed 1.5 mm.

Initially, the effect of seismic loads due to postulatedseismic event was considered in the design on thebasis of static structural analysis. The analysis wasdone using the moment distribution method bycalculating relative stiffness of various members.The end fixity conditions for various fixed boltedjoints were conservatively fixed between 0.5 and0.75. The calculations for bridge and column weredone to determine the worst position of bridge onthe column and FM head on the bridge, and todetermine maximum deflections at such a position.The values of natural frequencies of vibration indifferent directions were computed. Similarcalculations were done for all other structuresincluding gimbals and vertical spindle. Thesecalculations resulted in the following modifications inthe design:

a) Originally an integral radiation shield on top of thebridge was contemplated for shielding of FM servicearea from FM vault. The concept was changed anda separate Roll-on shield was provided on Fm Vaultfloor.

146


b) The columns were tied at a number of points tothe wall by tie members to provide more rigidity byreducing lengths of unsupported spans.

c) Additional bracings were provided in the column,bridge and gimbals to strengthen the structures.

d) The rating of X and Y drive brakes wereaugmented considering the forces generated byseismic disturbance.

4.0 DETAILED DYNAMIC SEISMICANALYSIS

Detailed dynamic seismic analysis using finiteelement technique for both coupled mode anduncoupled mode has been carried out for the newdesign of fuelling machine bridge and carriage byusing SAP-IV computer code (Ref.1)

4.1 FINITE ELEMENT MODELLING

The finite element model using 3-D beam and 3-Dstiffness element of SAP-IV computer code hasbeen developed for both coupled and uncoupledmode of fuelling machines. Fig.4 shows the modelfor the coupled mode. Fig.5 shows closer details ofmodel for FM head and carriage. The model hasbeen evolved so as to depict the structures and thejoints as realistic as feasible.

The fuelling machine bridge and carriage has anumber of connecting node points where theconnecting members are having multiple 'FREE'degrees of freedom. Suitable 'end release code'facility of SAP IV computer code is used at suchnodes to simulate the exact end conditions andrelative motions of various elements in the abovemodels. The connection between ball screws andcolumns, ball screws and bridge support, bridgesupport and bridge, drive plate and vertical spindle,top beam and upper gimbals etc. are some of thepoints where end conditions and relative elementmotions have been simulated in detail.

4.2 QUALIFICATION METHODOLOGY

Fuelling Machine Bridge and Carriage has beenanalysed for dead load and two levels of earthquakeviz. operating basis earthquake (OBE) and safeshutdown earthquake (SSE). The dynamic analysisusing response spectra method has been doneusing SAP- IV code. SSE analysis has been carriedout using 4 % damped SSE response spectra

(Fig. 6) for horizontal direction.(Ref.2) The SSEresponse spectra used is envelope of responsespectra for different floors/wall elevations supportingthe FM bridge and carriage structures. Multiplicationfactor of 0.67 has been used for vertical directionspectra accelerations. The value of damping waschosen as per the ASME code (Ref.3) OBE analysishas been done using 70 % of 2 % damped SSEresponse spectra (see fig.7).

Qualification of the structures has been done as perrequirements of ASME III division 1 sub section NF( Ref. 4 ) . The load combination for supports fordifferent loading categories are taken as follows:

Design condition - Dead wt.+OBE loadsLevel C - Dead wt.+SSE loads.

4.3 RESULTS

The dynamic analysis of fuelling machine bridgeand carriage has been conducted to extract 32modal frequencies for coupled mode and 16 modalfrequencies for uncoupled mode ( see table-1). Thestresses and deflection experienced by variousstructures in different elements were calculated andchecked. Table 2 and 3 show the maximumstressed structural members. The deflections forvarious modal frequencies were combined by SRSSmethod. Table 4 shows the maximum deflection forcoupled and uncoupled mode.

5.0 CONCLUSION

The results of the detailed dynamic seismic analysisshows that the structures are relatively rigid for theseismic excitations. The results have also givenstrength to the view that initial conservative staticanalysis calculations remain a useful tool in theprocess of designing such structures.

147

MRS I IN ILRNAIIUNAL CONIEKJiNCL ON CANDU WILL HANDLING SYSTEMS, MAY-1990

REFERENCES :

1. Computer Code SAP- IV ( A structural analysisprogramme for static and dynamic response oflinear systems by K.J. Bathe et.al,EngineeringAnalysis Corporation, Berkeley, USA)

2. Earthquake Engineering studies Eq-7771,Seismic Analysis of RB of NAPP under horizontalground motion including flexibility of internalstructures, A.S.Arya et. al. University of Roorkee.

3. ASME Boiler and Pressure Vessel Code, SectionIII, Division 1, Appendices, Para N-1230.

4. ASME Boiler and Pressure Vessel Code SectionIII Division 1, subsection NF, Components Supports,para NF-3320.

RAPS /MAPS FUELLING MACHINEFIG. 1

148

RST INTERNATIONAL CONFERENCE ON CANDU FUEL HANDLING SYSTEMS, MAY-1996

1 F H HEAD2 GIMBALS3 BRIDGE« BALL SCREW5 COLUMN

NAPS FUELLING MACHINEFIG. 2

I FUELLING MACHINE HEAD WI1HSUPPORT FRAME

I IOWER GIMBAl SUB ASSY3 UPPER &IMBAL ANO DRIVE PIATC ASSrI TROUEV5 IEVEUJNG ASS*S CENIREUSlrC ASSr7 X DB1VE ASSVa FiNEYonivf5 90'BOlAItOW MECHANISM10 7 DRIVE

F M. HEAD & CARRIAGE

NO OF NODES A35NO OF ELEMENTS 592NO OF BEAM ELEMENTS 5 90NO OF SPRING ELEMENTS 2

FINITE ELEMENT MODEL OFFUELLING MACHINES IN COUPLED MODE

FIG- 4

H

3sno

1n§n

O

a

O


FINITE ELEMENT MODEL OFF. M. HEAD & CARRIAGE

FIG. 5

> 1 2 - 3 UPERIOD IN SECOND

5SE ENVELOPE RESPONSESPECTRUM WITH 4% DAMPING

FIG. 6

IN SECONO

QBE ENVELOPE RESPONSE5PECTRUM WITH 2 % DAMPING

FIG.7

151


MODENUMBER

1,23,45-1617181920212223242526,272829303132

]

TABLE -1

MODAL FREQUENCY DATA

COUPLED MODE

FREQUENCY(CYC/SEC)

3.6373.6403.6635.9336.0186.1396.1919.36910.2310.6715.4215.4918.2322.2222.7531.3431.3534.48

PERIOD(SEC)

0.27500.27470.27300.16860.16620.16290.16150.10670.097780.093700.064840.064580.054840.045010.043950.031900.031900.02900

UNCOUPLED MODE

MODENUMBER

12345678910111213141516

FREQUENCY(CYC/SEC)

3.2613.6373.6403.6633.6633.8723.8733.8743.8746.3269.04210.28017.23023.94031.21040.920

PERIOD(SEC)

0.30660.27500.27470.27300.27300.25820.25820.25810.25810.15810.11060.097240.058030.041760.032040.02444

TABLE-2

MAXIMUM STRESS FACTORS(UNCOUPLED MODE)

STRESS TYPE

SHEAR STRESS

AXIAL COMPRESSIVE STRESS

AXIAL TENSION STRESS

COMBINED AXIAL COMPRESSIONAND BENDING STRESS

COMBINED AXIAL TENSIONAND BENDING STRESS

DESIGNCONDITION

0.6053

0.1428

0.3399

0.9668

0.9654

152

LEVEL CCONDITION

0.4484

0.1200

0.2974

0.6953

0.6941

STRUCTURALMEMBER

RAM HOUSING

SUPPORT STUD

SUPPORT STUD

RAM HOUSING

RAM HOUSING


STRESS TYPE

TABLE-3

MAXIMUM STRESS FACTORS(COUPLED MODE)

DESIGNCONDITION

SHEAR STRESS 0.0798

AXIAL COMPRESSIVE STRESS 0.7679

AXIAL TENSION STRESS 0.1499

COMBINED AXIAL COMPRESSION 0.9300AND BENDING STRESS FACTOR

COMBINED AXIAL TENSIONAND BENDING STRESS FACTOR

0.7080

LEVEL CCONDITION

0.06317

0.7176

0.1102

0.8230

0.5734

STRUCTURALMEMBER

UPPER GIMBAL

COOLANT TUBE

SUPPORT STUD

COOLANT TUBE

END FITTING

TABLE-4

MAXIMUM DEFLECTIONS

STRUCTURAL MEMBER

l.RAM HOUSING TAIL END(Y-DIR)

2.BALL SCREW (X-DIR)

UNCOUPLED MODE( CM.)

3.3

1.8

COUPLED MODE( CM.)

0.8

1.8

153

gss

£J£XT PAGE(S)I! Sef t BLANK


STRUCTURAL ANALYSIS OF FUEL HANDLING SYSTEMS

by

L. S. S. LeeAtomic Energy of Canada Limited

2251 Speakman DriveMississauga, Ontario, Canada L5K 1B2

CA9700735

ABSTRACTThe purpose of this paper has three aspects: (i) toreview "why" and "what" types of structuralanalysis, testing and report are required for the fuelhandling systems according to the codes, or neededfor design of a product, (ii) to review the inputrequirements for analysis and the analysisprocedures, and (iii) to improve the communicationbetween the analysis and other elements of theproduct cycle.

The required or needed types of analysis and reportmay be categorized into three major groups: (i)Certified Stress Reports for design by analysis, (ii)Design Reports not required for certification andregistration, but are still required by codes, and (iii)Design Calculations required by codes or needed fordesign.

Input requirements for structural analysis include:design, code classification, loadings, andjurisdictionary boundary. Examples of structuralanalysis for the fueling machine head and supportstructure are given.

For improving communication between the structuralanalysis and the other elements of the productcycle, some areas in the specification of designrequirements and load rating are discussed.

1. INTRODUCTIONQuite often the analysis part of the product cycle ishidden in the underworld or undervalued if notignored, or is taken for granted. This paper reviewsand highlights the main features involved in thestructural analysis and hopefully promotes mutualunderstanding and appreciation among the analysisand other elements of the design-analysis-fabrication-installation-operation-and-maintenanceproduct cycle of the fuel handling systems in aCANDU nuclear power plant.

2. WHY AND WHAT TYPES OF ANALYSIS ANDREPORT

Why and what types of analysis and report arerequired or needed? As part of the ComponentDesign Document (per CSA N285 series. References1 to 4) or the Design Output Documents (per ASMEBoiler and Pressure Vessel Code, Article NCA-3550,Reference 5), types of analysis, testing and reportrequired for structural integrity achievement for fuelhandling components and supports are summarizedbelow:

(1) Certified Stress Reports for design by analysis(together with Certified Review of Design Report bythe owner or his designee) are required for designregistration for the following components andsupports:

(a) Class 1, 1C and 4 components and theirsupports.

(b) Class 2 and 2C vessels designed to ASME NC-3200.

(c) Class 2 or Class 3 components designed toService Loading greater than design loading.

Due to the early contract award requirements forsome of the major fuel handling equipment aProvisional (Preliminary) Design Registration isusually applied and obtained when, at the time ofapplication, the final Design Report or the finaldrawings are not completed. This allows forcomponent fabrication or system installation tocommence before a final design registration isobtained. In such a case, a Certified ProvisionalReport (or Base Design Report) may be substitutedfor a final design report. The Provisional Report onlyhas to include the Design and Testing Conditions,not any Service (A, B, C and D) Conditions (seeSection 4 for the design requirements). The purposeof the stress calculations is to show that primarystresses are within allowable limits to proveadequacy of material thickness.

155


(2) Certified Design Report Summary may befurnished in lieu of a Certified Design Report forstandard supports (Class 1, 2, 3 or 4) designed byanalysis (provided by the manufacturer).

(3) Design Reports not requiring certification andregistration are required by the codes for:

(a) Class 2 and 2C non-standard support.

(b) Class 3 and 3C pressure-retaining systems andcomponents.

(4) Load Capacity Data Sheet and catalogue forClass 1, 2, 3, 4 and 6 standard supports to bequalified by load rating method. The Load CapacityData Sheet shall identify the tests and calculationsused to establish the load capacity (per Article NCA-3551.2). This is important in order that the user(designer) can interpret correctly the load ratingsused in the design and analysis (see Section 6below). Also, it shall be certified by a RegisteredProfessional Engineer for supports for Class 1components, Class 4 vessels, and Class 2 vesselsdesigned to NC-3200.

(5) Design calculations are needed for design andupon regulatory request are required for:

(a) Class 3 and 3C component supports.

(b) Class 6 pressure-retaining components and theirsupports.

(c) Design deviation due to fabrication errors (e.g.during post-order engineering).

(d) Modification of an existing design, e.g. change ofconfiguration or material.

(e) Change of operating conditions, e.g. hot orambient temperature F/M D2O supply.

(f) Change of installation conditions, e.g. deviationof bolt pre-load torque, material substitute, layoutchange, etc.

(g) Item replacements or repairs due to fabricationdiscrepancy, actual or anticipated item failure (e.g.fatigue life shorter than the plant life).

(h) Conceptual design for firming up the membersizes.

(i) Others as needed, e.g. for intervening elements.

The Fueling Machine (F/M) Head Pressure Boundaryis classified as Class 1 pressure-retaining componentin CSA N285.O and N285.1, and by reference, isdesigned to the requirements of the ASMESubsection NB. The Fueling Machine Supportsystems are usually composed of structuralsupporting elements and mechanisms which areclassified as Class 1C supports in CSA N285.0 andN285.2, and by reference, are designed to therequirements of the ASME Code, Subsection NFClass 1 component support. The classification of1C, instead of Class 1, component support in CSAN285.2 is because that there is no rules in theASME code for mobile support. Portions of the F/Msupports, (e.g. an elevating bridge and carriage, anda mechanism such as a ball screw and nutassembly) have a mobility not usually found insupports for pressure-retaining components. Oneimportant consequence of the CSA classification isthat materials specified in CSA-N285.6.9 can beused, which otherwise may not be ASME NFmaterial.

No matter why or what type of analysis anddocument is required, the essential purpose ofanalysis is to ensure the structural integrity and theadequacy of the product design under variousservice conditions.

3. CLASSIFICATION AND JURISDICTIONARYBOUNDARIES OF COMPONENTS AND SUPPORTSClassifications of pressure-retaining systems,components, and their supports are stipulated inCSA N285.0. The jurisdictionary boundaryconsideration involves boundaries betweencomponents and their supports, attachments,intervening elements, and building structures. Theboundaries for Class 1 , 2 , 3 , 4 and their supportsare shown in NB-1130, NC-1130, ND-1130, NE-1130 and NF-1130 respectively. The specificboundaries of jurisdiction between these structuresshall be clearly defined, e.g. in the DesignSpecification.

For the purpose of defining the jurisdictionalboundary between a component or piping supportand the building structure, they should be shown inrespective drawings (civil/structural drawings orsupport drawings). The key criterion for NF supportstructure is that they are installed and used for the"primary purpose" of supporting piping orcomponents. In general, a bolted connection

156


between the NF support structure and the buildingstructure should be designed as part of the buildingstructure. Whereas, if the means by which the NFstructure is connected to the building structure is aweld, the weld shall fall within the jurisdiction of theNF support. One simple rule for determining whetherthe connection design belongs to NF can be that:the analysis of the connection design need notinvolve the design of the building structure,otherwise, the connection design should be part ofthe building structure.

An attachment is an element in contact with orconnected to a component or support structure. Itmay have either a pressure retaining or nonpressure-retaining function and either a structural ornonstructural function. Structural attachment whichhas pressure retaining function or is in the supportload path should be treated as part of that pressure-retaining component or the support structure.

Sometimes confusion may be caused by intermittentstructural elements as to whether they should betreated as NF supports or Intervening Elements.Intervening Element by definition is a structuralelement in the support load path for a pressure-retaining component for which a major purpose isother than to "passively support" the component. Itis noted that the "means" (bolted or welded) bywhich the component or piping support is connectedto the intervening element shall fall within thejurisdiction of component or piping support.Intervening Elements are not registered, but whenrequired by the regulatory authority, design criteriaand supporting calculations shall be submitted.

4. DESIGN REQUIREMENTS AND ANALYSISPROCEDURESInput requirements for structural analysis must beprovided before analysis can proceed. The designbasis of plant and system operating and testingconditions is stipulated in NCA-2140 and is basedon the system safety criteria and operability ofcomponents and supports. Based on the plant andsystem operating and test conditions. Design,Service (A, B, C and D), and Test Loadings areestablished; its criteria are explained in NCA-2142.4

The structural analysis is performed by a designer oranalyst as an N Certificate Holder. The structuralintegrity and safety achievements are ensured andstipulated in the codes partly by:

(a) control of material.

(b) only certain types of design or construction areallowed, e.g., for welded and bolted flangedconnections.

(c) acceptable stress analysis procedures and stresslimits with safety margins which are compatible withthe class of construction and specification ofloadings.

On the other hand, the operability of components(whether the design works for the functionalpurposes), including leaking, seem not to beemphasized by the codes. The assurance ofoperability is up to the owner (or his designee) todefine the appropriate limiting parameters. However,code rules apply to the operability of pressure reliefvalves.

4.1 Design RequirementsThe rules for construction of Class 1 , 2 , 3 and 4(MC) nuclear components and their supports aregiven in the ASME Code, Subsection NB, NC, ND,NE and NF respectively. The owner shall provide orcause to be provided the design requirements and adesign verification report for intervening elements.Design requirements should be defined indocuments, e.g. design specifications. Article NCA-3252 stipulates the required contents of DesignSpecification. As a minimum, the designrequirements should include the followinginformation:

(a) Design, including drawings, and materialspecifications including impact tests.

(b) Code classification of components.

(c) Jurisdictionary boundary.

(d) Loadings for Design, Service and TestConditions.

The loading conditions that shall be taken intoaccount in designing component or support arespecified in Article NB-3111, NC-3111, ND-3111,NE-3111 and NF-3111 respectively for Class 1, 2,3, 4 components and their support structures.

