Foreword
The OECD Halden Reactor project is fuel, including mechanical deformations, an agreement between OECD member thermal behaviour, fission gas release, countries. It was first signed in 1958 and and corrosion. since then regularly renewed every third In the area of computer application, year. The activities at the Project is cen- the studies of the communication bet- tred around the Halden heavy water ween operator and process, and the sur- reactor, the HBWR. The research pro- veillance and control of the reactor core, gramme comprizes studies of fuel per- arc of particular interest for rcactor ope- formance under various operating condi- ration. tions, and the application of computers 1988 represents the 30th year since for process control. the Project was started, and this publica- The HBWR is equipped for exposing tion is produced to mark this event. It fuel rods to temperatures and pressures, contains a reprint of major part of the and at heat ratings met in modern BWR's publication produced at the 25th anni- and PWR's. A range of in-core instru- versary. In addition an account of the ments are available permitting detailed developments through the years 1983- measurements of the reactions of the 1988 is given. Contents
J 958-83 HOW IT STARTED 1 ORGANIZATION OF THE PROJECT 5 Agreements 5 Steering Bodies " Staff 6 RESEARCH AND DEVELOPMENT REVIEW 7 FUEL PERFORMANCE STUDIES 10 Background 10 Fuel Rod Thermal Behaviour 11 Fuel Rod Mechanical Interaction 14 Fission Gas Release and Fuel Defection Studies 16 Loss of Coolant Experiments 19 COMPUTER CONTROL RESEARCH 21 Background History 21 The Computer Age 22 Present Activities 24 Operator Interfaces 24 Man-Machine Communication . . . 25 Highly Reliable Computer Systems 26 Software Reliability 28 Computerized Operator Manual for Reactor Safety Systems ... 28 Alarm Handling 29 Core Surveillance and Control 31 Computer Control Experimental Facility 32 FUEL RESEARCH FACILITIES 34 HBWR as a Fuel Test Facility 35 Fuel Handling and Inspection Equipment 36 Irradiation Rigs 38 In-Core Instruments 44 Test Fuel Data Bank System 48 1983-88 FUEL PERFORMANCE STUDIES 1983-88 49 Introduction 49 Fuel Rod Thermal Behaviour 49 Fission Product Release 50 Pellet Cladding McchanicaJ Interaction 51 Cyclic and Transient Operation 52 Extended Burnup Operation 52 COMPUTERIZED MAN-MACHINE COMMUNICATION 54 Introduction 54 Man-Machine Interaction Research 54 The Alarm Handling System HALO 54 The Success Path Monitoring System SPMS 55 System for Early Fault Detection 55 System for Plant Diagnosis 56 The Process Operator Station CAMPS 56 The Core Surveillance System SCORPIO 56 Training Simulator for Off-Shore Oil/Gas Production Platform . 57 The Power System Simulator SASIM 57 Control Room Design 57 The Computerized Procedure Manual System COPMA 58 Software Reliability 58 FUEL RESEARCH FACILITIES . . . 59 High Pressure Light Water Loops in HBWR 59 PWR Waterside Corrosion Test Loop 60 Automatic Control of Experimental Conditions 60 In-Core Instruments '.,. - 61 Irradiation Rigs - - - • 62 Test Fuel Data Aquisition and Processing 63 STEERING BODIES AND PROJECT STAFF ...... 64 How it started
Cunnar Randers, astrophysicist by edu- cation, internationally recognized as a pioneer in the development of nuclear energy for peaceful applications, founder of the Norwegian Institute for Atomic Energy, ivas the key person behind the developments which resulted in the design and construction of the Halden Boiling Heavy Water Reactor and the establishing of the OEEC (later OECD) Halden Reactor Project in 195S. He has been hind enough to supply the editors of this publication with his own account of the events leading up to the successful international co-operation around the Haldcn Reactor:
The Halden Reactor is truly a remarkable during the 25th year of the Halden Pro- water reactors by Dr. Zinn in the U.S.A. animal! It was planned, designed and ject, became the first leader of the neu- Within a couple of months the NUPOP under construction before any nuclear tron physics division at JENER). plans were replaced by a boiling heavy power rcactor had been built commer- After three years of operating the water concept, and the proposed Joint cially in the world. Today it is still an reactor JENER wanted to move forward Dutch - Norwegian project was replaced activc experimental station for advanced by building a power experiment with by a pure Norwegian project even studies of problems in nuclear power. high temperature and pressure. There after the original project had been in- It has therefore functioned during the was a gentlemen's agreement that this se- cluded in the budget proposal in 1955 in whole life of nuclear power production. cond reactor should be built in the Norway. The new project was, as men- It seems like using the first spinning Netherlands. The power level should be tioned, approved preliminarily on 14th machine from the industrial revolution at least 20,000 kVV at an operating tem- June 1955, but with a condition that the for experiments in textile production perature of at least 200°C. Since ura- detailed