4.2 Analysis ProceduresThe rules of analysis procedure for Class 1,2,3 and4 (MC) nuclear components and their supports aregiven in the ASME Code, Subsection NB, NC, ND,NE and NF respectively. Requirements for

157


acceptability of design are stipulated in Article NB-3211, NC-3211.2, ND-3300, NE-3211 and NF-3131. They can be demonstrated by analysis orexperiment tests. The analysis method can be eitherdesign-by-analysis or by design rules (e.g. ND-3300). For design by analysis, the classical methodand/or finite element method can be adopteddepending- on the complexity of the structuregeometry, the load types (pressure, temperature orseismic loads, etc.) and the requirements of thecode allowable stress limits (primary vs. secondarystress, allowable stress intensity vs. allowablemaximum stress, etc.). Highlights of theconsiderations for structural analysis in accordancewith Subsection NB and NF are given below.

Class 1 Components (NBI(1) The design details shall conform to the generalrules given in NB-3130, including the minimumrequired thickness of shells.

(2) The stress limits for Design, Service, and TestConditions are based on the stress intensity (i.e.maximum shear stress theory). Fatigue evaluationshall be considered for Service Lever A, B and TestConditions.

(3) Protection against nonductile fracture shall beprovided.

(4) Buckling should be evaluated, e.g. under externalpressures.

NF Support Structure(1) Types of supports are given in Article NF-1212,NF-1213, NF-1214. Standard supports and catalogitems are supplied by a Quality System CertificateHolder as material, including Certification of LoadCapacity Data Sheets and Design Report Summary.

(2) Analysis procedure by: (i) design by analysis, (ii)experimental stress analysis, and (iii) load ratingmethod.

(3) NF support needs not include thermal or peakstress, except for high cycle fatigue, n > 20,000cycles, for Class 1 Linear Type support.

(4) Buckling should be evaluated, e.g. for beam typeelements of F/M support bridge and columns.

(5) Protection against nonductile fracture for Class 1component and piping support should be considered.

(6) There are three types of supports: (i) Plate-andShell-Type support, (ii) Linear Type Support, and (iii)Load Rated Support. The stress limits for Class 1Plate-and Shell-Type supports are defined by thedesign stress intensity (Sm) which is based on themaximum shear; others are defined in terms of theallowable stresses (S) which are based on the yieldstrength of material and the maximum stress(principal stress). For bolting, the limits are based onthe yield strength and the ultimate strength.

5. ANALYSIS EXAMPLES

5.1 F/M and Support Structure Seismic AnalysisEarthquake loads are part of the loading conditions(as Level C Service loads) for the F/M head pressureboundary components and the support structure.They are also required for the interfacing systems,i.e., F/M process system, reactor structure, fuelchannels, and feeders of the PHTS. Seismic analysisof the F/M and the support structure is thereforecarried out to generate the seismic loads.

The seismic analysis methodology follows therequirements and procedures of the NationalStandard of Canada CAN3-N289.3-M81 (Reference6). Seismic models have been constructed usingbeam and spring elements for various systems, e.g.the F/M, the support structure and the reactor. Forexamples, see Figures 1, 2, and 3. To account forvarious operation modes during the re-fuelingprocess, seismic models representing variousconfigurations have been developed:

(a) F/M attached or unattached to the reactor in thereactor vault area, with the F/M located at sevenrepresentative fuel channel locations (AH, E03,E20, K11, P02, S20, W11).

(b) F/M on the maintenance lock track, fiveconfigurations were considered: (i) unattached atcentre of track, (ii) unattached at new fuel portlocation, (iii) unattached at spent fuel port location,(iv) attached to new fuel port, and (v) attached tospent fuel port.

For the F/M seismic analysis, the input earthquakesare the F/M support points motions, in terms of floorresponse spectra or acceleration time-histories,which are generated from the reactor buildingseismic analysis by the Civil design group. The inputmotions take into account the effects of thevariation of soil conditions at the site and thesensitivity due to the uncertainties of the structural

158


properties (frequencies). The seismic loads resultedfrom the seismic analysis are represented in terms ofnodal accelerations, beam end loads and third-levelfloor response spectra. These seismic loads are thenused in the seismic qualifications (by analysis or testmethod) of the affected systems.

5.2 Stress Analysis of F/M Head Pressure BoundaryThe F/M head assembly consists of a number ofmajor sub-assemblies: a snout assembly, a magazineassembly, a ram assembly and two separators. Thehousings for these sub-assemblies form the pressureboundary of the F/M head. Loads and loadcombinations for Design, Service Level A, B, C, andTest Conditions for the F/M head pressure boundarywere defined in a design specification. Final detailedstress analysis and report was prepared and certifiedby the author and the third-party reviewer. It formedpart of the submission for the final designregistration in accordance with the requirements ofReferences 1 and 5.

The methodology adopted in the stress analysis ofthe F/M head pressure boundary components makesuse of an optimum combination of classical andfinite element (FE) methods. The stress analysis isbased on linear elastic static analysis except forsome assemblies in which non-linear gap elementsare used in the FE method to simulate theinteraction behaviours under various loadingconditions at the contacted face between twocomponents.

In the finite element method, the trend is to utilizecomputer-aided capabilities for modeling andmeshing. Ideally, the mechanical design automationtools used should able to provide direct interfacebetween the design models, the drafting models andthe analysis solid models. The analysis solid modelsare usually simplified to remove unnecessary details.Figure 4 shows such a FE model for the magazinehousing, which was generated by using I-DEASsoftware package (by vendor SDRC).

5.3 Stress Analysis of F/M Support StructureThe F/M head assembly and the cradle assembly aresupported from the F/M carriage. The carriage issuspended from rails on the bridge in the reactorvault area (Fig. 2) or on the track frame in themaintenance lock area. The F/M support structure isanalyzed in accordance with the analysis proceduresstipulated in Subsection NF (as described in Section4.2) as:

(a) Beam (linear) type elements inassembly, the bridge and the columns.

the cradle

(b) Non-beam type elements (plate and shell) in thecradle, the carriage, the bridge and the columnassemblies.

(c) Load rated mechanisms of manufacturer'sproprietary components in the carriage, and thebridge-elevator interface. There are 19 load ratedcomponents used in the F/M support structure aslisted in Table 1.

6. INTERACTIONS BETWEEN ANALYSIS ANDOTHER ASPECTS OF PRODUCT CYCLEVarious elements of the product cycle need theservice of structural analysis (see Section 2) whilethe latter requires input from the former (see Section3 and 4). Mutual understanding, appreciation, andefficient communication among them are importantfor a successful product. Experience indicates thatsome areas within the interfaces warrantimprovements. Examples are given below:

(1) The required information for analysis should beprovided "timely and adequately". This is vital inorder to avoid repeated analysis. Prior to contractaward, it is essential that the requirements (e.g.design specifications) be clear and available.

(2) Have a section on "Requirements for Analysis"included in the design specifications to providespecific instructions for analysis. Examples can be:

(a) The component can be defined as piping, whilethe analysis, for convenience, can be based on therequirements for vessel.

(b) Analysis can be optionally based on a higherclass (usually means higher allowable limits) thanthe class of which the component is classified. Inthis case, the material requirements for the higherclass shall be satisfied.

(3) For non-standard NF supports, do not specify(e.g. on drawings) the type of support, whetherPlate- and Shell-Type or Linear-Type. The type ofsupport to be assumed in the analysis should bedetermined by the analyst depending on thegeometry complexity and the load distributions.

(4) Procedures for determining load ratings for Plate-and Shell-Type and Linear-Type support are providedin ASME NF-3280 and NF-3380 respectively. The

159


load ratings are defined for Level A, B and C(including seismic) Service Loading. The load ratingsprovided by the manufacturers, in terms of testreports. Load Capacity Data Sheets or catalogues,should be compatible with the NF definitions. Forinstance, the load ratings given in catalogue as"static" or "dynamic" can be determined by testingconditions and failure mechanisms which may bequite different from the required ASME NF testingprocedures or the actual operating conditions.Similarly, the "capacity" rating given in somecatalogues can be established based on the fatiguelife while the load ratings stipulated in the ASME NFprocedure are based on the ultimate load failurecriterion. Any discrepancy in the compatibility of thedefinition between the ASME NF and manufacturer'scan create confusion and mis-application foranalysis.

(5) If a Provisional (Preliminary) Design Registrationhas been applied and obtained, see Section 2,therefore, component fabrication might havecommenced before the results of the final structuralanalysis are completed. It can happen that theService Conditions (considered in the final DesignReport, but not in the Provisional Report), e.g.seismic loads or fatigue life, indicates that thepreliminary design may not be adequate. Thisrequires an extra effort on the structural analysis toremove any fictitious overstress that might result inorder not to modify design during manufacture.Close coordination between the structural analysis,design, manufacturing and maintenance (e.g. itemreplacement due to short fatigue life) are vital inorder to find an acceptable solution.

7. CONCLUSIONS

This paper describes the basic requirements andprocedures for structural analysis in accordance withthe codes. To satisfy the code requirements isnecessary, however, to have the analysis work donein a cost effective manner is vital to the overallsuccess of a project. This paper has highlightedsome areas for the improvement of the workmethod, especially the interface between thestructural analysis and the other elements of theproduct cycle.

8. REFERENCES(1) National Standard of Canada CAN/CSA-N285.0,"General Requirements for Pressure-RetainingSystems and Components in CANDU Nuclear PowerPlants".

(2) National Standard of Canada CAN/CSA-N285.1,"Requirements for Class 1, 2, and 3 Pressure-Retaining Systems and Components in CANDUNuclear Power Plants".

(3) National Standard of Canada CAN/CSA-N285.2,"Requirements for Class 1C, 2C, and 3C Pressure-Retaining Components and Supports in CANDUNuclear Power Plants".

(4) National Standard of Canada CAN/CSA-N285.3,"Requirements for Containment System Componentsin CANDU Nuclear Power Plants".

(5) ASME Boiler & Pressure Vessel Code, Section III,Subsection NCA, NB, NC, ND, NE and NF.

(6) National Standard of Canada CAN3-N289.3,"Design Procedures for Seismic Qualification ofCANDU Nuclear Power Plants".

Table 1 NF-Grade Load Rated Components in F/M Support Structure

No.

123456789

F/M Support Structure

Cradle Trunnion Bearing (Fixed End)Cradle Trunnion Bearing (Free End)Cradle Ram Assembly Cam FollowersCarriage WheelCarriage Wheel BearingCarriage 'Z' Guide Cam FollowersCarriage Cam FollowerCarriage Fine 'Y'-Drive Screw JacksGimbal Roundway Bearing

Load Rated Component

Spherical Roller BearingCylindrical Roller BearingRollerwayWheelWheel BearingCamrol BearingCamrol BearingScrew JackRoundway Bearing

160


TableNo.101112131.41516171819

1 (continued)F/M Support Structure

2" Gimbal RoundwayUpper Gimbal Turntable BearingCarriage Seismic ClampCarriage 'Z' Motion DriveElevator Roundway BearingElevator RoundwayBridge Ball ScrewBridge Screw JackBridge Screw NutBridge/Elevator Camrol Bearing

Load Rated ComponentRoundwaySleeving BearingThrust BearingHydraulic CylinderRoundway BearingRoundwayBall ScrewScrew JackBall Nut Thrust BearingCamrol Bearing

fUSLUKMACS1MS

FIGURE 1 FUELING MACHINE AND CARRIAGE SEISMIC MODEL

161


FIGURE 2 BRIDGE AND COLUMN SEISMIC MODEL (for Fuel Channel Location W11)

FIGURE 3 ATTACHED FUELING MACINE AND REACTOR SEISMIC MODEL

162


(a) Global Model

(b) Submodel of the End Cover

FIGURE 4 FINITE ELEMENT MODEL OF F/M MAGAZINE

163

NEXTloft EL AUK

FIRST INTERNATIONAL CONFERENCE ON CANDU FUEL HANDLING SYSTEMS, MAY '96

PICKERING DRY STORAGECOMMISSIONING AND INITIAL OPERATION

by

STEVE JONJEVTECHNICAL SUPERVISOR

FUEL HANDLING UNITONTARIO HYDRO - PICKERING NUCLEAR DIVISION

P.O. BOX 160, PICKERING, ONTARIO, L1V 2R5

CA9700736

ABSTRACTHaving commissioned all individual conventional andnuclear systems, the first Dry Storage Container (DSC)was loaded with four modules of 17 year cooledirradiated fuel (366 bundles) in the Auxiliary IrradiatedFuel Bay (AIFB) on November 29, 1995. Afterdecontamination of the outer surface, and draining ofwater, the DSC was transported to the Used Fuel DryStorage Facility (UFDSF) workshop, where it wasvacuum dried, and then the lid was welded on.Following successful radiography test of the lid weld,the DSC was vacuum dried again and backfilled withHelium to a pressure of 930 mbar(a). The Helium leaktest showed zero leakage (allowable leak rate is 1x10'5

cc/sec). Finally, after loose contamination checks wereperformed and permanent safeguards seals wereapplied, the DSC was placed in the UFDSF storage areaon January 23, 1996.

Radiation fields at contact with the DSC surface were<0.6 mrem/hr, and at the exterior, surface of thestorage building wall only 33 micro-rem/hr (far belowthe target of 250 micro-rem/hr). Therefore, the actualdose rates to general public (at the exclusion zoneboundary) will be well below the design target of 1 % ofthe regulatory limit.

1. INTRODUCTION

Implementation of the dry storage concept at PickeringNuclear Division (PND) followed a successfuldemonstration program, which used two concretecontainers loaded with six and ten years cooledirradiated fuel from the AIFB (1988-1992). Results ofthis program confirmed both the technical andeconomical viability of the dry storage system as analternative to the conventional wet storage system.The current DSC design offers two main features:

• interim dry storage of irradiated fuel for up to 50years

• a means of transportation to its final disposaldestination without the need to re-packageirradiated fuel

The UFDSF will provide additional on-site storage forabout 576,000 irradiated fuel bundles, or approximately1,500 DSCs by the end of station operating life (year2025). The UFDSF currently has an operating licenseas well as a transportation license (for DSC off-siteshipment) granted by the Atomic Energy of CanadaBoard (AECB).

Safeguards are applied to the UFDSF and DSCs as perthe requirements of the International Atomic EnergyAgency (IAEA) regulations. Surveillance cameras areinstalled in the UFDSF workshop and storage area, andpermanent safeguards seals are applied to each DSC byinstalling two fibre optic cables and attaching "cobra"seals.

2. COMMISSIONING

2.1 DSC Pre-Loading ActivitiesNew DSCs are delivered to the UFDSF workshop wherethey are unloaded from a truck and temporarily stored.Prior to transferring the first empty DSC (#OH-0O01) tothe AIFB, it was inspected for corrosion and otherphysical damage. A lid and transfer clamp wereinstalled on the DSC body to verify a proper fit. Also,four empty modules were placed into the DSC to verifyproper clearance between them and the DSC cavity,and that the lid would not contact the uppermostmodule. Each DSC has two ports, a vent at the topand a drain at the bottom. The internal threads ofthese two ports were inspected for any damage andrepaired as necessary. With the lid taken off the DSCbody, the main weld groove was carefully inspected forany rust. None was found. The prepared DSC wasplaced in the temporary laydown area, ready to betaken to the AIFB (ref. Fig. #1).

At the same time as the DSC inspection, two operatorswere recording at the AIFB serial numbers of thebundles to be loaded into the DSC. This was doneusing a telescope and recording equipment speciallyfabricated for this task. The process was quite lengthy(it took about 3 days vs. the 4 hours expected), but, at

165

the end, a complete record of the 366 bundle serialnumbers was available. This requirement came fromthe AECB and IAEA in mid 1995 and was agreed uponby PND. Before the loading of the DSC, IAEA staffperformed the following verifications:

• 100% item count, (i.e. 366 bundles)• Non Destructive Analysis (NDA) of irradiated fuel• correctness of serial numbers by checking 8 out

of 366 bundles

This task was very cumbersome, and took about 12working hours vs. the 2 hours originally planned.

2.2 Loading of DSC in AIFBThe fully prepared DSC was transported from theUFDSF workshop to the AIFB along the route north ofthe powerhouse (Fig.#1) by the on-site transportervehicle (Fig.#3). In the AIFB, the lid was taken off andthe DSC was filled with demineralized water. OnNovember 29, 1995 it was submersed into the AIFB,placed on an impact pad, and loaded with four modulesof approximately 17 years cooled irradiated fuel; a totalof 366 bundles. This operation was performed undersurveillance of the IAEA and AECB staff and variousstation personnel. Once the lid and in-bay clamp werepositioned and engaged on the DSC, it was hoisted outof the water. Total in-water time was 1 hour, 43minutes, which is considerably less than the originallyestimated 3 to 4 hours. Radiation fields on the DSCafter loading were measured as 0.5 mrem/hr at contactand non-detectable at working distance. It should benoted that radiation fields at contact rose marginallyafter draining as the shielding from the water was lost.

The DSC was then immediately washed in thedecontamination pit for about 15 minutes with waterheated to about 75°C. After placing the DSC back onthe handling pad, it was allowed to dry before takingsmears from the surface. The smears showed zerosurface contamination, which was much better thanexpected. It reflects the positive aspects ofmaintaining clean bay water, minimizing in-water timeand ensuring a prompt washing with hot water afterremoval from the bay water.

2.3 Draining and Vacuum Drying in AIFBOn November 30, 1995 the water in the DSC wasdrained back into the AIFB by gravity, with both drainand vent ports open. It took 15 minutes to reach thepoint where no water was flowing through the drainport. The drain port was then closed, and the transferclamp was installed on the DSC flange.

This method of draining, although it had worked wellduring "dry" commissioning, left about 55 litres ofwater in the DSC cavity versus the expected 10 litres.The draining of the water from subsequent DSCs wasvastly improved by tilting the DSC in two different

directions. It is believed that most of the water wastrapped amongst the fuel bundles in the module tubes.

The vacuum drying system was connected to the DSCdrain port and allowed to run for 14 hours. Water wascontinuously removed from the DSC, and internalpressure was lowered from atmospheric to 13 mbar(a).Although the target of 2.5 mbar(a) was not achieved,and the DSC had obviously not been completely dried,it was decided to transport it to the UFDSF and resumevacuum drying there.

2.4 Transport of DSC to UFDSFFinal contamination surveys found only one area on thebase of the DSC with 1000 cpm of loose contaminationwhich was easily removed. This is a sign that thesweating phenomenon that had been observed duringthe demonstration program is not taking place. Gammasurveys found fields still at about 0.5 mrem/hr atcontact and zero at working distance.

By means of a special "O" ring seal placed in thewelding groove, a sub-atmospheric pressure of lessthan 830 mbar(a) was established in the DSC cavity.This sub-atmospheric pressure was maintained duringthe DSC transport through the station yard to preventthe possible spread of loose contamination.

On December 1, 1995, having been picked up by theTransporter vehicle (fig.#3), the DSC was taken to theUFDSF workshop. Along with station and AECBpersonnel, IAEA staff accompanied the DSC as part of"human surveillance" on its 18 minute trip to theUFDSF (fig.#1). The Transporter worked flawlesslythroughout this process. At the UFDSF, the DSC wasfirst dropped off by the Transporter, then moved bycrane to welding station #3. IAEA staff then applied asecurity seal to the base, which was an additionalsafety measure to being monitored by the permanentsurveillance cameras. The DSC internal pressure roseonly marginally during the trip.

2.5 Vacuum Drying in UFDSFUpon arrival in the UFDSF, about 20 litres of water wascollected through the drain port. It is believed that thiswater was forced out from the module tubes byvibrations during the trip through the yard. This wasfollowed by two days of vacuum drying in an effort tocompletely dry the DSC internals. However, internalpressure only got down to 7 mbar(a) and the presenceof water was still evident. In an attempt to attain therequired dry condition that had been observed duringthe testing stage, vacuum drying of the DSC proceedednon-stop for two more days. On December 5, 1995 aninternal pressure of 5.5 mbar(a) was achieved with nofurther water being removed. This was deemed to bean acceptable condition for the initial stage of dryingprior to welding the lid to the DSC body.

166

2.6 Welding and X-rayThe flange weld, about 30 feet long, is made by twoautomatic welding heads, each traveling along thecircumference of the lid on opposite sides (fig.#4).Prior to welding, the DSC flange must be preheated to105°C, while a slight vacuum is drawn in the DSCcavity to prevent any spread of contaminants. Duringearlier tests, it was confirmed that a pre-heat period of16 hours is required to literally "soak" the DSC lid andbody with heat so it will not cool down before the firstwelding pass is done. The design estimate had been 8hours.

Preparation of the DSC for welding of the flange tookseveral days, and some of the work was done inparallel with vacuum drying. First, primer paint had tobe ground off the weld groove down to bare metal.Then, the welding machine was installed and putthrough a series of checks. A problem with one of thetrailing cameras was discovered and quickly rectified.By December 7, 1995 the preheaters and thermo-couples were installed and function tested.

Having sufficiently pre-heated the DSC flange, the pre-weld vacuum system was disconnected and the firstweld pass completed on December 12, 1995.Constant monitoring of the process from the controlroom, as well as frequent local visual inspection by thewelders, was required to ensure a flawless weld. Ittook about two days to complete eleven passes, sevenof which were done on the first day.

The welding machine was removed from the DSC, theX-ray machine installed and the weld allowed to cooldown for 48 hours. In the meantime, the DSC internalswere maintained slightly sub-atmospheric by connectingthe drain port to the active ventilation system. Theflange weld was inspected using the X-ray machine(radiography). To reduce the X-ray hazard, it wasperformed in the evening hours, and was completed onDecember 16, 1995. There were no flaws found andthe weld was deemed acceptable.

2.7 Helium Backfill and Leak TestDuring the final vacuum drying, which took less than 2hours, a pressure of 1 mbar(a) was reached, indicatingthat practically no water remained inside the DSCcavity. It was subsequently backfilled with Helium toabout 930 mbar(a).

Manual welding of the vent and drain ports tooksubstantially more time than expected. However, thewelds were all found to be good and did not require anyre-work. Each weld required 8 hours vs. the 2 hoursoriginally estimated.

The Helium leak test requires repositioning the DSC tostation #4, which has the necessary equipment andinstruments for this test (fig.#5). The unique test

method uses a two piece "bell-jar" so that the wholeDSC can be tested at once. A vacuum is drawn in thespace between the DSC outer surface and the bell-jarso that any leakage from the DSC cavity can be quicklydetected. Having placed the DSC into the bell-jar, ittook 26 hours to pull the required vacuum of 15-20micro-bar(a). This lengthy process was expected and itis mainly due to outgassing of the DSC outer surfacewith its coating of epoxy paint. The Helium leak testwas performed within the next five minutes, and itshowed zero leakage (acceptable leak rate is 1x10"5

cc/sec). For all practical purposes, the DSC was readyto be put into final storage on December 21, 1995.

Having painted the welded surfaces, and havinginstalled the safeguards "cobra" seals, the actualplacing of the DSC in the storage area and a ribboncutting ceremony took place on January 23, 1996.

2.8 DSC History RecordsDetailed information will be computerized, with hardcopies kept in the station's Q.A. vault. Included will beeach DSCs serial number with manufacturing data, themodules' serial numbers, the age of the irradiated fuel,the position of modules within a DSC, the number ofbundles in each module, date of loading, as well as X-ray and Helium leak test results. This has all beenrecorded in a checklist and verified by station and IAEAstaff. The UFDSF storage area has a simple gridsystem to keep track of the DSCs' locations, and eachDSC has a nameplate.

All information listed above is currently kept in the formof a "DSC history docket".

3. RESOURCE REQUIREMENTS

About 314 personhours were required to complete thefirst DSC, from start to finish. Contrary to originalestimates, mechanics were in the highest demand,followed by operators and quality control staff (seeTable 1).

Once the whole process becomes routine, it isexpected to take about 160 personhours per DSC.Table 2 shows individual process activities, type andnumber of resources required and the number of hoursneeded to complete the first DSC.

TABLE 1 - RESOURCE REQUIREMENTS$n person hours}

OP's105

33.5%

MM125

39.8%

SM16

5.1%

QC35

11.2%

IAEA19

6%

P&M12

3.8%

RC2

0.6%

167

4. OCCUPATIONAL RADIATION SAFETY

4.1 Station PersonnelThe shielding provided by the DSC proved to be muchbetter than anticipated, and resulted in gamma fields ofless than 0.6 mrem/hr at contact with the DSC surfaceand zero at 1 metre distance. Also, by keeping the baywater very clean, the gamma background fields in theAIFB are very low {about 0.25 mrem/hr). As acombination of these two factors, the gamma dosereceived by operators was 10 mrem per DSC (5operators were involved). Similarly, the gamma dosereceived by mechanical maintainers was about 10mrem per DSC. Other work groups received nomeasurable gamma dose. There was no Tritium dosereceived while working in the AIFB.

Therefore, the total radiation dose received by all workgroups was about 20 mrem per DSC, which comparesfavourably with the original estimate of about 75 mremper DSC.

4.2 General PublicRadiation fields at the outside surface of the storagebuilding wall were about 33 micro-rem/hr, which is farbelow the target of 250 micro-rem/hr. These radiationfields may marginally rise as the number of DSCs in thestorage building increases. However, the actual doserates to the general public at the exclusion zoneboundary will be well below the design target of 1 % ofthe regulatory limit.

Air emissions to the environment from the UFDSFworkshop (zone 3 from the radiation point of view) aremonitored and were found to be zero. There has beenno active liquid discharge from the UFDSF workshopthus far.

5. INITIAL OPERATION

The initial production runs took place from the "B"Irradiated Fuel Bay (IFB-B) because of the proximity ofits fill-in date. They yielded a transfer capability of four(4) DSCs per two months. Full production transfercapability of eleven (11) DSCs every two months willbe required by early 1997, just before the AIFBbecomes full as well. A total of sixty-six (66) DSCs willhave to be transferred from then on every year,assuming the eight Pickering reactors are operating atan average 85% Capacity Factor.

Major difficulties encountered thus far are listed below:• Draining of water from the DSC was inadequate

causing very long vacuum drying periods. This hasbeen overcome by tilting the DSC in two differentdirections. However, the DSC Lifting Beam sufferedminor damage since it had not been designed fortilting. A design change is expected to take placeshortly.

• Welding of the vent and drain ports took two days. Itwill be reduced by using modified procedures.

• Welding of the lid to the DSC body (flange weld) wasnot satisfactory on the first 3 DSCs. Extensive re-work has been required due to porosity and lack offusion, predominantly in the root of the weld (the firstof eleven passes). An investigation is in progresswhile the welding of the flanges carries on at a slowand resource-consuming pace.

• The helium leak test on its own does not take morethan 5 minutes. However, it requires a vacuum in thebell-jar of 15 - 20 micro-bar(a), which needs in excessof 24 hours to achieve. It is believed that re-calibration of the leak detector to operate at avacuum of 1 mbar(a) will reduce the total test time toless than 4 hours.

• A lack of manoeuvring space in the IFB-B is slowingdown the process. Particularly, the rationale for theportable impact pad is being questioned in view oflimited DSC movements. Also, the manoeuvringsequence of equipment such as the in-bay clampstand and lifting beam stand will have to be improvedupon.

All of the above, plus other minor factors, contribute tothe current transfer rate being about 15 calendar daysper DSC. This transfer rate must be improved to about5 calendar days per DSC by early 1997 in order tomaintain the power level of the eight Pickering reactorsat an average 85% Capacity Factor.

ACKNOWLEDGMENTS

The author wishes to thank Ms. C. Walker and Mr. D.Hunter for their help in preparing this paper.

REFERENCES

1. "Pickering UFDSF Safety Report", Ontario Hydro tothe Atomic Energy Board, 1994.

2. R.N.Sumar and S.Jonjev, "Dry Storage of IrradiatedCANDU Fuel at Pickering NGS", Proceedings ofAnnual Conference, Canadian Nuclear Society,June 1994.

3. D.Hunter, Work Ran #2-95-10768-00 "Commis-sioning of DSC in AIFB", November 1995.

168

TABLE 2 - UFDSF PROCESS

......jf. „,,. X HOURS \

1

2

3

4

5

6

7

8

9

to

11

12

13

14

• RECORD SERIAL #'S OF ALL BUNDLES TO BE LOADED (4 MODULES).. INSTALL 4 MODULES ON THE INSPECTION STAND.• IAEA TO VERIFY W I N 4 MODULES ON THE INSPECTION STAND.• UNLOAD NEW DSC, INSPECT FOR CORROSION OR DAMAGE.• MM TO INSPECT VENT AND DRAIN PORT THREADS AND WELD GROOVE.• PERFORM DSC TEST MODULE CLEARANCE CHECKS. INSTALL DSC LID AND

TRANSFER CLAMP.CRANE PREPARED DSC TO TEMPORARY LAYDOWN AREA. FILL OUTINDIVIDUAL DSC RECORD SHEET.

• TRANSPORT PREPARED EMPTY DSC TO AIFB OR BIFB USING DSC ON-SITETRANSPORTER.REMOVE DSC LID AND TRANSFER CLAMP. REMOVE TRANSFER CLAMP FROMLID AND INSTALL INBAY CLAMP.FILL DSC WITH DEMINERALIZED WATER, SPRAY ALL EXTERNAL SURFACES OFDSC WITH DEMINERALIZED WATER.MOVE DSC INTO IFB, LOAD FOUR (IAEA) VERIFIED MODULES OF USEDIRRADIATED FUEL.INSTALL DSC LID, ENGAGE INBAY CLAMP.

• REMOVE LOADED DSC FROM IFB. MONITOR CONTINUOUSLY AND SPRAY ALLSURFACES OF DSC WITH DEMINERALIZED WATER.

• DECONTAMINATE DSC IN DECONTAMINATION FACILITY. REMOVE INBAYCLAMP, INSTALL TRANSFER CLAMP AND ENGAGE.DRAIN WATER FROM DSC TO IFB VIA DRAIN CONNECTION. PULL PARTIALVACUUM AND VERIFY ACCEPTABLE TRANSFER CLAMP LEAK RATE.

• CONTAMINATION CHECK OF DSC. (LESS THAN 200 CPM ABOVE BACKGROUND)• COMPLETE CONTAMINATION CHECKS OF DSC & TRANSPORTER. DRIVE

LOADED DSC TO UFDSF, UNDER IAEA HUMAN SURVEILLANCE.• POSITION DSC AT SELECTED WELD STATION. REMOVE TRANSFER CLAMP AND

"O" RING SEAL.• PREPARE WELDING GROOVE AND INSTALL WELDING MACHINE ON DSC

• INSTALL FLANGE PRE-HEATERS AND ACTIVATE (16 HRPRE-HEAT). DRAWWELD VACUUM AT DRAIN PORT, WITH VENT PORT OPEN.

• SHUT OFF WELD VACUUM BEFORE THE FIRST WELD PASS IS STARTED.WELD LID TO DSC BODY (AUTOMATED WELD 11 PASSES).

• WHEN FLANGE WELDING COMPLETED, REMOVE WELDING RIG AND PRE-HEATERS.CONNECT VACUUM LINE FROM DRAIN PORT TO HV SUCTION HOLE ON WALL.

• ALLOW 48 HOURS FOR ALL WELDS TO COOL.• INSTALL X-RAY MACHINE.

RADIOGRAPH FLANGE WELD.

MANUALLY WELD VENT PORT.• DRAW VACUUM TO <1 MBAR, AND BACKFILL DSC WITH HELIUM.

MANUALLY WELD DRAIN PORT.• QC VENT AND DRAIN WELDS.

POSITION DSC AT WELD STATION #4. PERFORM DSC He LEAK TEST (BELL JARMETHOD).

• INSTALL SAFEGUARD SEALS (IAEA COBRA SEALS).

• PREPARE DSC FOR WELD AND TOUCH UP PAINTING.COMPLETE ALL PAINTING AND DSC IDENTIFICATION LABEL APPLICATION.COMPLETE FINAL DSC CONTAMINATION CHECKS (SPOT DECONTAMINATION ASREQUIRED).

• ENGAGE DSC WITH TRANSPORTER AND MOVE TO STORAGE AREA.• COMPLETE DSC INDIVIDUAL RECORD SHEET (HISTORY DOCKET).

OPERATORS• IAEA STAFF

• OPERATORS• MM

OPERATORS

• OPERATORS

• OPERATORS

• OPERATORS• IAEA STAFF• MM

RAD CONTROL

• OPERATORS• MM• P & M

• OPERATORS• MM• P & M• OP's (MOVE X-RAY

MACHINE)• QC (RADIOGRAPH

FLANGE WELD)• OPERATORS• MM• OC

• OPERATORS

• OPERATORS• IAEA STAFF• SM (PAINTERS)

• OPERATORS

2X 16HRS.2X8HRS.

2 X 3 HRS.2X2HRS.

2X6HRS

3X2HRS.

3 X 3 HRS.

2 X 3 HRS.1 X I HRS.2 X 4 HRS.1 X 2 HRS.

2 X 1 HRS.3 X 2 4 HRS.3X2HRS.

2 X 1 HRS.3 X 3 HRS.3X2HRS.2 X 1 HRS.

2 X 1 6 HRS.(2 NIGHTS)2X4HRS.2 X 1 6 HRS.1X3HRS.

2 X 6 HRS.(26HRTOPULLVACUUM)1X2HRS.1X2HRS.2 X 8 HRS.

2 X 3 HRS.

169

Legend

o

15 CmergtficyWalerSupply Plrnptouse

16 E/reigencyft>«*SLipf»y Generators

17 Tempe«ngW«i«9yiiennf\fTphouso

10 trnaQetKyWSaSj^Vete9aton(<rmfxx\rA)f OiV

19 Un«emagnncyCor«dCenka(creperu«)

20 0emnaatedW.it«3aageTw»V$(aicpair»i)

MOT I* anstonT w (one pa u»i)

?? SysicxnScfviceTianslo'meffoiieperurvi)

23 GcraQ)orScfviccIiarttlorn«i(onapcrLnl)

29

30 OJFal

31 Emergency Pbwr Gtneralor 01 links

32 SmalCralnMngDocli

33 Compu tti Dock

34 Warehousa

35

363/3939 WestAnraxBuUrig40 DySlorageModtJarard41 laslKna&iiy42 HdiariigB Fuel BayDSC Iitnsla Rome _ ^

nudorOukingsM3 POMrhauwdlMsl-445 SavloeWng6 VacumBliUng

10 SlandbyGenereuOlTanks

11 O K i u Management IXAtny

12 Water lieaknemOuidng

13 SeKfcigOasn

14 AuxtoyhadialedruelOay

aaaaacaoo®moaooooooDOOOOOODOO

so 100

FIGURE 1Site Layout

Vent Port

Shield Plate LidLocating

Pin

Lilt Plate

Fuel Bundles(Total 384)

Drain Port

Data

Seal/StructuralWeld

Height 3.55 mFront 2.12 mSide 2.42 mWeight (Empty) 53 MgPayload 10 Mg

Fuel Modules(Total 4)

SaleguardsTube

ReinforcedHigh Density

Concrete

Steel InnerLiner

- Steel OuterLiner

Mechanical Seal

Cobra Seal

CobraSeal

Safeguards SealsConfiguration

FIGURE 2Dry Storage Container for Used Fuel Storage

UFDSF DSC DNSITE TRANSPDRTER

7% GRADE

4 00'

18.2 38.4 78.8

7 PERCENT GRADE DPERATIDNMAXIMUM

WHEEL ttl&<2 DRIVE/STEERWHEEL tt3&4 STEERWHEEL tt5&6 FIXEDWHEEL #7a<8 STEER

FIGURE 3On-Site Transport Vehicle

UFDSF DSC LID TD BODY REMDTE WELDING

u>

QwcMngPoMr Supply (2) Q CnrtagtTnckmv«reN«kr(2) ( f l OsdMorQr-wfe)ft Equipment Boom (2) tjj) Madia** Satin Traddng@ Cooing Unit (4) Q RamotoAitCniMn@W«tdngQun<i) 60 TradingkopwIonCwMnir«)TravtiC*ni*g* @ Prahttflng Ekm«nl»

LID REMDTE VELDILLUSTRATION

LID PLATE

BODY FLANGEPLATE

MECHANICAL SEAM TRACKINGREFERENCE SURFACE

(FUME OR PLASMA ARC CUT)

FLAME Oft PLASMA-ARCCUT SURFACE

DIMENSIONS mmSCALE t : l

DSC LID-TO-BDDY WELD JOINTFIGURE 4

DSC Lid to Body Remote Welding

UFDSF DSC HELIUM LEAK DETECTIDN (BELL JAR)

A) He LEAK DETECTOR

B) VAC. PUMP (He DETECTIDN VAC.)

C) VAC. PUMP (He DETECTIDN VAC.)

D) VAC. PUMP (MAIN BELL JAR VAC.)

E) VAC. PUMP (MAIN BELL JAR VAC.)

F) VAC. PUMP (CAL.LEAK VAC.)