financial and technical plans today. nium was impossible to buy, and we had should be placed before the parliament The design and plans for construction only a few tons of natural uranium by the end of the year for the final go were placed before the Norwegian »Stor- (from the Netherlands' remaining store) ahead signal. ting» by the government on 14th June we had to build a heavy water reactor. The fall of 1955 became a hectic 1955 and preliminarily approved. Some The thought of risking boiling in the period of negotiations for arranging bar- of you will discover by using your heavy water led us to a variety of diffi- tering between U.K., the Netherlands, pocket calculator thaL this is three years cult designs from pressure tank to the U.S.A. and Norway of heavy water, too early to fit with todays 25 years organic cooling and beryllium oxide or uranium, fuel element production services anniversary. The reason is simple: We are aluminium oxide for cooling channels. and securing industrial participation in celebrating the silver wedding of Norway IF A and our wi/.x.ard engineer Odd Dahl Norway. Norsk Hydro, the heavy water and the countries participating in the settled in 1954 on a form of aluminium producing company in Norway, became OECD Halden Rcactor Project, but the foil around the cooling channels. It was our basic economic supporter. Saugbrug- Halden Rcactor is not a legitimate child called the NUPOI1 (natural uranium foreningen, the paper and pulp factory of that union. It was conceived and power only pile). But the Dutch wizzard in Halden, at the time considering nu- practically born before the wedding. Dr. Wendt then produced a series of clear energy for process steam supply, Maybe a few short words about the more exotic proposals like »slurry» provided the site. On the 4th November, prehistory are appropriate: reactors and »suspension» reactors. This the last day, we delivered the complete The Institute for Atomic Energy (IFA) resulted in a year of discussions and package to the minister of industry, and in Norway had put its first experimental indecision in the Netherlands. the government decided the same day to heavy water reactor in operation in 1951. In the meantime came the U.S-A.'s place it before parliament. Earlier that year, the Senior Dutch astro- Atoms for Peace programme in 1955,and Considering the unending bureaucracy phycisist H. Kramers had combined the Dutch decided to use the opportunity involved in such nuclear decisions one efforts with his younger astrophysicist to buy a ready made American testing may today look upon this with pleasure friend in Norway to form JENER (Joint reactor. The Norwegians were not inter- or horror according to ones belief in Establishment for Nuclear Energy Re- ested in this, but highly interested in nuclear power. However, the speed of search. Dr. J.A. Goedkoop, who happens another new development: the demon- the approval process initiated a concer- to be the chairman of the Halden Board stration of the stability of boiling light ted effort by all the opposing quarters,
1 and only after two months of public protests and newspaper discussions, mainly from university quarters, and a public discussion with 500 people in the Norwegian Society of' Engineers, the plan was approved by parliament. The interesting point, when looking back, is that safety was not an issue. It was the question of distribution of funds between university research and i n d LI s t ri a I d e ve I o p me n t. We had already started blazing the big hole in the mountain in Maiden before the final permission from the parliament. We could do that because we had esta- blished IFA as an independent non-profit organization with its own capital, which wc could risk even if the final answer should be 110. In that case, however, we would have been sitting with a big hole in the mountain and no money, not a very attractive situation. Now we could turn Lo the technical problems. The first pressing one was: Would we dare to construct a flat lid on top of the pressure vessel? Some of the engineers with a good university training said: j\'o, it must be curved for safety reasons! However, our basic idea was to build a reactor where you could change the fuel elements easily without unscrew- ing the lid. Only a flat top with holes in il could achieve that. Odd Dahl was con- vinced that a thick flat lid with an alumi- nium packing against the tank flange would work. Others were convinced it would curve, leak or blow off. The de- cision had to be made to meed the dead- line for the tank order. And it was made: Flat lid! Today we realise this is the reason why the Halden reactor has been Odd Dahl giving instructions during the installation of the reactor vessel lid (and is) a success. At this time, the young generation with Nils Hidle in the lead began to show began to realise more clearly the size of jointly by a group of interested European that they had gone to a good school in the project, and particularly the running nations. The Council approved the idea. IFA. They look over more and more the budget and specialist