G) VACUUM GAUGE

BELL JAR (UPPER)

TD ACTIVEVENTILATION

t/llllll //AWNWWX.//' III11 III AN WWW

CAL. REF.LEAK

BELL JAR(LOVER)

Hx3-VENT VALVE

FIGURE 5DSC Helium Leak Detection


BRUCE USED FUEL DRY STORAGE PROJECTEVOLUTION FROM PICKERING TO BRUCE

by

R.E. YoungOntario Hydro

Bruce Nuclear Power DevelopmentNuclear Waste & Environment Services Division

Tiverton, Ontario NOG 2T0

CA9700737

Problem Statement

By 1999, the Bruce Generating Station A willhave filled all of its current wet pool used fuelstorage and by 2003, the Bruce GeneratingStation B will have filled all of its current wet poolused fuel storage. Additional used fuel storagespace is required to allow these two stations tooperate beyond these two dates.

The recommended option was to utilize a drystorage container (BDSC 600) similar to thatused at the Pickering Nuclear GeneratingStation. The changes made to the PNGScontainer included:

• fuel stored in trays• container capacity increased to 600 fuel

bundles• changed the container lid to a metal lid• changed the single concrete lid to a double

metal lid system• made the container non transportable• the container would be dry loaded

Project Description

The BDSC 600 dry storage system is anevolutionary development of Pickering DryStorage Container (DSC) system design but isnot licensable for transportation off-site undercurrent regulations. The Bruce container is a dryloaded concrete cask similar in size to thePickering DSC. However, it is much simpler toload, seal and test. The primary benefit of thisalternative is that the fuel remains in trays for theduration of the storage period. Handlingefficiencies accrue with this alternative becausetrays are loaded directly into the container. Adry load method has been developed for theBruce Site, thus eliminating the need for fuel bayfloor reinforcement, impact pads and crane

capacity upgrading, with an associated reductionin costs.

The BDSC 600 will stand about 13 ft. tall andhave a foot print of 8 ft. x 7 ft., weighing 76 tonsempty and 93 tons fully loaded. The walls are 21inch thick concrete with a 14 inch steel inner linerand outer shell. The cavity is of rectangularshape with similar dimensions to the quadrantsof the stacking frames in the SIFB. All exteriorsurfaces are painted with 3 layers of high solidsepoxy paint to reduce the possibility ofcontamination during loading.

The double lid arrangement of the BDSC is alsodifferent from the Pickering single lid "top hat"design. The lids are an all-steel plug type designand the closure welds will be situated on top ofthe container in a circular pattern. The primarylid consists of a 9" thick circular steel plate roughmachined around the edge and fine machinednear the weld joint at the top edge. A small ventis provided in the lid to aid vacuum drying. Theprimary lid is welded to the container body with aVi inch bevel weld reinforced with a % inch filletweld. The 1" thick steel secondary lid is placedon the container body after the primary lid hasbeen welded in place. This secondary lid is alsocircular and is welded to the body with a % inchfillet weld. The second lid facilitates leak testingof the primary containment weld and providesredundant containment. This lid has a singlepenetration to aid leak testing. The container isbackfilled with inert helium cover gas after thefuel is loaded.

This design offers many advantages over thePickering DSC system design. The constructionof the container is simpler and this fact isreflected in a 28% reduction in the life cycle costover a Pickering repeat. The circular lids areeasy to machine simplifying fit-up with noconcrete pouring, thereby reducing welding,

175


testing, and machining cost as well aseliminating rebar and concrete itself. Using steelalso reduces cask height and hence a slightweight reduction. The circular lid also permitslifting provisions to be placed on top of thecontainer, reducing the lifting bean span andweight, as well as simplifying the connection intothe container body. Another major advantage ofthis lid design is the simplification of the closurewelding machine. The machine will simply rotatearound a centre point laying in weld materialfrom above the joint in two or three passes. Theweld equipment will most likely utilize a MIGprocess similar to the Pickering system,however, the machine will only put down two orthree passes, instead of the 11 passes requiredat Pickering. No preheat will be required and thetotal weld time (including set-up, inspection andfinal cleaning) could be less than one shift,compared to the Pickering 72 hour system.

176

^ ^^

177

00

PICKERING-;• • " ;^: r: ! "1 384 bundlesm 140" highm 62 tons\ Ttaltspoftable off site

m ModulesSlab type squate lid

1 Wet fttel tad• 11 pass pet teeter V~gf oove

(11/2")• ASMENC ; ...t Itidoor stdMge

600 bundlesISg.S^high93 tonsNoti-tfansportabte off siteTmysCitculat plug lidt)ty fuel load4 pass to top circtilaf V^gtoove

Z, ASMEME "•H Outdoot stdirage

BRUCE OS(RRAO8ATFD FI9F6 STORAGE TRAV

Bruce Dry Storage Container - 600BDSC 600

Ten Fort

Vent (not visible)

Secondary Lid

PrimaryContainment l id

Fuel in tray*

25 trays

Fainted SteelOuter Shell

Reinforced ConcreteShielding Wall

Sted Inner liner

180

TOOLS MAINTENANCE WINPOWCHOICE I 1 A ff 2

BRUCE DRY LOflO EpUiPfiENT

WATER LEVEL 629' - 6"

00

6 0 2 ' - 0

3ij^S^^ce^a8^W**fcfco*tfwfc3JSB*ScSiM*fcjS««So«*M«&iB63*:

Independent If £$^^ t roh systemsTested before it is brought into theIrradiated Fuel BayMo eomponerit is | f e||er than the 35 toncapacity of the existing crane

! Local control centre will beindependent of other station systems

BRUCE SECONDRRV IRRRDIHTEDFUEL BflV


PND FUEL HANDLING DECONTAMINATION-FACILITIES AND TECHNIQUES

byR.Y. Pan

Ontario HydroNuclear Technology Services

CA9700738

The use of various decontamination techniques and equipment has become a critical part ofFuel Handling maintenance work at Ontario Hydro's Pickering Nuclear Division. This paperpresents an overview of the set up and techniques used for decontamination in the PNDFuel Handling Maintenance Facility and the effectiveness of each.

1.0 INTRODUCTION

The current PND Fuel HandlingMaintenance Facility (FM Shop) wasconstructed in the late 1970's as part ofthe Pickering 'B' expansion that increasedthe size of the Pickering station from fourto eight units. It is located in the Unit 0operating island just west of Unit 5.

Despite various other improvements, FMShop personnel still experienced the sameproblem with inadequate decontaminationequipment that existed in the originalPickering A Fuel Handling MaintenanceShop. Decontamination was still limited tomanual scrubbing and dry wiping forsmaller items like closure plugs, andmanual scraping and vacuuming for largeritems like FM rams and head internals.

A renewed effort to upgrade Fuel Handlingdecontamination capabilities was initiatedin December 1987, after excessive levelsof Co-60 were identified in the stationlaundry and traced back to protectiveclothing from the FM Shop. This effortfocused on two key result areas: (1)eliminating the existing contaminationhazard, and (2) developing more effectivedecontamination capability.

The FM Shop was thoroughlydecontaminated over a four month periodin order to eliminate the loosecontamination hazard.

An assessment of the facility wasconducted with the assistance ofpersonnel from what is now Ontario HydroNuclear Technology Services (NTS) in

order to facilitate development of betterdecontamination equipment.

1.1 Performance Criteria

A set of performance criteria wasdeveloped from this assessment and usedas a general guideline for improving FHdecontamination capability. They are asfollows:

(a) The cleaning activities carried outin the decontamination centres mustreduce loose activity to less than1,000 cpm. Fixed activity to bedealt with by disposal.

(b) The decontamination equipmentshould be selected to minimize theamount of "hands on" work in orderto minimize personnel radiation doseduring decontamination.

(c) Decontamination equipment musthave sufficient capacity to meet theanticipated work load of the FMshop in a timely manner.

(d) The layout for the FHDecontamination Facilitymust satisfy the following.

(i) A managed flow of personnel andequipment through thedecontamination centre. This is bestachieved by one-way materials flow.

(ii) Contaminated and cleanequipment must be kept separate toprevent cross contamination.

185


(iii) Ease of inventory control, sothat parts brought in fordecontamination are easily retrievedand processed.

(iv) Sufficient space to work andmanoeuvre materials along thetransfer flow paths.

(v) Sufficient storage space forcontaminated equipment awaitingdecontamination. The location ofcontaminated equipment storagemust not pose a radiation hazard toany staff working in the shop orstation.

machine components and tooling. Thedegree of contamination removal dependson the aggressiveness of the mechanicalmethod selected as well as the applicationtechnique. An aggressive manualapplication of a mechanical cleaningtechnique can be highly effective atremoving fixed contaminants. However, itcan also remove the upper layers from thesurface of the contaminated object whichis usually not desirable unless the part is tobe scrapped. Less aggressive mechanicalmethods may not result in material loss,but may also not be effective at removingmore adherent deposits.

(e) Radiation monitoring of equipment andpersonnel entering and leaving thedecontamination centre is required.Appropriate monitors and meters must bein place and operational.

(f) Procedures for facility operation,housekeeping, equipment use andmaintenance, waste handling, andradiation control must be available.

(g) Proper management of the Shop isessential to ensure on-going successfuldecontamination facility operation.Periodic assessments of the facilityoperation should be carried out to assist inidentifying deficiencies and requiredcorrective action.

2.0 DECONTAMINATION TECHNIQUES

Many techniques can be used to cleancontaminated items. Some of the methodslooked at for use in the FM Shop aredescribed briefly below.

2-1 Manual /Mechanical

Manual/mechanical techniques rely onphysical contact between the surface of acleaning tool, such as a metal wire brushor scouring pad, and the surface to becleaned.

The process can be quite effective atremoving loose and some fixedcontamination from the surface of fuelling

2.2 Chemical

Chemical cleaning relies on the use of asolvent solution to dissolve the radioactivedeposit or contaminant. Chemical cleaningprocesses are usually classified as eitherdilute or concentrated. In general, dilutechemical cleaning refers to processes usingsolutions with concentrations < 1 %(w/w); whereas concentrated chemicalcleaning typically refers to processes thatuse solutions with concentrations > 1 %(w/w).

Chemical cleaning effectiveness isdependent on a number of factors:concentration and composition of thecleaning solution, application method(temperature, pH, flow rates, contacttime), and composition of the deposit orcontaminant are key among these.

2.3 High Pressure Water

High pressure water jets with dischargepressures ranging from 1 to 15 kpsi havebeen used to clean a variety ofcontaminated equipment surfaces, andtools.

The decontamination effectiveness of thehigh pressure water stream depends onthe shape, flow rate, and dischargepressure of the jet and stand off distance.

186


Water jets employed at the lower end ofthe pressure range ("2 to 3 kpsi) areeffective at removing some loosecontamination and will generally notdamage protective coatings or paintedsurfaces beneath the deposit. A typicalsystem operating in this pressure rangewill use a "fan" type spray nozzle (30° or45° angle of dispersal on the spray), havedischarge water flow rates of 3 to 6USgpm, and be manually operated.

Water jets employed at the higher end ofthe pressure range ("9 to 15 kpsi) areeffective at removing all accessible loosecontamination and some fixedcontamination. (For the purposes of thisdiscussion, "fixed" refers to anycontaminant that cannot be removed bymanually wiping with a soft cloth.)Damage to underlying protective coatingsand painted surfaces can result if care isnot taken when decontaminating withwater jets in this pressure range. Typically,a system operating in this pressure rangewill employ fan or pencil type spraynozzles (jet can be angled or straight),have discharge water flow rates of 6 to 15USgpm, and be either automated ormanually operated.

2.4 Vibratory Finishing

Vibratory finishing uses small metal orceramic cleaning media immersed in acleaning solution to decon small metalliccomponents and hand tools. Duringcleaning using this method, contaminatedcomponent, cleaning media and solutionare all held in a large tub with a system tocontinuously filter and recirculate thecleaning solution. The tub vibrates on itsspring mounted base, creating physicalcontact between the cleaning media andcomponent in the tub. A thoroughinspection should also be performed aftercleaning in the VFM to ensure that none ofthe ball cones have become lodged withinthe part cleaned.

The cleaning action obtained using thismethod is a combination of the mechanicaleffect created by contact between themetal cleaning media and component/

contaminant surface, and the chemicalaction of the cleaning solution. Eachprocess enhances the action of the other.

Vibratory finishing is effective at removingloose and some fixed contamination fromthe surfaces of small fuel handlingcomponents and hand tools.

2.5 Ultrasonic Cleaning

Ultrasonic cleaning combines the chemicalaction of a cleaning solution with thephysical cleaning action of cavitation toremove deposits/ contaminants fromcomponent/ tool surfaces. The basicsystem consists of a tank (which containsthe cleaning solution), a heater, and anultrasonic generator and transducerassembly. A recirculating clean-up systemwith in-line filter is also sometimesemployed.

The effectiveness of ultrasonic cleaning onany given deposit or contaminant isdependent on the cleaning solution used,and method of application. In general,ultrasonic cleaning enhances theeffectiveness of chemical cleaningsolutions by producing a physical effect(cavitation) on or near the contaminatedsurface. This is due to the multitude ofbubbles that are produced in the cleaningsolution as a result of vibrations generatedby the transducer assembly. These bubblesare unstable, localized low vapour pressureregions and collapse quickly. The cavitated"collapsed bubble" region produced is atemporary, localized pocket of hightemperature and pressure that can enhancethe chemical cleaning properties of theliquid solution by up to 10 times,according to results reported by EPRI.

(f) Strippable Coatings

Strippable coatings are used for cleaning/contamination control in two ways: [1] asa protective coating (to preventcontamination of the material underneath)which can be easily removed when loadedwith deposits, or [2] as decon technique toremove contaminants already present on a

187


surface. Typically, a strippable coating willbe applied as one thick layer (0.020-0.080in.) directly to the clean or contaminatedequipment surface. When applied as aprotectant, the strippable coating may beleft in place for up to a few days beforeremoval. When applied to decon a surface,the strippable coating is removed as soonas it has dried or cured sufficiently toobtain maximum strength.

The effectiveness of strippable coatings isquite good in general. Decontaminationfactors as high as 1000 have beenreported by EPRI using this method.

(g) Dry Ice Blasting

Dry (CO2) ice blasting is a decontaminationmethod that relies on physical and thermaleffects to remove deposits fromcontaminated surfaces. Systems on themarket today typically use compressedpellets of C02 "snow" in the cleaningprocess. The pellets are entrained in an airstream that is directed via a jetting tool tothe contaminated surface to be cleaned.The cleaning mechanism is dependent onseveral factors: [1] the impact energy ofthe pellet, which is a product of thepellet's mass/density and velocity; [2]pellet mass flow rate/ number of impactsper unit time; and, [3] thermodynamiceffects.

Factors [1] and [2] are controllable bymaking adjustments to the blastingmachine. Increasing any one of thesefactors (pellet density, velocity, or pelletflow rate) will increase the effectiveness/aggressiveness of the decontamination.

Factor [3], thermodynamic effects, iscaused by the temperature differentialbetween the dry ice pellets and thecontaminated surface being impacted. Aneffect known as "tracking" (freezecracking), in which the contaminant layerbecomes embrittled by the cold CO2 is onesuch effect. Fracking results in a break upof the active deposit layer, making it easierto dislodge.

Another thermodynamic effect occurswhen the CO2 pellet sublimates on impact.The CO2 gas from the sublimating pellet isable to penetrate beneath the deposit layer(which has been made accessible by"fracking") and dislodge contaminantsfrom underneath as it expands.

Dry ice blasting can be effective atremoving most loose contaminants, oil andgrease, and some paints and adhesives. Itis not effective at removing oxides,scales/mineral or other hard deposits.

3.0 DECONTAMINATION EQUIPMENTSELECTION

Field testing indicated that high pressurewater jetting was effective at removingmost of the contamination normallyencountered in the course of Fuel Handlingmaintenance. This contamination consistsmainly of activated corrosion and somefission products picked up from theprimary heat transport system duringfuelling.

In addition to being effective, highpressure jetting also offered theadvantages of being non-corrosive, andbeing relatively inexpensive to start up andoperate. Waste disposal was also easysince no chemicals, abrasives, or otheradditives are ordinarily needed in order toachieve good results.

Decontamination using manually operated10 kpsi jetting equipment commenced in1989. Very careful control of the waterjetting operation was maintained duringthis time in order to ensure worker safetyas well as to obtain good decontaminationfactors.

The FM ram was the most frequentlydecontaminated item using this methodover the next 18 months. After severalrepetitive manual cleans of the ram usingFM Shop personnel in plastic suits, itbecame apparent that this physically tiring,somewhat tedious task would be ideallysuited to automation. This observationeventually led to the development of the

188


automated Ram Spray Booth (RSB), aventilated, totally enclosed automatedwater spray booth.

4.0 Ram Spray Booth

The RSB was custom designed andconstructed to decontaminate PND fuellingmachine rams, and can accommodateother long cylindrical items with diameters< 17".

Some key design features incorporatedinto the RSB are as follows:

• accommodates an FM ram assemblywith the drive housing gray lockattached

• use of overhead crane to load /unloadindividual components

• high accessibility to the ram wheninside the booth

• all high pressure spray, drain, and filtersystems located within the booth

• built to facilitate decon of interiorsurfaces and minimize potential forcrud traps

• able to use robotic arm to perform themajority of the ram housing and ballscrew decon with rotary water jets

• capability to manually water jet areasnot accessible by the robotic spray armor as backup

• built in safety devices to limit accessto the booth interior when roboticspray arm in operation

• PLC access codes to restrict use toauthorized personnel

• connectable to active ventilation

The height of the booth (92" including footpads), required construction of a raisedplatform around the booth to provideoperators with easy access to controls andglove ports.

The Automated Water Lancing Machine(AWLM) uses a two nozzle rotary sprayjetting tool (~9kpsi) for cleaning. Separateelectric motors inside the booth move thespray arm longitudinally and radially alongseparate tracks to provide spray coverageto most areas of the ram. An Allen Bradley

PLC directs both motors to control thespraying pattern when in use.

A separate, smaller 3 kpsi water jettingpump is also built into the RSB to permitmanual jetting of any areas missed. Thehand-held spray lances are operatedthrough glove ports down the length ofboth sides of the spray booth.

The Ram Spray Booth was placed in-service in mid-1994 at a total cost ofapproximately $300,000.

Decontamination results achieved usingthe RSB have been very good. Rams withsmearable activity > 1 million cpm havebeen reduced to virtually zero loose.Further, time required for a full ramdecontamination has been lowered from 7-10 days using manual spraying to about 3days using the RSB.

4.1 Ram Spray Booth Modifications

A number of modifications were carriedout during the RSB commissioning processto address deficiencies and enhanceperformance.