staff that would be Everybody among us knew that achie- final completion of the reactor, enthusi- necessary in order to exploit ihe instru- ving agreement within half a years time astically assisted and »reigned» by Ragnar ment to its full capacity when it was between ten lo twelve nations, their re- Strand, the resourceful technical disciple ready. search institutions, their administrative of Odd Dahl. »We can have the reactor When the question therefore finally organs, their governments and finally running in the spring of 1959» was had to be faced whether we should parti- parliaments, would be no straightforward Hidle's estimate. cipate in the OEEC reactor programmes, task. However, we were lucky to have During 1958 a new development had wc were aware of the fact that our capa- two main players on our teams who begun to colour the activities around the city in Norway was not sufficient for steered the next months negotiations Halden Reactor Project. Already in 1957, working with several reactor develop- and avoided the many pitfalls that lay the nuclear agency of OEEC (later OECD) ments simultaneously. On the other along the path, Pierre Huct in OEEC had worked out a suggestion for com- hand there was a strong political pressure and Jens Chr. Hauge in Norway. Hnuge mon European nuclear energy projects. by our government to support OEEC. was Norway's member of the Steering Under the preliminary talks about what The obvious solution was to get the Committee in OEEC and very highly kind of project would be appropriate, OEEC nations to cooperate with us in regarded from before in western Euro- the French pioneer Lew Kowarski had the Halden reactor. pean countries. These two actors carried mentioned the possibility of turning the After some lively discussions inside out an incredible efficient round of nego- progressing Halden reactor into a com- IFA we produced a memo to our Atomic tiations during the last part of 1957 and mon venture for Europe. At the time we Energy Council with a proposal that we the beginning of 1958 and at a meeting took a rather cool view of such a sugges- offer the nearly finished Halden reactor of. the steering committee in Paris 22nd tion in IFA, feeling we were far ahead in to the OEEC without compensation on May 1958 there was on the table a pro- this purely national project which we the condition that the running budgets posal for a common financing of the felt wc owned ourselves. During 1958 we for the next three years be supplied Halden Project for three years amounting
2 to 3.66 Mill, dollars. Norway should con- tribute SI Mill., the six Euratorn coun- tries SI Mill., England 5660,000,Sweden and Switzerland §350,000 each, Denmark and Austria $150,000 each. Each country should sign an agree- ment with Norway who would continue to own and have the responsibility for the reactor. Norway invited all participants to a signature meeting in Oslo on 11th June. The Halden Committee and the Halden Working Group were then esta- blished and given the authority to decide about the future programmes to be carried out in Halden. The last, and undoubtedly the most important problem to overcome, was LO finish the reactor and make it work. Thcchnically wc had few difficulties, but there were other problems: Even in Norway, where decisions were taken with breathtaking speed a couple of years before, a new phenomenon had entered the scene: the concentration upon the magic word »the maximum possible accident;). We had so far no atomic energy act in Norway, and the proposals for Concentration in the HBWR control room commiLtees, bureaus and organizations during the initial criticality experiments. Behind that had to be consulted to permit the the control desk from the left: Viking Oluer starting up of the reactor were endless Eriksen, Henri It Ager-Hanssen, Cunnar Randers and Heinz Braun. Sitting at the dash: Nils and growing. A safety report was prepared I! y dell. as was usual in countries with atomic energy laws. And it was commented Shown in facsimile the entry in the HBWR ope- upon and critizised by all proposed in- rator's logbook on 29th June 1959, as the stances, and new ones demanded to come world's first (and only) boiling heavy water into the picture. When, after a year of reactor went critical for the first time. this, the reactor start-up day approached, and the approvals were called for, the whole picture suddenly changed 180 A q Tfc. degrees around. Nobody wanted to sign the approval and nobody considered ^KXy. /da themselves competent to be responsible for permitting start-up. In the end the ministry of justice decided that it would /7m! be up to the Institute itself to take the responsibility for the start-up, exactly as,we had always planned and expected. QgirtCR/- try And on 29th of June 1959, about M three years after the decision to build the reactor.it was made critical. A formal opening by the King of Norway followed. frcj A celebration that included the whole of )7QD — Very . the city of Halden with school-free children lining the street with Norwegian ho flags. In front of the mountain entrance /t'/THC of the Reactor twelve European flags 9'ccttc - marked Norway's First European tech- nical project. During the 24 years which have passed since that day the group of engineers and / scientists in Halden has turned out a con- tinuous flow of new and valuable infor- mation about properties and behaviour of water cooled reactors. While the 'l reactor originally was conceived as a first experimental heavy water boiler, already ho? (hrO the decision to install the flat lid with easily changeable fuel elements paved