4.1.1 L-225Installation

High Pressure Pump

The existing FM Shop high pressure jettingpump was modified to allow it to beinstalled beneath the Ram Spray Booth foruse as part of the Automated WaterLancing Machine, resulting in savings onequipment costs when purchasing theRSB.

4.1.2 Light Curtain

Difficulties were encountered setting upthe electric light curtain. The light curtainis a device consisting of an emitter, areceiver, and two mirrors. As designed,this device was supposed to stop theAWLM in the event that one of theoperator's arms was extended into thebooth while in operation. Although the unitdid eventually function as designed, itproved to be exceedingly difficult and timeconsuming to maintain the proper set up.

189


This was due to the fact that the lightcurtain model selected by the manufactureris at the limit of its range when used as aperimeter detector for the RSB. Slightjarring of the mirrors from personnelwalking by was enough to set them out ofalignment. The corrective measureeventually used to resolve this problemconsisted of disconnecting the light curtainsystem, and using shield doors over theglove ports to prevent access while theAWLM is in operation. Safety interlocks,which trip the Emergency Stopinstantaneously when activated, are inplace on all RSB doors and hatches toprevent all other access routes to thebooth interior while the AWLM is inoperation.

4.1.3 Automated Water Lancing MachineProgramming

Customization of the PLC software for theAWLM was carried out in the FM Shopbetween April and July 1994. The numberof passes and the spray pattern of therobotic arm were modified in order tomaximize coverage during FM ramdecontamination. The spray pattern wasmodified so that washing occurred from

the top down (i.e. the first pass at 90° tohorizontal, then at increments on eitherside of top centre down to horizontal) asopposed to the original pattern which had

sprayed from 0 ° to 180 (i.e. fromhorizontal to the top then back tohorizontal).

4.1.4 Tooling and Accessories

Ram trolley cart modifications andnumerous other mechanical changes weremade to the Ram Spray Booth duringcommissioning to help achieve betterperformance. These changes includedthings such as:

fabrication /installation of spray cradlesand supportsfabrication /installation of ballscrew /tube racksinstallation of safety supports (for sidedoors) in the RSB

repairs to defective air valves (for sidedoors)repairs to drip trays (for side doors)addition of glove bar/ hangerrepairs to the cable trayaddition of an access port (2" dia.) tothe north end loading hatch to enablewater jetting of ram housing and tubeinternals without removing them fromthe RSB

5.0 Other Decontamination Equipment

Other equipment/techniques that hadproven to be effective -such as ultrasoniccleaning and vibratory finishing-- were alsointroduced into the FM Shop at around thesame time as water jetting. Figures 1 and2 show the layout of the renovatedfacility.

6.0 Decontamination EquipmentEffectiveness

Effectiveness of the equipmentinstalled was assessed as part of thecommissioning process. A summaryis presented in Table 1 below.Regular checks since that time haveshown that the equipment is stillmeeting performance targets (i.e.able to decontaminate componentsto < 1000 cpm loose).

7.0 CONCLUSION

Improvements in productivity andequipment reliability have been achieved asa result of upgrading Fuel Handlingdecontamination capability.

Ram performance, for example, hascontinued to improve as decon capabilityhas increased. This is at least partlyattributable to better rebuilds duringmaintenance as a result of [1] being ableto fully see the part, and [2] being able tospend more "quality" time working on thepart.

There are also other indirect benefits ofdecontamination. The maintenance side ofthe FM Shop, for example, is no longer arubber area, which makes for more

190


comfortable working conditions. As well, since the commencement of the program,dose pickup has dropped by a factor of 5 an important health and safety gain.

Table 1: Decontamination Equipment Effectiveness

DESCRIPTION

Ram Spray BoothVibratory Finishing MachineGlove Box Spray BoothVentilated Sinks/BenchesAutomated Parts WasherUltrasonic CleanerVentilated Solvent Tank

USED FOR

rams, other cylindrical partsclosure plugs, small partssmall parts (< 18 kg)small parts, hand toolssmall parts, hand toolssmall parts, hand toolsoil, grease removal

EFFECTIVENESS* (cpm)> 1,000,000 to < 10,000

> 25,000 to<2,500> 100,000 to < 2,500

> 10,000 to < 2,500> 55,000 to < 30 ,000"> 25,000 to < 2,500

fully degreased

Maximum smearable activity levels detected before and after decontamination. Total DFs for each pieceof equipment will vary depending on the nature of the contamination on the component being cleaned.Most components can usually be decontaminated to zero loose activity levels.

##Measurements after one wash cycle. Most components can usually be decontaminated to < 1,000

cpm with repeated washing (two to three additional cycles).

FIGURE 1: SMALL PARTS DECON ROOM LAYOUTVENTILATED

BENCH

• • •VIBRATORY.FINISHER

OVERHEADDOOR

VENTILATEDSOLVENTTANK. VSTI

DOUBLE DOORS

CLOVE BOX V E ' sTNfesT E D

SPRAY BOOTH VENTILATED i VENTILATEDCB1 BENCH I BENCH

VENTILATEDBENCH

\ \SMALL WINDOWSPARTS (I OF41

WASHER < l u f 4 >

fSTORAGE CABINET

WITH CONTAMINATIONMETER

SMALL PARTS OCCON ROOM

f 9DP

UMMAVfeDOlM

EAST FU SHOP

NOTTOSCAUE

FIGURE 2: EAST FM SHOP EQUIPMENT LAYOUT

191

FiRST INTERNATIONAL CONFERENCE ON CANDU FUEL HANDLING SYSTEMS, MAY 1996

EIQURE3;RAM SPRAY BOOTH DETAILS

AWLMPLC*ELECTRICAL

'CONTROLS

VERTICALSIDE DOOR

(lOFfl

SOUTH LOADINGHATCH

WINDOW(1OFO

TOP LOADINGROOF DOOR

(FULL LENGTH OF BOOTH)

CONTROLSFOR3KPSIJETTINGSYSTEM

3KPSI PtMPV(FOR HAND

LANCING)SIDE DOOR

RELEASE VALVE

9 KPSI PUMPFOR AWLM

GLOVE PORTS

FIGURE 4:

AUTOMATED WATERLANCING MACHINE

DETAILS

{END VIEW)

ROTARY SPRAY HEADTIMING BELT

GLOVE PORT

FMRAM(17" MAX. DIAMETER)

AWLM TROLLEYSPRAY HEADRAIL SYSTEM

RAM TROLLEY

192


WINDOW GLOVE PORTSI

VENTILATION INLET WIPER ARM ' FRONT HATCH (OPEN) SIDE HATCH (1/2)DAMPER

SAFETY DOOR INTERLOCK (1/3)

FIGURE 5:GLOVE BOXSPRAY BOOTH

3.000 PSI JETTING PUMPDETERGENT DISPENSER"""DRAIN PUMP

DRAIN CONNECTION

•ACKET

TRANSDUCE* MEATS)

SOH(ELECTRIC

POWER ULTRASONIC HWH FREQUENCY

FIGURE 6:ULTRASONIC CLEANERDETAILS

INJECTSOECONTAMMATBI

FICURE 7: VENTILATED SOLVENT TANK DETAILS

UCBT T2££2i£«

MNSEHOStAND MWSII

STAKCtR AJUH

TKO500WHtATUt

ELIM1VTS

193


FIGURE S:VIBRATORV FINISHER DETAILS

HGURE 9-A:VENTILATED BENCH DETAILS

SLUOGC

SETTUNQTANK

.MIMDOII iO AXCLE T T F .

FIGURE 9-B:VENTTLATED SINK DETAILS

SPtAT ItlNSCI

»OT TO < f ^ l f

1

r TO sc^n SJEXJilQ

NOTE: All other deuilt a n idenlkal to 2m TentiUlcd bench units.

KPT •*> v i u

194

1ST CNS INTERNATIONAL CONFERENCE ON CANDU FUEL HANDLING SYSTEMS

CA9700739

DARLINGTON NGD FUEL HANDLING

HEAD EIGHT ACCEPTANCE PROGRAM

P.H. SKELTON (DNGD)T. SIE (DNGD)

J. PILGRIM (GE-CANADA)

ABSTRACT

Darlington NGD requires eight fuelling machineheads to fuel the 4*932MW reactors. Sixheads are used on the three fuelling machinetrolleys for normal fuelling operations. Afurther two heads are required to allow formaintenance and to provide for such reactorface activities as PIPE and CIGAR.

Seven heads were successfully delivered tosite from the head supplier. During acceptancetesting, stalls on the charge tube screwassembly of the eighth and final headprevented its delivery to site. Replacement ofthe charge tube screw with a spare screw didnot alleviate the problem.

An in depth series of tests were undertaken atsite, at the supplier and at the screw sub-supplier to determine the root cause of theproblem.

These tests included taking torquemeasurements under different operatingconditions and using different components toassess the effects of the changes on torquelevels. An assessment of the effects ofchanging chemical conditions (particularly crudlevels) was also made.

To ensure that the results of the testing werewell understood, additional torque testing wasalso completed on a head and screw assemblyat site that was known to work well.

Based on all of the above series of tests, arecommendation was made to re-machine thecharge tube screw(s). The original charge tubescrew from Head eight was subsequentlyreturned to the sub-supplier for re-work.Follow-up torque measurements andacceptance testing showed that the screw re-work was effective and that Head eight couldbe successfully delivered to site.

This paper focuses on the results of thehead/screw test program. Results of theacceptance testing are also discussed.

INTRODUCTION

The Darlington fuel handling system providesautomatic on power fuelling capability to allfour reactors. Under normal operatingconditions, each reactor unit can be servicedby any one of three fuelling machine (FM)systems.

A fuelling machine system consists of one pairof fuelling machine heads and their associatedheavy water and air auxiliary systems allmounted on a transport trolley.

During a typical fuelling operation, one fuellingmachine head inserts four fuel bundles into achannel two at a time. This causes the stringof 13 bundles in the channel to be shifted andthe four end bundles to enter the secondfuelling machine head. Either of the fuelling

195


machine heads can insert new fuel or receiveirradiated fuel as all the heads are identical (seefig D.

Each head consists, in part, of a charge tubeand a concentric ram which transfers fuel intothe reactor under normal operating conditions(310oC and 10MPa). The charge tube has anaxial drive (CTA) and a rotary drive (CTR). Theram has an axial drive only.

Figure 2 depicts the layout of a Darlingtonballnut/screw arrangement. The charge tube ishollow with the ram positioned inside thecharge tube. Both charge tube and ram axialmovements are generated through the ballscrew/nut arrangement. The charge tube axialdrive rotates the charge tube ballnut which isstationary with respect to the fuelling machine.This motion causes the charge tube screw toadvance or retract axially depending on thedirection of the ballnut rotation. All actions arein heavy water. There is no lubrication (otherthan the water).

For a ball screw/nut arrangement to work withminimum torque, it is crucial that the balls donot bunch up. Adjacent balls in a screw cancome into contact with each other by:

- vibrations- bending of the screw due to loading- imperfect screw profiles- poor deflectors or guides

When any of the above occur, higher torquesare required to overcome the bunching up ofthe balls. An effective ballscrew design willhave less potential for ball bunching or if itdoes have some ball bunching, the balls willseparate easily.

DISCUSSION

Six heads are used on the three fuellingmachine trolleys for normal fuelling operations.A further two heads are required to allow formaintenance and to provide for such reactorface activities as PIPE and CIGAR.

Seven heads were successfully delivered tosite from the head supplier. During final

acceptance testing of the eighth head, stallson the charge tube screw assembly preventedits delivery to site.

Prior to the final acceptance testing, a series ofpreliminary tests of the head was completedon a test rig that simulated a Darlington fuelchannel. During this initial testing, frequentcharge tube axial stalls (due to higher thanexpected torques) were noted. Note thataverage torques typically seen on anacceptable charge tube are 5O-8Oinch.lbswhile peak torques of 150-200inch.lbs weretypically seen on Head 8 charge tube.

The fuelling machine was removed from thetest rig, disassembled, inspected, cleaned, re-built (with a spare screw/nut assembly) andreturned to the test rig for a second series oftesting. The original screw/nut assembly wasreturned to the sub-supplier for evaluation.

During the second series of testing, cold runs(260oC) were successful with no stallsobserved. However, when the test rigtemperature was increased to 300oC, frequentstalls began to appear. When the drive torquewas plotted against time, many torque peaks(as high as twice the average torque) werenoted.

Clearly there was a problem with the chargetube assembly.

Although the major efforts were directed atresolving whether the charge screw was theroute cause of the problems, it was also feltthat the root cause could possible lie with headcomponent misalignment. It should be notedhere that Head 8 was at one time overheatedwhen it was used in an unrelated test program.

Ontario Hydro Technology was contracted toperform a series of tests on Head 8.Accelerometer probes were located at severalpoints along the outer housing of the chargetube axial drive (see figure 3(1>). Torque profileswere also generated. A series of thirty twotests were completed. Sixteen of the testswere in advance mode and sixteen were inretract mode. These tests were completedunder several different conditions including:

196

1ST CNS INTERNATIONAL CONFERENCE ON CANDTJ FUEL HANDLING SYSTEMS

- off channel- on channel, cold and de-pressurized- on channel, cold and pressurized- on channel, cold pressurized with two fuelbundles loaded into fuel carrier.- on channel, hot and pressurized- on channel, hot pressurized with two fuelbundles loaded into the fuel carrier

Results of this series of tests121 appeared toindicate that the high torques seen with Head8 were not caused by:

- the screw touching at the front end of thefuelling machine- alignment problems at the rear of the fuellingmachine- the input drive or gear box

The testing did appear to indicate that thehigher than expected torques were caused byeither pitch/thread depth variations from thefront to the back of the screw or pre-loadingon the ballnut. In addition, there was evidenceof double contact with the balls on the screwthread. This double contact indicated thatthere was some imperfection in the screwthread profile.

In later discussions, test rig crud levels wereraised as a possible additional contributingfactor since the rig was predominantly carbonsteel with minimal oxygen control.

The evidence now clearly pointed to aballnut/screw problem and not a misalignmentproblem.

A set of experiments were then conducted tonarrow down the problem further. A series oftests were completed on a ballnut/screwarrangement that was located away from thehead on a test bench. Seven variables werealtered. These variables included:

- changing the screw (two spare screws wereavailable)- changing the nut (two spare nuts wereavailable)- changing the screw bearings

- changing the pre-load on the nut (lOOIbs and6001 bs)- changing the cleanliness of the screw (byadding fine paniculate crud)- changing ballnut transfer tube- changing ballnut deflectors

Results of this testing are shown in figure 4.Two parameters appeared to predominate,screw pre-load and rig cleanliness (crud).

Head 8 was then re-built. The pre-load on thescrew was reduced from the normal level of600lbs down to lOOIbs. Efforts wereundertaken to ensure that the test rig was asclean as possible. The rig was run for severaldays with purification flows maximized, Asmuch crud as possible was removed from therig.

With Head 8 now rebuilt using optimumconditions as determined from the benchtesting (lower ballscrew pre-load and with aclean test rig), final head acceptance testingrecommenced.

Again, higher than expected torques wereobserved. This was somewhat disappointingsince all the evidence from bench test workhad indicated that the problem was wellunderstood and that we had completedeffective fixes (although bench test work wasnot completed at normal temperatures andpressures).

It was now evident that Head 8 was notadequate for normal production use. Furtherwork was required on the ballnut/screwarrangement.

The screw/nut supplier was contacted forfurther input. The original screw from Head 8(which had been previously returned to thesub-supplier) was inspected more closely.Indications of double contact on the screwthread were evident. In addition the screw wasfound to be bent by about 13 thousandths ofan inch. Although this bending was acceptablefor use in normal fuelling machine operation, itwas too far out for a precise grindingoperation.

197


A decision was made to straighten the screwand regrind the screw thread. It wasrecognized that the screw could be badlycracked or possibly broken during thestraightening process. However it was equallyrecognized that we would not be able to usethe screw in the current state. The screw waseventually straightened to within 2thousandths of an inch but could not bestraightened further. A decision was thenmade to machine the screw straight.

Following the straightening of the screw, thethread profile was re-ground. The materialremoved from the thread profile was mostlyremoved in the axial direction (about 5thousandths of an inch). The depth of thethread was only altered in the area where thedepth was shallowest (by about 2 thousandthsof an inch).

The screw (with its nut) was then returned forre-building into Head 8.

Subsequent acceptance testing with Head 8with the re-machined screw was successful.No high torques or stops were observed duringthe acceptance testing. Head 8 was finallydelivered to site during the summer of 1995.

directed at developing torque sensingequipment for use on the fuelling machinesduring normal operation.

ACKNOWLEDGEMENT

The authors wish to thank technician staff atGE (Canada) for all their hard work at the testrig which allowed us to get a betterunderstanding of the problems beingexperienced with Head 8. In particular, theauthors wish to thank Art White, Don Foley,Dan Finnegan and Dan Ayotte for theirdedication and commitment.

The authors would also like to thank Phil Daleof Ontario Hydro Technologies for his valuableinsight which allowed us to better focus in onthe route cause of the problem.

REFERENCES

1. Darlington GS-A Fuel Handling SystemTechnical Document.

2. Head 8 Acceptance Testing Minutes ofMeeting dated 11 August 1994.

CONCLUSION

During final acceptance testing of the eighthhead, stalls on the charge tube screwassembly prevented its delivery to site.

An in depth series of tests undertaken at site,at the supplier and at the screw sub-suppliereventually determined the root cause of theproblem to be a badly machined screw.

The screw was returned to the manufacturefor re-work which included straightening thescrew and re-machining the thread profile.

Subsequent testing of Head 8 using the re-machined screw was successful. Head 8 waseventually delivered to site during the summerof 1995 and is presently available for service.Because the testing program showed theimportance of using torque monitoring as adiagnostic tool, future activities are being

198

1ST CNS INTERNATIONAL CONFERENCE ON CANDU FUEL rtANL>jUXlNL>

CHANNEL SHOOT

closure PIUG

V ~ 1 1 K L _

\JJB5.MAGA2WE

M MSPIINE

\FUEL

BUNDLES

CHARGE TUBIM U SCREW

mijf

FUELCARRIER

CHARGf TUBEM U N U T

/

j

»

DRIVES

RAMM U N U T

RAMBALL SCREW

RAMDRIVE SUEVE

RAMOMVE GUIDE

Figure 1: Schematic of Darlington Fuelling Machine.

CHARGE TUBEBALL SCREW

\

Figure 2: Schematic of a Charge Tube with a Single Ball Nut.