3 The king of Norway, Olav I', and the first From the HBWR opening ceremony at the entrance of the tunnel leading to the reactor hall. Project manager, O.H. Kasa, in the reactor hall during the start-up arrangements. the way for the varied studies of interest great achievement during the 25 years. to all types of water cooled reactors. So many names would have to be in- The first core consisted of 7 tons of cluded and some would undeservedly be natural uranium canned in aluminium. left out. Most of them, however, are The heavy water charge was 16 tons. present at our 25 years celebration, it The operation was limited to 6 MW and is only to look around. Water reactors as the temperature to 150°C. With the designed on the basis of all these years sccond charge of two tons of 1.5% en- of study have proven themselves both riched uranium, the full design capacity economically superior to other gene- of the reactor, about 20 iVHV and over rally available energy producers, and 200°C could be achieved. This came completely safe even when carelessly after a couple of years, at which time it operated. When the present widespread was clear that the real great future of the panic due to misinformation and lack of reactor lay in its versatility and ability to information subsides, and nuclear re- run with a great number of different fuels actors take over the main load all over at the same time, and the possibility of the world, contribution from the quarter inserting all kinds of in-core instrumen- century pioneering work of the OECD tation to study the mysteries unknown Halden Reactor Project will stand out at the time, like dynamics of a boiling as one of the main pillars of the deve- lopment. reactor, void effects, flux distribution, water speeds, pellet/cladding interaction, temperature distribution and so on. All these studies demanded design of new in- core instruments, and even more impor- tant, complete data control of all para- meters and of reactor operations. The complete data automated control of reactors and a set of new instruments came out of this, and instead of dying from old age, the reactor improved its abilities both bodily and spiritually. it would be impossible to mention by- name all those who contrituted to this
4 Organization of the Project
AGREEMENTS
The initial participants in the Project A fourth agreement for the period ment period, 1979 - 1981, Central were: January 1967 to December 1969 was Electricity Generating Board of U.K. • Institutt for Atomenergi, Norway entered into by the same Signatories, in joined as new Signatory member on be- • The Federal Chancellery of the addition the Austrian Studiengesellschaft half of a British industry and research Republic of Austria fur Atomenergie and Comitato Nazionale group. United States Nuclear Regulatory • The Danish Atomic Energy Com- per l'Energia Nucleare changed their Commission changed its status from mission membership from associated-party to associated party to Signatory member • The Euratom Commission (repre- Signatory status while Japan Atomic whereas ElcctricPowerResearch Institute, senting Belgium, France, Germany, Energy Research Institute joined as a Inc. continued its participation as an Italy, Luxembourg, and the Nether- new Signatory. The programme con- associated member together with lands) ducted through this period included fuel General Electric Co. and Combustion • Aktiebolaget Atomenergi, Sweden performance testing, high fuel heat Engineering, Inc. Austria terminated its • The Government of the Swiss Con- rating studies, in-core instrument de- participation in the Project at the end of federation velopment, thorium physics and fuel the agreement period. • The United Kingdom Atomic Energy studies and on-line computer control During the ninth, 1982 - 1984, and Authority. studies. the tenth, 1985 - 1987, agreement Soon afterwards the .Finnish Atomic A fifth agreement period covering period, the participation in the Project Energy Commission and the United January 1970 to December 1972 was remained unchanged. States Atomic Energy Commission were entered into by the same Signatories ex- The Signatories to the present and associated to the Project. cept for the Swiss and U.K. participants. eleventh agreement period, 1988 - 1990, The joint research and development The German participation was changed are the following: programme of the first three year agree- to include a German group comprising • The Norwegian Institutt for Energi- ment entered into in 195S comprised, in KFA Jiilich, Siemens, AEG, NUKEM teknikk addition to the final