199


Figure 3: Schematic of Darlington Fuelling Machine Charge TubeDrive Mechanism.

20 r — ^

15

10

0 L

i

CLEAN NUT SCREW BALL PRELOAD T.TUBE DEFLECT

LEVEL 1 LEVEL 2

Figure 4: Result of Experiments to Determine Which MajorVariable(s) Affect Ballnut/Screw Performance.

200


FUEL HANDLING SYSTEM OF INDIAN 500 MWe PHWR-EVOLUTION AND INNOVATIONS

A.SANATKUMAR, INDER JIT AND G. MURALIDHAR

Nuclear Power Corporation of India Ltd.,Mumbai-400 085, INDIA

CA9700740

ABSTRACT

India has gained rich experience in design,manufacture, testing, operation and maintenance ofFuel Handling System of CANDU type PHWRs.When design and layout of the first 500 MWePHWR was being evolved, it has been possible forus to introduce many special and innovativefeatures in the Fuel Handling System which arefriendly for O&M personnel. Some of these are:

• Simple, robust and modular mechanisms forease of maintenance.

• Shorter turn-around time for refuelling achannel by introduction of a transit equipmentbetween Fuelling Machine (FM) Head and lightwater equipment.

• Optimised layout to transport spent fuel instraight and short path and also to facilitatedirect wheeling out of FM Head from ReactorBuilding to Service Building.

• Provision to operate the FM Head even whenPrimary Heat Transport (PHT) System is openfor maintenance.

• Control-console engineered for carrying outrefuelling operations in the sitting position.

• Dedicated calibration and maintenance facilityto facilitate quick replacement of FM Head as asingle unit.

The above special features have been described inthis paper.

1. INTRODUCTION

Pressurised Heavy Water Reactors (PHWRs) builtin India have evolved from the CANDU typeDouglas Point reactor. The On-power FuelHandling Systems (FHS) at Rajasthan (RAPS-1&2)were built with Canadian collaboration and areidentical to that at Douglas Point. The design atMadras (MAPS-1&2) is similar to that of RAPS-1&2but manufacture, testing, commissioning andinstallation were carried out in India. All thesubsequent reactors of 220 MWe capacity atNarora (NAPS-1&2), Kakrapar (KAPS-1&2), Kaiga(Kaiga-1&2) and Rajasthan (RAPP-3&4) have FHSwhich are significantly different from the Rajasthanand Madras reactors. These systems were entirelydesigned and built in India. Thus, we have gainedrich experience in design, manufacture, testing,operation and maintenance of FHS of PHWRs.

Now we have embarked on the detailed design ofIndia's first 500 MWe PHWR to be built at Tarapur(TAPP-3&4). For a 500 MWe PHWR, the refuellingrequirement is about two channels per full-power.

day as compared to about one channel per full-power day in case of a 220 MWe PHWR. Theincreased refuelling demand for 500 MWe reactorrequires better availability of FHS. The FHS designshould also strive to achieve shorter turn-aroundtime. Further, there is a need to economise on theman-rem requirements during maintenance.

In order to meet the above requirements, the FHSfor 500 MWe PHWR has been designed to besimple and rugged. Modular construction has beenfollowed for the various sub-assemblies of FuellingMachine Head and other equipment of FuelTransfer System. In addition, a full fledgedCalibration and Maintenance Facility has beenprovided.

201


When design and layout of the plant were beingevolved, it has been possible for us to introduce inthe FHS, many unique features which are friendlyto operation and maintenance personnel, apartfrom being more rugged and reliable.

The above features are described in detail in thispaper.

2. FUELLING MACHINE (FM) HEAD

FM Head is an important equipment of the FHS.The FM Head consists of three major sub-assemblies, viz., the magazine assembly, the snoutassembly and the ram assembly.

2.1 Magazine Assembly

The magazine assembly is made up of a pressurehousing and an end cover, which when joinedtogether by a high pressure coupling, form apressure vessel. A rotary magazine is mountedinside the pressure housing to store new or spentfuel bundles and various other accessories requiredfor carrying out the on-power refuelling operations.

The magazine rotor is made of twelve machinedtubes positioned around a central shaft and held inplace at each end by two jig bored plates. This rotorassembly is mounted over a central shaft which isbolted to pressure housing at one end andsupported in the end cover at the front end. Thusthe magazine rotor is a simply supported assembly.The internal gear attached to the rotor is driven bytwo pinion gears mounted on water lubricatedbearings in the rear end of the pressure housing.Each pinion gear, in turn, is driven by an indexingdrive gear box through a rotary balancedmechanical shaft seal cartridge. The backlash freeindexing drive gear box is driven by an oil hydraulicmotor. (Fig 1)

Some special features of this assembly are:

• In the rotor design, welding is completelyavoided. Distortion control during welding andsubsequent heat treatment, as also theprecision machining of the fabricated structure,has been a demanding and difficult operation inearlier design. Thus the 'bolted design' makes itmanufacturer-friendly.

• The simply supported rotor design is more rigidand avoids the alignment errors that are

possible in case of cantilever mounting used inthe earlier design.

• The magazine drive shaft seal has been madeas a modular cartridge' type unit. Thus, a fullyassembled and tested replacement cartridgecan be fitted in place of the existing one withoutdis-assembling too many components.

2.2 Snout Assembly

The snout assembly is attached to the front end ofthe magazine assembly. It consists of a centresupport which carries a bellow like metallic seal foreffecting a high pressure joint with the end fitting. Amechanism consisting of four cam actuated jawsegments, a clamping barrel with internaltrapezoidal (acme) threads, and a worm gear withmatching external trapezoidal threads, are used formoving and clamping on to an end fitting. (Fig.2)


• An oil hydraulic motor is used to drive themechanism instead of the piston actuated rackand pinion drive used earlier. This enables useof single start acme threads in clamping barrel,giving the required irreversibility to avoidunclamping on loss of oil pressure.

• The snout assembly is modular in construction.This allows removal of the snout assemblywithout the need for the disassembly of thehigh pressure coupling between the end coverand the pressure housing (also referred to,some times, as magazine housing).

2.3 Ram Assembly

The ram assembly is mounted at the back end ofthe magazine assembly. The ram assembly isrequired for operating the different accessoriesduring fuelling operations and to push new fuel intoor receive spent fuel from the channel. In this B-Ram and C-Ram are powered by two separate (butco-axial) rack and pinion drives. Each of thesedrives consists of a dual rack system driven by twopinions which in turn are driven by two high torquelow speed oil hydraulic motors, thus providing abalanced, duplicated drive. Latch ram is alsooperated by oil hydraulic motor. In addition, asystem for emergency movement of B-Ram and C-Ram separately, by using heavy water pressure inthe ram housing, has been incorporated. This

202


would be particularly useful in boxing up thechannel in abnormal situations (Fig 3).


• The ram assembly is of modular construction,allowing its removal as a single unit from thepressure housing.

• The rack and pinion drive selected is of arobust design and expected to perform bettercompared to ball screw drive used in the earlierdesign.

• C-Ram position monitoring scheme is fullyoutside water environment and hence would bemore reliable than the wire rope design usedearlier.

3. FUEL TRANSFER (FT) SYSTEM

The FT system requires to perform the basicfunctions of storing the new fuel in reactor building(RB), exchange of new fuel with spent fuel andtransport of spent fuel using the well-proven'Shuttle Transport System' to Spent Fuel StorageBay (SFSB), located outside the reactorcontainment. It comprises of two sets ofmechanical equipment, one for the north and theother for the south side of the reactor.

The time taken for the FM Head to discharge spentfuel received from the core, and to accept new fuelfor refuelling the next channel, has been reducedby having a common Fuel Transfer Port andincorporating an 'Exchange Programme'. In theexchange programme, while two spent fuel bundlesare discharged into the Transfer Magazine (TM),two new fuel bundles are received from the TM intothe FM Head. The spent fuel in the TM can be sentto the SFSB simultaneously with the refuelling ofnext channel. Thus the overall cycle time forrefuelling a channel has been reduced to about twoand half hours.

All the equipment in the spent fuel path are closedvessels and do not freely communicate with theatmosphere outside the equipment. Thus, spread ofactivity (mostly from the occasional failed fuelbundles) into the FT Room and other areas in thereactor building is considerably reduced.

TM and the other important equipment of FTsystem namely Shuttle Transfer Station (STS) aredescribed below;

3.1 Transfer Magazine (TM)

TM (Fig 4) interacts with three different equipmentnamely New Fuel Magazine, FM Head and ShuttleTransfer Station for the receipt of new fuel bundles,exchange of new fuel bundles with spent fuelbundles and discharge of spent fuel bundlesrespectively.

It consists of a rectangular stainless steel housingin which a carriage moves linearly up and down.The carriage comprises of six fuel tubes which canbe aligned to three ports of the housing to receive /discharge the fuel bundles. The carriage is drivenby an oil hydraulic motor through a rotating ball nutand translating ball screw. The position of thecarriage is controlled / monitored by duplicate setof potentiometer assemblies. The aligned positionof carriage is also confirmed by an independentreed switch assembly.

Some special features of TM are :

• Capability to receive 8 spent fuel bundles fromFM Head and load 8 new fuel bundles into theFM Head in one sequence.

• Simple fuel stops for preventing accidentaldropping of spent fuel into the housing.

• Drive system and position sensing devices arebrought outside the TM Housing to enable easymaintenance. Shielding provided around TMHousing allowing maintenance even when theTM contains spent fuel.

• Provision for cooling of spent fuel bundlesstored in it under various conditions.

3.2 Shuttle Transfer Station (STS)

STS (Fig 5) comprises of a shuttle carriage whichmoves vertically up and down in a cylindricalhousing partly filled with light water. It interacts withTM under dry environment to receive the spent fuelinto a special carrier tube called Shuttle' housed inthe carriage. The Shuttle is then lowered andtransported to Spent Fuel Receiving Bay throughShuttle Transport Tube (STT) using hydraulic force.The STT is kept straight and short for smoothmovement of fuel.

203


The aligned position of the carriage is achievedagainst mechanical stops. The position ismonitored continuously by a duplicate set ofpotentiometer assemblies.

Some special features of this assembly are :

• Simple wire rope drive mechanism powered byoil hydraulic motor.

• Fuel shutter for preventing accidental droppingof fuel into the housing.

• Special dampener attached to shuttle carriageto reduce impact force on bundle due topostulated free fall of carriage.

4. ISOLATION OF FMSYSTEM FROM PHT

HEAD D2O

As heavy water storage tank is common to FMHead and PHT System, FM Supply pumps cannotbe run to meet FHS needs when the PHT systemhas ben opened for maintenance. This puts alimitation on the preventive maintenance andshutdown maintenance jobs to be taken up on theFHS as heavy water supply would not be available.Therefore, an additional 5 tonnes capacity storagetank together with associated circuits have beenprovided so that FHS can be operated in an'isolated mode'.

5. FUEL HANDLING CONTROLS

The controls for the FHS have been based onProgrammable Logic Controller (PLC). In addition,a sitting control console has been provided for FHSwhich would help to reduce the operator fatigue. Adedicated Data Acquisition System (DAS) inconjunction with Expert System would be designedto work as Operator Support System (OSS) and asa tool for maintenance and health monitoring ofcertain critical actuators.

6. CALIBRATION & MAINTENANCEFACILITY (CMF)

In order to meet the enhanced refuellingrequirement for 500 MWe, the on-line maintenanceactivities are required to be minimised. This can beachieved by adopting modular design of the

assemblies so that a faulty assembly can be quicklyreplaced with a well maintained, tested, calibratedand proven reactor-worthy assembly. Thisapproach also avoids any repeat work, whichsometimes becomes necessary due to possiblemaintenance errors in inconvenient locations andradiation fields.

With the above in view, a full fledged off-lineCalibration and Maintenance Facility has beenprovided for the first time, in case of 500 MWePHWR. This facility has the necessary features forcarrying out maintenance, testing and calibration ofimportant components, critical sub-assemblies andcomplete FM Head with simulated reactorconditions outside the reactor building in a tritiumfree environment. This would enable to keep an FMHead in the poised condition ready for replacementin a short time whenever required.

In addition, care has been taken to include thefollowing desirable features also in the CMF.

i) Simulating and analysing, when necessary,the mal-operations in the FM Heads asthey occur in the reactor,

ii) Providing training facilities for operationsand maintenance personnel of the fuelhandling unit.

6.1 Layout of CMF

The FM service areas are located below therespective FM Vaults. The two FM service areas ineach reactor unit are inter-connected with an FMtunnel which runs below the calandria vault. TheFM tunnel permits the movement of FM Head, fromany of the two service areas through a single FMairlock, directly to CMF located in the servicebuilding (Fig 6). The tunnel has adequate concreteand water shielding to allow passage of FM Headeven when the reactor is operating. The CMF islocated in between the two reactor buildings and iscommon for a twin unit station. See Fig 7 for thevarious facilities provided in the CMF, some ofwhich are described below:

6.2 FMHead Test Facility

Three test channels have been provided for testingof FM Heads. The central channel is similar to areactor coolant channel. On this channel, two FMheads can be tested in unison either in hotcondition or at ambient temperature with either ofthe two FM Heads as upstream or downstreamhead. Each of the other two channels can be used

204


for testing an FM Head independently either asupstream or downstream under ambienttemperature. The necessary forces (hydraulic dragforces / ram forces of the opposite end FM Head)can be simulated by using a test ram housed in theopposite end of the test channel

6.3 Ram Assembly Test Facility

It is desirable to test a Ram Assemblyindependently before it is assembled to an FMHead. For this purpose a Ram Assembly Test Rig(RATR) has been provided. A test piston withlabyrinth seals is housed inside the RATR togenerate forces required to adjust/calibrate theforces of the B-Ram/C-Ram/latch.

6.4 Other Facilities

Several test rigs have been located in the CMF toindependently assemble, test, calibrate and fine

tune various sub-assemblies of the FM Head (suchas snout assembly, coolant channel closure plugsetc.). Further, provision has been made in the CMFto decontaminate the FM Head before taking upany work on it.

7. CONCLUSIONS

It has been possible for us to briefly cover in thispaper only some of the improvements andinnovative features of the 500 MWe Indian PHWRFHS design.

It is well recognised that the performance of the on-power FHS play a major role for the success ofPHWRs. We hope that these features willcontribute in a large way for a troublefree andoptimal performance of the FHS.

205

FIRST INTERNATIONAL CONFERENCE ON CANDU FUEL HANDLING SYSTEM. MAY 1996

S3o

//////?/*////' S ' f SSS SJ^*^ .—H

I PRESSURE HOUSING. S MAGAZINE ORIVE W0EX1N0 ASSY,

Z ENO COVER. . fc HIGH PRESSURE COUPLING.

3 MAGAZINE ROTOR iSSY. 7. BEAR.NG HOUSING

4.INTERNAL GEAR 8.CENTRAL SLEEVE

FIG. I SIMPLIFIED VIEW OF MAGAZINE ASSEMBLYI3OO UW| PHWRI

1 ENO FITTING.

2 CENTRE SUPPORT.

3 SKINNER SEAL.

4 CLAMPING BARREL.

3 SCREW

6 WORM AND WHEEL ORIVE.

7 SNOUT EMERGENCY LOCK.

8 MAGAZINE HOUSING ENO COVER

9 OUTER SUPPORT

K> CLAMPING POST

11 WEDGE SEGMENT

FIG. 2 SIMPLIFIED VIEW OFSNOUT ASSEMBLY

(500 MWa PHWRI

FIRST INTERNATIONAL CONFERENCE ON CANDU FUEL HANDLING SYSTEM, MAY 1996

REED SWITCH

O

C-RRM TUBE 1 C-RRM PINION

FRONT HOUSING-

-RERR HOUSING

C-RRM LRBYRINTH

•B-RRM LRBYRINTH

C-RRM RRCK-

-8-RRM TUBE

B-RRM PINION — '

FIG. 3 SIMPLIFIED SKETCH OF RAM ASSEMBLY

( 5 0 0 MW«. Pi-

-B-RRM RRCK

• BALL SCREW HOUSING

-POTENTIOMETER

-GEAR BOX

-MOTOR

•SPUR GEXR BOX HOUSING

-SHIELDING SLAB

-SHIELDING BLOCK

TM HOUSING

CARRIAGE

FUEL GUIDE

FIG. A SIMPLIFIED VIEW OF TRANSFER MAGAZINE

(500 MWe PHWRI

FIRST INTERNATIONAL CONFERENCE ON CANDU FUEL HANDLING SYSTEM. MAY 1996

LEAKASE COUJICTION

Ooo

ROPE 2H.VE MECHANSM

FUEL SHUTTER

TOP HOUSING:

SETTLETRANSPORT „•RISE - V

BUSH OVIOCS '

SPENT FUELSTORAGE BAY

SPENT FUEL-s,RECEIVING BAY

FUEL TRANSFERROOM

RB-1

FM VAULT/—IFM SERVICE AREA

FM AIRLOCKFM TUNNEL

RB-2

FIG. 5: SIMPLIFIED VIEW OF SHUTTLETRANSFER STATION

(500 MWo PHWR)

FlG:6 LAYOUT OF FUEL HANDLING EQPT./AREAS(500 MWe PHWR)


MSTRWCMTCALBRATMN

FACUTYMETROLOGY SUB ASSEMBLY RU.

TEST FACILITIES

OKP-t CHP-M

WHC»-» J WHAO-t

M A M *UAKAOCOOUXOTXM

TESTMO AREAFOR SOUTH

FMHEAD

RAH ASSEMBLY TIST BO

FM KCONTAMMATKW AREA(SOUTH) FM OECONTAMMATION AREA

(NORTH)

PROCESS EQUIPMENT AREAm-i

FM MAMTENANCE AREA FM STORAOE AREA

DECOHTAMMATIONFACILITY

ACRONYMS:

WHCTF - WATER HYDRAULIC COMPONENTS TEST rACUTYOPP - OL POWER PACKOHCTF - OL HYDRAUUC COMPONENTS TEST FACUTYVHCS - WATER HAORAUUC CONTROL STATIONWHAC - WATER HYDRAULIC ACTUATOR CASKETRATF - RAM ASSEMBLY TEST FACLITYOHP - OL HYDRAULIC PANELTC - TEST CHANNEL

HCRfCOPtPJP3HXHRFRTK

- H O T CHANNEL- RAM FORCE CAU8RATKW DEVICE- TRIPLEX PUMPS•CENTRTUQAL PUMP• HEAT EXCHANGER- ELECTRIC HEATER-FLTER-TANK

FIG:7 CMF LAYOUT AT FLOOR EL. 93.00 M.

209!5EXT PACJ-(S)

left BLANK


ANALYSIS OF FUEL HANDLING SYSTEM FOR FUEL BUNDLE SAFETYDURING STATION BLACKOUT IN 500 MWe PHWR UNIT OF INDIA

MADHURESH R., R. NAGARAJAN, EVDER JIT AND A. SANATKUMAR

Nuclear Power Corporation of India Ltd.,Mumbai-400 085, INDIA

CA9700741

ABSTRACT

Situations of Station Blackout (SBO) i.e. postulatedconcurrent unavailability of Class III and Class IVpower, could arise for a long period, while on-powerrefuelling or other fuel handling operations are inprogress with the hot irradiated fuel bundles beingany where in the system between the ReactorBuilding to the Spent Fuel Storage Bay.

The cooling provisions for these fuel bundles arediverse and specific to the various stages of fuelhandling operations and are either on Class III oron Class II power with particular requirements ofinstrument air. Therefore, during SBO, due to thelimited availability of Class II power and instrumentair, it becomes difficult to maintain cooling to thesefuel bundles. However, some minimal cooling isessential, to ensure the safety of the bundles.

As discussed in the paper, safety of these fuelbundles in the system and/or for those lying in theliner tube region of the reactor end fitting isensured, during SBO, by resorting to passivemeans like 'stay-put', 'gravity- fill1, 'D2O- steaming'etc. for cooling the bundles. The paper alsodescribes various consequences emanating fromthese cooling schemes.

1.0 INTRODUCTION

The on-power refuelling operation is commencedonly when Class IV power is available. However,Station Blackout (SBO) f.e. postulated concurrentunavailability of Class IV and Class III electricalpower, may occur when refuelling or other fuelhandling operations are in progress. Under suchsituation hot irradiated fuel bundles could bepresent anywhere in the Fuel Handling System(FHS) and/or in the liner tube region of the

downstream reactor end fitting i.e. out of thethermosyphoning flow path of the coolant channel.

For the safety of these fuel bundles, necessarycooling is required in order to avoid theiroverheating due to the decay heat. However,during SBO, when Class III power is not availableand availability of Class II power as well asinstrument air is limited, the requirement of coolingthese fuel bundles becomes difficult to achieve,since:

a) The normal cooling provisions requiresubstantial electrical power.

b) The fuel handling operations are performed in anumber of stages, with specific coolingprovisions for each stage.

c) The equipment of the FHS are located atdifferent places from the Reactor Building (RB)to the Spent Fuel Storage Bay (SFSB), Fig.1.

Therefore, in order to assess the effect of SBO onthe safety of the irradiated fuel bundles present inthe FHS, various scenarios within the system, likelyto exist at the time of SBO, have been identifiedand analysed. Further, a detailed strategy has beendeveloped to provide adequate cooling to thebundles with the minimal power consumption.

2.0 POINTS CONSIDERED IN SBOHANDLING STRATEGY AND COOLINGSCHEME FOR FHS:

The SBO handling strategy would greatly beinfluenced by the exact stage of fuel handlingoperations being carried out at the time of SBO.

Following general requirements were set up forevolving the SBO handling strategy for FHS :

211


a) The scheme to be conceptualised for coolingthe fuel bundles should make best use of thealready available system, the interconnectionsbetween FHS and Primary Heat TransportSystem (PHT) and the inherent coolingcapabilities of FHS.

b) In case a cooling scheme can not be evolvedwith the existing hardware, or, the schemeusing the existing circuit is not practicable,modifications in the design should be kept tothe bare minimum.

c) Scheme should be general, to the extentpossible so as to be applicable for most of thestages of fuel handling operations.

d) The electric power and pneumatic powerrequired should be low.

e) Loss of D2O and transfer/leakage of heavywater from high pressure circuit of PHT to lowpressure circuit should be as low as possible.

f) The FHS should be brought to safe and stablestate, at the earliest, requiring minimaloperations.

g)

h)

Maintaining of the cooling to the spent fuelbundles should be with minimum interventions.

Safety of the system should preferablyestablished with minimum feed backs.

be

i) Entry to and stay in the RB should be minimal.

3.0 PRINCIPLE ADOPTEDCOOLING OF FUEL BUNDLES:

FOR

The SBO would be declared if Class III power is notrestored even after 6* minutes of Class-IV powerfailure. This implies that the hot irradiated fuelbundles in the downstream liner tube region, if any,or in the FHS would have seen at least 6 minutesof forced water cooling due to the FM supply pump,on Class II power, continuing to operate till thedeclaration of SBO. The decay power of the fuelbundles at this instance would be almost 3% of thefull power. Such hot irradiated fuel bundles ifexposed to nearly stagnant air environment, would

" 6 minutes is reasonable time for the operator to attemptmanual connection of Diesel Generator supply to thebuses, in case of failure of automatic restoration system.

reach a steady state sheath temperature of theorder of 1500 K [1]. The order of the temperature instagnant steam environment has been found to bethe same. In the process of rise in temperature tosuch values, sheath splitting due to oxidationswelling of fuel, is probable [2], as the oxidisingenvironment could exist due to the thermal andradiation decomposition of water or due to thepresence of air.

However, it has been found that as long as suchfuel bundles are kept immersed in water, even ifthe water is boiling without any forced circulation,the clad surface temperature of the fuel bundlewould be close to the saturation temperature ofwater corresponding to the prevailing pressure.This is established through the following calculationprocedure:

a) Calculation of sheath temperature for themaximum power fuel element of the maximumpower fuel bundle (with 6 minutes of coolingperiod) assuming the element to be immersed,in isolation, in the boiling pool of water at asaturation temperature of 583 K correspondingto the maximum operating condition at thereactor outlet (The sheath temperature isevaluated to be approximately 5 K higher thanthe saturation temperature).

b) Verification of the above result, on theextrapolated boiling curve, Fig. 3, obtained bydata from Fig. 4 (The sheath temperature isfound to be nearly same as that obtained instep 'a' above).

c) Extrapolation of the result for the bundlegeometry by estimating the critical heat flux forthe bundle with the help of equation 10.112 of[3] (point A in Fig.3) and using the conceptdepicted in Fig.5 (It is concluded that thesheath temperature of any element in the fuelbundle would be of the same order as that ofthe single fuel element immersed in water, inisolation, as stated in 'a' above).

From the above calculations it is established that inthe system, sheath temperature can go to about588 K, at the most. This is considered acceptablefor the safety of the fuel bundles.

Therefore, in order to dispense with the use ofpumps during SBO, in absence of forcedcirculation, the fuel bundles are required to be keptimmersed in water for their safety.

212


4.0 POSSIBLE SCENARIOS FOR FHSDURING SBO:

In order to understand the behaviour of the systemduring SBO, especially from the point of view ofsafety and cooling of the fuel bundles with thetruncated cooling provisions, and also to evolve thecooling scheme for these fuel bundles, followingscenarios have been analysed:

A) Fuelling machines (FMs) hydraulicaliyconnected to channel (i.e. on-reactor mode)and downstream shielding plug yet to bewithdrawn.

B) FMs hydraulicaliy connected to channel (i.e.on-reactor mode) and downstream FMreceiving (or has just received) spent fuelbundles.

C) FMs hydraulically isolated from channel (i.e.off-reactor mode/park mode) with spent fuelbundles in one of the FM.

D) FM hydraulicaliy connected to transfermagazine (i.e. fuel transfer mode) with spentfuel bundles in FM and/or transfer magazine(TM) and/or in the tube connecting the FM andTM.

E) Spent fuel bundles under transit storage in TM.

F) A pair of spent fuel bundles in the process oftransfer from TM to shuttle transfer station (i.e.dry transfer mode), Fig 2.

G) A pair of spent fuel bundles in the process oftransfer from shuttle transfer station (STS) toSFSB.

H) I) Spent fuel bundles in an emergency tube1

(ET) of Rehearsal Tube Facility (RTF).

ii) FM with spent fuel bundles in hydrauliccommunication with an ET of RTF.

1 Among the 3 tubes in the Rehearsal Tube Facility 2 areEmergency Tubes which are same as reactor channeland are meant for storing spent fuel bundles inexigencies

5.0 ANALYSIS OF VARIOUSALTERNATIVE COOLING SCHEMESCONSIDERED FOR THE IDENTIFIEDSCENARIOS:

Various possible cooling schemes for individualscenarios were analysed in detail. Salient featuresof the analysis for a typical scenario are given inTable 1.

The cooling scheme adopted for various scenariosis summarised in Table 2.

6.0 IMPORTANT RESULTS OF THEANALYSIS:

6.1 Thermosyphoning Between FHS and PHT

Thermosyphoning between the FHS and theinterconnected PHT is possible to be establishedduring off-reactor mode / park mode, fuel transfermode and during the scenario pertaining to the ET(i.e. Scenario C, D and H respectively as given inSection 4.0). However, owing to the inadequatedriving force, thermosyphoning in these casesleads to exposure of the fuel bundles to nearlystagnant steam environment which is unacceptablefor fuel bundle safety as discussed in Section 3.0.

6.2 Cooling of Hot Irradiated Fuel Bundles

Cooling of hot irradiated fuel bundles present in thedownstream liner tube and in FHS is ensured in linewith the principle described in Section 3.0. The fuelbundles present in FM/TM/downstream liner tuberegion/ET would be cooled as follows:

a) Initially by sensible heating of the equipmentand the hold-up water and later by steaming ofthis water.

b) Venting of steam from FM/TM/ET as the casemay be.

c) Make-up of water (lost due to steaming andventing) fromi) the channel, in case FM and channel are inhydraulic communication;

ii) the PHT storage tank (by gravity flow) in allthe remaining cases.

213


The fuel bundles in SFSB and in case present inSTS during SBO, would be cooled by thesubcooled light water of SFSB/STS.

In case, a pair of spent fuel bundles is undergoingthe dry transfer operation at the time of Class IVpower failure, the pair would be quickly brought towater environment in TM or STS, depending uponthe situation (motive power for this would beavailable as it would take about 5 minutes).Afterwards the cooling of the pair would be asdescribed above.

6.3 Loss ofD^O Inventory

The escape of D2O in the form of steam, from thesystem to the FM vault, in 8 hours (the designtarget for SBO handling capability) could be at themost 1200 Kg., considering the following:

a) 13 irradiated bundles of maximum powerchannel (5.5 MWth) with an average coolingperiod of 10 minutes, out of thethermosyphoning loop, with FM on On-reactormode.

b) 8 spent fuel bundles of the maximum powerchannel with an average cooling period of 45minutes, being in the TM.

In case (a), 700 Kg. from high pressure circuit ofPHT and in case (b), 500 Kg. from FT system /lowpressure circuit of PHT would escape to FM vault.

6.4 Tritium build up in accessible area (V2 area) ofRBas a result of Release ofDfi Steam from the FHS

The steaming from FHS would commence onlyafter at least one hour from the declaration of SBO,when equipment containing the fuel bundles andthe water inventory submerging the fuel bundles inthese equipment, get heated-up. Therefore, build-up of tritium in FM vault and in other high enthalpyareas (V1 area) may occur only after the first hourof SBO. Due to the stipulated leak paths betweenV1 and V2 areas, the tritium build-up in V2 area isconservatively evaluated to be about 22 DAC at theend of 4 hours and about 70 DAC at the end of 8hours, Fig. 8. This is considered acceptable underSBO conditions.

7.0 CONCLUSION

A detailed SBO handling strategy for FHS, hasbeen developed, to bring the system to stable statewith minimal energy consumption.

The safety of the irradiated fuel bundles presentanywhere in the FHS, is ensured during SBO, byresorting to the inherent, passive cooling capabilityof the system. However, this will eventually resultin some escape of D2O and subsequent build-up oftritium.

REFERENCES

1. G.PAL, S.G.MARKANDEYA andV.VENKATRAJ, "Analysis of the TransientBehaviour of 37 Rod Fuel Cluster of 500 MWePHWR During Spent Fuel Handling Operation",Proceedings of TENTH National Heat & MassTransfer Conference; paper no. 111-14;Srinagar, India, 1989.

2. J.NOVAK and G.MILLER, "Dry Fuel Handling:Station Experience and Ontario Hydro (CNS)Program", Proceedings of InternationalConference on CANDU Fuel, pp 322 -345,Chalk River, Canada, 1986.

3. ROBERT H. PERRY and DON GREEN,"Perry's Chemical Engineers Handbook", SixthEdition, McGraw Hill.

4. COULSON J.M. and RICHARDSON J.F."Chemical Engineering, Volume 1, Fluid Flow ,Heat Transfer and Mass Transfer", ThirdEdition, Oxford Pergamon Press, 1977.

5. WARREN L. McCABE , JULIAN C. SMITH andPETER HARRIOT, "Unit Operations ofChemical Engineering", Fourth Edition,McGraw Hill.

6. ERNST U. SCHLUNDER et al, "HEDH- HeatExchanger Design Handbook, Volume 2, FluidMechanics and Heat Transfer", HemispherePublishing Corporation, 1986.

214


TABLE 1

ANALYSIS OF VARIOUS COOLING SCHEMES TO EVOLVE THE SCHEME TO BEADOPTED FOR HANDLING SBO FOR THE SCENARIO OF,

"FUELLING MACHINES HYDRAULICALLY CONNECTED TO CHANNEL ANDDOWNSTREAM FM RECEIVING/RECEIVED SPENT FUEL BUNDLES"

Highlights of Alternative Schemes considered for handling SBO :

I. SWITCH OFF ELECTRICAL LOADS and STAY-PUT:• Scheme Leads to exposure of fuel bundles to stagnant steam environment, Fig.6.

II. 1PUSH THE IRRADIATED FUEL BUNDLES PRESENT IN LINER TUBE REGION, BACK IN TO THETHERMOSYPHONING FLOW PATH OF THE CHANNEL :

• Feasible only when 17 or less fuel bundles in the channel.• Operation of fuel stop will be required needing instrument air.• Scheme takes nearly 10 min. for execution and consumes 30 kwh energy.• Bundles will stay in thermosyphoning flow as the drag force on them is negligible.

III. 1BOX-UP THE CHANNEL AT DOWNSTREAM END FITTING :• Possible only when 17 or less fuel bundles in the channel.• Guide sleeve to be removed, seal plugs to be installed.• Scheme takes nearly 10 min. for execution and consumes 30 kwh energy.

IV. CIRCULATION OF WATER IN FM BY FUEL TRANSFER (FT) D2O PUMP :• Pump would be required for almost entire duration of SBO.• Heat sink would be required for removal of decay heat from FT heat exchanger.

V. INJECT LIGHT-WATER INTO THE FM :

• Scheme would downgrade entire PHT.

VI. VENT THE STEAM FROM FM and MAKE-UP THE W A T E R IN FM BY CHANNEL WATER.

a) VENT BY VENT VALVE• Constant monitoring of D2O level in FM and constant adjustment of vent valve will be required.• Scheme leads to exposure of hoses to steam.

b) VENT BY AIRTRAP, FIG. 7• Passive scheme, needs only vahong-in of the trap.• No exposure of hoses to steam.

SCHEME ADOPTED : VENTING BY AIR TRAP AS PER VI(b) ABOVE IN LINE WITH THE POINTS STATED INSECTION 2.0 AND THE PRINCIPLE OF COOLING DESCRIBED IN SECTION3.0.

1 The scheme has been worked out with the assumption that no operation is required to be carried out on the upstreamFM. Hence, the scheme is applicable only when 17 bundles are existing in the channel as 13 irradiated bundles,originally in the core, would be brought back, allowing additional 2 pairs to be residing in the space between the FMand the channel.

215

N>


TABLE 2SUMMARY OF COOLING SCHEMES FOR VARIOUS SCENARIOS'

ON-REACTORMODE,DOWNSTREAM FMRECEIVINGRECEIVEDBUNDLES

SCENARIO B"

ORFUEL

VENTING OF FM BYAIR-TRAPVENTINGSTEAM,LIQUIDWOULDINGRESSEDFM FROM

(ONTHETHED2O

BEINTOTHE

CHANNEL & HENCEKEEPING THEFUEL BUNDLES INFM IMMERSED INLIQUID D2O)

OFF-REACTORMODE, FM WITHEIGHT SPENT FUELBUNDLES

SCENARIO C*

SAME AS THATFOR SCENARIO B(ON VENTING, THELIQUID D2OWOULD COMEINTO THE FMFROM PHTSTORAGE TANKDUE TO GRAVITY &HENCE KEEPINGBUNDLES IN FMIMMERSED INLIQUID D2O)

FUEL TRANSFERMODE FUELBUNDLES IN FMAND/OR IN TM

SCENARIO D#

SAME AS THATFOR SCENARIO B(AFTER VENTING,THE D2O IN FM &TM WOULD BESUFFICIENT TOKEEP THEBUNDLESSUBMERGED INLIQUID D2O.HOWEVER,WHENEVERREQUIRED BASEDON TM LEVELTRANSMITTERSIGNAL, THE TMMAY BE FILLED BYGRAVITY FLOWFROM PHTSTORAGE TANK)

FUEL BUNDLESUNDER TRANSITSTORAGE IN TM

SCENARIO E*

FILLING UP OF TMBY GRAVITY FLOWFROM PHTSTORAGE TANK &SIMULTANEOUSVENTING OF TM BYAIR TRAP.

PAIR OF FUELBUNDLESUNDERGOING DRYTRANSFER

SCENARIO F#

THE PAIR OF FUELBUNDLES TO BEBROUGHT TOWATERENVIRONMENTEITHER BYRETRACTING THEPAIR TO TM OR BYTAKING THE PAIRTO STSDEPENDING UPONTHE SITUATION.AFTERWARDSCOOLING WOULDBE AS PER EITHERTHE PREVIOUSSCENARIO OR THENEXT SCENARIOAS THE CASE MAYBE.

PAIR OF FUELBUNDLES INSHUTTLETRANSFERSTATION

SCENARIO G*

FOR KEEPING THEBUNDLESIMMERSED INWATERENVIRONMENT,SUFFICIENTQUANTITY OF H2OABOVE THE PAIROF BUNDLESWOULD BEAVAILABLE.

FUEL BUNDLES INEMERGENCYTUBE, OR FM WITHSPENT FUELBUNDLES INCOMMUNICATIONWITH ANEMERGENCY TUBEOF REHEARSALFACILITY

SCENARIO Hi* & Hii*

VENTING OFEMERGENCY TUBEBY AN AIR-TRAP(ON VENTING THETUBES, BUNDLESWOLD BE KEPTIMMERSED IN D2OBY FILLING THETUBE DUE TOGRAVITY FLOW INTHE MANNERSAME AS OFF-REACTOR MODE).