assembly and com- and GKSS working in agreement with • Riso National Laboratory, Denmark missioning work on the reactor, low the German Ministry of Scientific • The Finnish Ministry of Trade and power and power physics experiments Research. General Electric Co. and Industry with the first fuel charge. The first agree- Nuclear Fuel Services, Inc., U.S.A. • Gesellschaft 1'Ur Reaktorsicherheit, re- ment period was prolonged with one-and- joined the Project as associated members presenting a German group, of compa- a-half years and included reactor physics while USAEC discontinued their asso- nies, working in agreement with the and dynamics experiments with the se- ciated party membership. The programme German Federal Ministry for Research cond fuel charge. comprised development of integrated and Technology A second agreement was entered into computer-based control systems, fuel by the same participants, and in addition performance studies, irradiation of test • The Italian Comitato Nazionale per the Atomic Energy of Canada Limited fuel assemblies for participating organi- 1'Energia Nucleare e Alternativa (AECL) as associated party for one-and-a- zations, and development of fuel perfor- • The Japan Atomic Energy Research half years. This agreement covered the mance models. Institute period from January 1963 through June The sixth agreement period covered • The Swedish Nuclear Power Inspecto- 1964. The agreed programme was de- the three year period 1973 - 1975. rate voted to the completion of the dynamics Austria left the Project and the Signa- • Central Electricity Generating Board experiments with the second fuel charge, tories to this agreement were now from representing a group of Nuclear a programme of water chemistry and de- Norway, Sweden, Italy, Japan, the Research and Industry organizations velopment of in-core instrumentation. Netherlands, Finland and Denmark. The in the United Kingdom A third agreement on the OECD German group of participants was • United States Nuclear Regulatory Maiden Reactor Project covering the further expanded, it being left to KFA Commission period from January 1964 through De- Julich to act as the only formal German joined by the associated party members: cember 1966 Was entered into by the Signatory organization on behalf of the above mentioned organizations in group. In addition the Electric Power • Combustion Engineering, Inc., U.S.A. Norway, Denmark, Finland, Sweden, Research Institute, Inc. of U.S.A. • Electric Power Research Institute, Switzerland and United Kingdom, while bccame a Signatory. USAEC and Com- Inc., U.S.A. the participation of EURATOM on be- bustion Engineering, Inc., U.S.A. joined • General Electric Co., U.S.A. half of the six European community as associated members together with • N.V. Tot Keuring van Elektrotcch- countries was terminated, but partly re- CBTN of Brazil. The programme items nische Materialen (KEMA). The placed by participation of Reactor Cen- during this period included fuel per- Netherlands. trum Nederland and a German group formance models, studies of integrated consisting of Siemens, AEG and NUKEM computerbased control and supervision The aim of the programme is to im- in agreement with the German Ministry systems and irradiation of test fuel for prove the operational safety and relia- of Science, as Signatories, and Comitato participating organizations. bility of water cooled reactor systems Nazionale per l'Energia Nucleare, Italy as Brazil terminated its association with and covers two main research areas: fuel associated party. USAEC and SGAE, the Project at the end of the sixth agree- performance and safety, and process Austria continued their participation as ment period, while the Austrian Studien- supervision and control. associated parties. The programme was gesellschaft fur Atomenergie, later The The joint programme gudgct for the devoted to fuel testing, water chemistry Austrian Forschungszentrum Seibersdorf 1988 - 1990 period is 210 million Norw. and corrosion, stability and burn-up rejoined the Project during the seventh kroner. In addition, contract work for physics studies, and in-core instrument agreement period, 1976 - 1978. The participating organizations amounts to development. Irradiation experiments on United States Nuclear Regulatory Com- about 110 million Norw. kroner. behalf of participating organizations mission replaced the USAEC were also started during this period. At the beginning of the eighth agree- - Participating ' ' OrganizationSK ltu''*?!' t, s Vr Sponsorship of the % Institutt for energi- . O.E.C.D. *Joint Programme Nuclear Energy teknikk, Norway Agency Owner of Facilities Creator of the Board Management Legal Liability Joint Project Approval of Yearly Safety Responsibility General Assistance Programmes and Budgets ;••• End Responsibility for Pre- and Consultation paration and Implemen- Programme Group tation of Programmes Formulation of Programmes s andTechnica! Supervision .<
Project Management Execution of Programmes
Computerised Fuel Research Man-Machine Reactor Operations Communication
Organization of the Project.