* Scenario A * does not have cooling requirements hence not indicated here.* For description of title of scenarios, refer to section 4.0.


NOTESi-

1. THE REFUELLING IS OONE IN THEDIRECTION OF FLOW. NORMflLLY EIGHTFUEL BUNOLES RRE REFUELLEDIN ONE REFUELLING OPERATION.

2. SHUTTLE TRRNSFEK STBTION ISTS)ON NORTH SIDE I NEW FUEL MRGRZINE ETC.ST SOUTH SIDE NOT SHOWN1 FOR CLBRITY.

3. NORMALLY NEW FUEL BUNDLES FROMTRANSFER MAGRZINE ITM)RRE EXCHANGED WITH SPENT FUELBUNDLES IN m. US SHOWNIN SOUTH SIOE, WHEN FM ISIN COMMUNICRTION WITH TM. FORCLRRITY. NEW FUEL LOROINGROUTE TO FM IS SHOWN HEREON NORTH SIDE ONLY.

4. FROM TM, RT R TIME. ONE rRIR OF SPENTFUEL BUNDLES IS DISCHARGED TO BRY VIRSTS. TRANSFER OF THE PRIR FROM TM TOSTS IS DONE IN ORY ENVIRONMENT BS SHOWNIN SOUTH SIDE flND DETAILED IN FIG. 2.

RCRONYMSi-

FM - FUELLING MRCHINE

ICW - INNER CONTRIKMENT VRLL OF RERCTOR SUILDINC

NFM - NEW FUEL MBGRZINE

OCV - OUTER CONTRINNENT WRLL OF RERCTOR SUILOING

SFSB - SPENT FUEL STORRGE BRY

SFTT - SPENT FUEL TRBNSPORT TUBE

STS - SHUTTLE TRRNSFER STRTION

TM - TRBNSFER MBGRZINE

NEW FUEL

SPENT IIRRROIRTEOI FUEL

OPERRTION IN 03D ENVIRONMENT

OPERRTION IN DRY ENVIRONMENT

OPERRTION IN H3O ENVIRONMENT

FIG. 1 SCHEMATIC OF MAJOR FUEL HANDLING OPERATIONS

• RRM BE(H20 OPERflTED)

FUEL BUNOLE-5HUTTLE

RflM BF-(H20 OPERRTEO)

/-CRRRIflGE TUBE

-ROM FIB(D20 OPERRTED)

TRRNSFER MHGflZINE

SHUTTLE TRRN5FER STRTION

FIG. 2 DRY TRANSFER OPERATION

44M447T/CV8X4/UM3

217


*

A _

C-

JLj

/

jj'/

1

/V

f

- • • • •

/

ti

1

If

7/

—

\\ —'"\

_

— -

—10

2 S W 20 80 WO

A • .TEMPERATURE DIFFERENCE. K

LEGEND:-

= STANDARD CURVE FOR WATER BOILING AT 373 K

EXTRAPOUTED CURVE. FOR WATER BOILING AT

583 K. BY EVALUATING POWT 0 FROM FIG. 4

(EVALUATION W FIG. 4 IS SHOWN IN CHAIN DOTTED LINE).

C •= HEAT a U X OF MAXIMUM POWER FUEL ELEMENTOF THE MAXIMUM POWER FUEL BUNDLE.

B " HEAT FLUX OF MAXIMUM POWER ELEMENTCORRESPONDING TO THE MAXIMUM POWER FUEL BUNDLE.

A - THE CRITICAL HEAT FLUX OF THE FUEL BUNDLEEVALUATED FROM EQUATION 10.112 OF [3 ] .

FIG. 3 EFFECT OF TEMPERATURE DIFFERRENCE (BETWEEN THE SURFACE & POOL WATER)

ON HEAT FLUX FROM THE SURFACE. KEPT SUBMERGED IN BOILING WATER.

(r«f. FW. 7J7 or [4])

/// ij ji ;

\

k\

250

X 200

5 ISO

50

3 0.1 0.2 0.3 0.4 0.B O.C 0-7 0.8 0-9

pr - p / p c _

(P-OPERATINO PRESSURE, Po • CRITICAL PRESSURE FOR WATER)

FIG. 4 EFFECT OF PRESSURE ON 'PEAK HEAT FLUX IN THE

BOILNO CURVE FOR WATER I.e. (q) mox' AND'COR-

-RESPONDING TEMPERATURE DIFFERENCE. I . e . A 0 C '(ref. FTO. 7.40 OF [4] • HO. O.7 OF [8])

20

10

<

BUNDLE—v

Jy<!//

SINGLE ELEMENT

/

A e

FIG. 5 CONCEPTUAL COMPARISION OF BOILING OUTSIDE

BUNDLE WITH THAT OUTSIDE SINGLE ELEMENT

(r»f. FIO. S M SECTION 2 . 7 . S Of [ 6 ] )

218


RIH

i r

T°3

SG

LINER TUBE HOLES —w

VROH

i

TO r-

UPSTREAM FM

NOTE:THE FIGURE OPICTS THAT 8 SPENT FUELBUNDLES ARE RECEIVED M DOWN STREAMFM AND PROLONGED iNTERRUPTION M SUPPLYHAS LED TO FLOW STRATOTCATON AND 9O8JNO.

ACRONYMS:

FM - FUELLING MACHKERM . REACTOR MLET HEADERROH - REACTOR OUTLET HEADERSO - STEAM OENERATOR

FUEL STOP

D20 STEAM •

DOWNSTREAM FM

FIG. 6 FLOW STRATIFICATION IN FUELLING MACHINE

•TO DRAIN TANK

SV343

020STEAM TO FM VAULT

AIR TRAP

NOTE:FOR HANDLING SBO. CLOSE SV 343AND OPEN SV347 (AFTER FMPRESSURE B LOW).

FM VENT LINE

FUELLING MACHINE (FM)

FIG. 7 VENTING OF FM BY AIR - TRAP

219

70-

8

TIME t (STARTING FROM CLASS IV FAILURE AT t - O ) , h r .

FIG. 8 TRITIUM BUILD UP IN V2 AREA IN 8 Hrs.OF SBO(CORRESPONDMO TO 1200 K«. 020 RELEASED FROM

&SEXY PAGE(S)left

44Btok/B2«/e3/223M


EVOLUTION OF THE DESIGN OF FUEL HANDLINGCONTROL SYSTEM IN 220 MWe INDIAN PHWRs

BYCA9700742

L.Dhruvanarayana, H.Gupta,& M BharathkumarNuclear Power Corporation of India Ltd,

Vikram Sarabhai Bhavan, Mumbai-400094, India

ABSTRACT

Following two CANDU type reactors at Rajasthan(RAPS-1&2), three nuclear power stations, each oftwo units of 220 MWe has been in operation atRajasthan (RAPS-1&2). Madras (MAPS-1&2).Narora (NAPS-1&2) and Kakrapar (KAPS-1&2).Two more stations, also of 220 MWe capacity, areunder construction at Rajasthan (RAPP-3&4) andKaiga (Kaiga-1&2). These are natural uraniumfuelled pressurized heavy water cooled and heavywater moderated reactors (PHWRs). The two unitsat Rajasthan viz RAPS-1&2, were built with thetechnical collaboration with Canada and the rest ofthe units have been designed and built indigenouslyincorporating a number of modifications,particularly in the on-power refuelling system. Theevolution of the design of the Fuel Handling Controlsystems of these reactors, taking into considerationoperational needs, safety aspects andmaintainability are highlighted in this paper.

A combination of hydraulic and electronic controlhas been provided to enable the operations. InRAPS-1&2, hardwired electronic controls wereprovided, while in MAPS-1&2, the hardwiredsystem was improved. From NAPS onwards, acomputerized control system with hardwiredinterlock logic has been provided. New devices likecoarse-fine potentiometers, special oil filledpotentiometer assembly, rectilinear potentiometersetc., were specified from NAPS onwards.Positioning logic is computerized providingflexibility and expandability. Digital panel meters

and indicating lamps have been provided formanual mode operations, while CRT monitors helpin computer mode operations.

Hydraulic controls which comprise D2O hydraulics,H2O hydraulics and oil hydraulics have beenimproved from NAPS onwards. Hydraulic panelshave been relocated in accessible areas to reducemanrem and for better maintainability. All electricdrives including X and Y drives were modified ashydraulic drives for better control. New types ofvalves have been employed.

The paper highlights the details of the changesincorporated.

1. INTRODUCTION

The pressurised Heavy Water reactors (PHWRs)with natural uranium fuel require regular andfrequent on-power refuelling. In Indian PHWRs of220 MWe, a complex Fuel Handling System mainlycomprising two Fuelling Machines and a FuelTransfer System is provided to enable refuelling.An elaborate control system with the following mainfeatures has became necessary for enablingautomatic and remote manual operations.

a) Automatic Sequential Logicb) Safety interlock logicc) Continuous feedback process controld) Accurate positioning at precalibrated positions.e) Torque, Force and speed control for the drives.

221


f) Manual operability and system statusindicators.

It has also been considered desirable to have anoperational logging system in addition to the above.

These requirements have been met by providingsuitable Electrical and Hydraulic controls. TheHydraulic controls meet the requirement (e)mentioned above whereas the rest of therequirements are met by Electrical controls. Thetwo CANDU type reactors at Rajasthan AtomicPower Station (RAPS-1 &2)and the two reactors atMadras Atomic Power Station (MAPS-1&2) wereprovided with hardwired Electrical control systems,whereas in subsequent reactors at Narora (NAPS-1&2) and Kakrapar (KAPS -1&2) computerisedcontrol systems have been provided. The abovereactors are in operation. Four units, two at Kaiga(Kaiga-1&2) and two at Rajasthan (RAPP-3&4) areunder construction.

2. ELECTRICAL CONTROLS

The Fuel Handling System is provided withextensive electrical controls to enable automaticsequential operation while ensuring safety ,reliability and maintainability.

2.1 CONTROL SCHEMES

In RAPS-1&2 and MAPS-1&2, a totally hardwiredsystem was provided. Discrete component logicunits, employing Germanium transistors wereemployed in RAPS-1 &2, whereas these wereimproved using silicon transistors in MAPS-1&2.The positioning controls employed hardwired tripunits.

From NAPS onwards, a computerised system wasadopted in view of the following advantages:

I. Better man-machine communication with thehelp of CRT messages.

II. Flexibility in positioning, interlock andoperational program logic provided by software,compared with wiring changes that wereneeded in previous projects.

III. On-line diagnostic capability.

IV. Possibility of performance evaluationsuitable data logging and analysis.

by

The computerised system provided in NAPS-1&2and KAPS-1&2, employed a 16 bit supervisory

master computer with two separate 8-bitmicrocomputers. The master computer analysesthe operational demands, as entered by theoperator and initiates the various tasks that themicrocomputers have to execute. The software isorganised in such a way that a full auto modeoperation, including automatic sequential operation,is possible with both master and microcomputersbeing functional, whereas, a degraded or semi-automode is possible with only the microcomputersbeing functional.

A manual mode is also provided, alongwith a fullfledged control console, so as to enable completionof operations to bring the system to a safe state, inthe event of any failures of the computerisedsystem.

The application software for the computerisedsystem is written in a specially developed higherlevel language called Process Control Language(PCL). A compiler has been provided with themaster computer, so that the codes, which arecompiled off-line, can be stored on the mastercomputer disks, and transferred to microcomputersas required. The transferred codes are interpretedby an Interpreter program in the micro computer. Itis possible to calibrate the various discretepositions, by bringing the drive to the requiredposition and invoking the calibration program. Themicrocomputer enables automatic positioning andalso checks the safety interlocks. No command isissued to the field device unless the interlocks aresatisfied, as checked by the microcomputer andalso by a hardwired safety interlock logic, whichdoes not depend on the computerised system. Thishardwired IC-based safety interlock logic alsoensures safety for the commands issued by theoperator in the manual mode. During the sequentialoperation, if any of the permissives are notavailable, these are indicated on the CRT monitorson the console by proper messages.

On the control console, digital panel meters havebeen provided to enable positioning of the drives, incases, where microcomputers develop faults duringoperation. The operator will be required to refer tothe calibration record in such cases.

The safety interlock logic in NAPS-1&2 employsTTL integrated circuits, as compared with discretecomponents like silicon transisters in MAPS-1&2.CMOS logic units are employed in RAPP-3&4 andKaiga-1&2. Comparator units are employed togenerate logic signals based on analog field

222


signals. These logic signals are used in interlocklogic.

While in RAPP-3&4/Kaiga-1&2, the controlphilosophy is mostly comparable to that of NAPS, afew improvements have been incorporated indesign. Arithmetic capability to some extent hasbeen added in the PCL. A 16-bit microcomputer willbe employed, thereby enhancing the computingpower considerably. This will enable limited use ofautomatic sequential operations even without thesupervisory computers. Positioning can be donebased on the signals from either the primarydevices or back-up devices, comparator units willhave the set point potentiometers locatedindependently, so that replacement of the defectivecomparater units does not require adjustment of setpoints. A PC-based operational logging sub systemwill also produce a summary print out apart fromstoring a detailed log in the computer.

2.2 SENSORS

In comparison with RAPS/MAPS, a coarse-finepotentiometer assembly has been employed as aposition monitoring sensor for many drives (e.g. B-ram of Fuelling Machines)from NAPS onwards.This has a ten turn potentiometer with a single turncontinuous rotation potentiometer on a commonshaft, thereby enhancing the resolution andpositioning accuracy with a good signal to noiseratio. At certain places where two Linear Variabledifferential transformers (LVDT) were employed inRAPS/MAPS, a single long stroke LVDT has beenemployed from NAPS onwards. For sensing smallmovements, rectilinear potentiometers have alsobeen used for better compatibility with thecomputerised controls. Oil filled potentiometerhousing for C-Ram has been modified to utilise asingle turn continuous rotation potentiometer,thereby reducing probability of potentiometerfailures resulting from Ram'C rope breakages.Further this type of design has simplified theprocedure to replace the failed assemblies.

3. HYDRAULIC CONTROLS

Hydraulic controls have been provided to ensurethat the hydraulic cylinders and hydraulic drivemotors function to meet the requirements of thedesign of mechanical equipment. Hydraulic controlscan be categorised under Water Hydraulic controlsand Oil hydraulic controls.

3.1 WATER HYDRAULIC CONTROLS

Water Hydraulic controls employ Heavy water orlight water as the fluid medium, so as to ensurecompatibility with the process. The water hydrauliccylinders require the force and speed to becontrolled. Solenoid operated directional valves areemployed to enable the advancing and retracting ofthe cylinders. In general, Differential PressureRegulating Valves with the outlet pressurereferenced to the operating pressure of theassociated equipment, are employed to providecontrolled forces. However, control valves areemployed for C-Ram force control, so that differentforces can be selected on program commands.Speed control is achieved by providing throttlingvalves or fixed orifices.

Designs of most of the stations as regards waterhydraulic controls are similar. However from NAPSonwards the layout of the valve panels has beenchanged. Control devices for new actuators havebeen accommodated and also catenary hoses havebeen standardized. Fuelling Machine (F/M) valvepanels have been rearranged such that thestructural frame work meets seismic requirements.All catenary hoses have been provided with eitherin-line excess flow check valves or flow limitingorifices, to prevent or limit the D2O escape duringfailures of hoses.

3.2 OIL HYDRAULIC CONTROLS

These controls enable speed and torque control, inaddition to direction of movement for the various oilhydraulic drives. Many of the Electric drives,provided in the design of RAPS-1&2 and MAPS-1&2 were changed to oil hydraulic drives fromNAPS onwards. For example, the drive motors of Xand Y motion drives of Fuelling Machines have nowbeen made oil hydraulic from NAPS, whereas thesewere electric motors in RAPS-1&2 and MAPS-1&2.Similarly New Fuel Magazine and TransferMagazine drives of Fuel Transfer System are nowoil hydraulic. The design philosophy of speedcontrol by using Pressure Compensated FlowControl Valves and force/torque control by pressureregulating valves has been extended to the newdrives. In certain cases relief valves have beenprovided for force control. Also additional deviceslike counterbalance valves (CBV) have beenprovided. For example, for F/M Y-motion drivecontrol, CBVs are provided to prevent uncontrolleddescent whenever brakes are released.In RAPS-1&2 and MAPS-1&2, F/M oil power packunits and valve panels were provided on the

223


Fuelling Machines. However from NAPS onwards,these have been shifted to accessible areas, byemploying increased number of catenary hoses.This has resulted in better maintainability andreduction in manrem. Oil coolers with chilled waterflow have been provided to limit the temperaturerise of oil, thereby improving the performance.

4. SAFETY ASPECTS

The safety of Fuel Handling System will be ensuredmainly if the accidental unclamping is preventedand coolant to the spent fuel bundles is madeavailable at all the times. However, additionally, itis considered necessary to prevent hot wateringress into the Fuelling machines and to providesufficient interlocks for preventing mechanicaldamages resulting in difficulties for retrieval apartfrom leading to manrem consumption. Controlsystem design has been improved in theserespects. Interlocks are checked both in thehardware and software from NAPS onwards.Autoinitiated action based on redundant timershave been incorporated so as to ensure coolantduring dry transfer. In RAPP-3&4/Kaiga-1&2devices which are important for safety has beenidentified and it has been ensured that fail-safefeature is provided to a large extent. For safetyrelated functions, back-up devices are provided.The cable routes have been made diverse for back-up devices to prevent common cause failures to the

extent practically feasible. In the computerisedsystems, watch dog timers are provided. Any powersupply failure leads to a "Tripped state" which hasbeen made safe. In this state, all bistable valvesretain the last energised condition, while most ofthe oil hydraulic drives come to a halt. The failureof set point power supply for F/M magazinepressure control leads to closure of return linecontrol valves, preventing hot water ingress, if theFuelling Machine is in hydraulic communicationwith the coolant channel.

5. CONCLUSIONS

As can be seen from above, NAPS design which iscomputerised has also been improved with regardto operability, man-machine interface and state-of-art technology. Safety has been of prime concern inaddition to operating features. It may be noted thatthe performance of the system at NAPS/KAPS hasbeen reasonably good.In RAPP-3,4/Kaiga-1&2, inaddition to the above mentioned features,maintainability has also been given importance andit is expected that RAPP-3&4/Kaiga-1&2 will givebetter performance.

224