STEERING BODIES STAFF
Under the Halden agreement an inter- The Project staff totals 204 of which national committee, known as the about 65 are graduate technical or scien- Halden Board of Management, approves tific staff, and the remainder technicians the research and experimental program- and other supporting staff. me. This is prepared and executed by a In the back of the publication a listing team of scientists and engineers, em- is given of the members of the Halden ployed by the Norwegian Institute and Board of Management and of the Halden drawn from the various participating Programme Group, and of all the staff countries, operating under the supervi- who have worked with the Project for sion of an international technical group more than two years through these 30 known as the Halden Programme Group. yeais, together with an organisation chart of the present Project staff. Research and Development Review
The original objectives of the Halden Project were to demonstrate the feasi- bility of the boiling heavy water reactor and to determine the important charac- teristics of such plants, including reactor physics parameters, dynamic behaviour and thermal properties. Reactor Physics Experiments have been performed with three different fuel charges to determine the important physics parameters. Special emphasis was placed on measurements of the reactivity coefficients associated with temperature and void, and of control rod worth. For the second fuel charge the im- portant physics parameters were also measured at the end of the core lifetime, and the bum-up of the fuel assemblies was determined. The initial third fuel core loading in- cluded seven centrally located thorium oxide assemblies, which were used to obtain information on the physics and fuel behaviour of ThOg fuel under con- ditions representative of large D^O reac- tors. The subsequent third charge core loadings have varied considerably accord- ing to the requirements of the test fuel programme, and physics experiments and calculations have been restricted to that necessary for determining test fuel parameters and important operating characteristics. Dynamics Experiments with the first and second fuel charges produced detailed characterization of the plant dynamics. The high frequency instabilities typical in boiling light Water reactors were not seen in the HBWR. However, at higher power levels an instability of very low frequency caused by steam void forma- tion (flashing) in the reflector and in the moderator was found. It could easily be counteracted through nuclear or pres- sure feedback control, and subcooling the reflector/moderator was also found to eliminate this instability.
Developing process control
Conducting experiments
Studying fuel performance 1805 When in-core instruments became available from 1962, direct measurement of fuel channel thermal and hydraulic parameters produced interesting infor- mation. Experiments in an electrically heated boiling test loop revealed serious dis- crepancies in comparison to in-pile re- sults, particularly regarding stability limits. This underlines the importance of being able to study such phenomena in- pile through use of special instruments, rather than relying on out-of-pile experi- ments. Water Chemistry Phenomena were Developing reactor plant dynamics models. studied in 1963 - 1965, and useful infor- mation was found on the radiolytic de- composition of the heavy water, on use of ammonia for alkalinity control, and on the distribution of deuterium and nitrogen between the water and steam phase. Radiolysis of heavy water is ob- viously an important problem for any heavy water reactor, while the presence of the highly gamma active N-16 in the steam will dictate shielding requirements for heat exchangers and steam lines. Distribution of iodine between steam and water was determined when the intensified fuel test programme led to an increased number of fuel failures with consequent release of fission products to the moderator. A test series was also car- ried out to determine the corrosion be- haviour of mild steel in the reactor pri- mary circuit. Performing reactor physics experiments using the first reactor simulator. Studying water chemistry problems. in-Core instrument Applications for measuring fuel channel hydro- and thermodynamic properties was success- ful, and iL was soon demonstrated that instruments capable of measuring fuel rod parameters necessary for understand- ing fuel rod performance could be made. As the reactor proved to be .well suited for this type of research work, an inter- active development activity was triggered ofr in 1964 leading to more and more advanced instruments and irradiation rigs, also including special in-corc high pressure flasks for studying fuel behaviour under modern light water reactor conditions. Fuel performance studies which was initiated by the success of the in-core instruments, have steadily expanded both in scope and detail hand in hand with the development of in-corc irradi- ation experimental technique. The overall aim of the fuel research has been to in- vestigate in pile how various parameters influcncc the integrity of a fuel rod. The studies have comprized fuel deformation characteristics, heat transport phenomena, fiction product release, and cladding de- fection mechanisms under steady and transient power and coolant flow con- ditions at different levels of fuel burnup. Process Computer Application investi- gations were initiated in 1967 through
8 [IlIIE
The UBW'R reactor is located inside a mountain across the street from the paper and puip factory which uses the steam produced by the reactor in their process the installation of a large process com- Operational Experience with the The control rods are overhauled at puter for the purpose of applying Halden Reactor has been very satisfac- -intervals, their reactivity worth are still modern control theories to control and tory. During the first charge operation adequate. Nuclear instrumentation has supervision problems in the reactor plant. period 1959 - 1960 the reactor power been renewed and modified according to During these experiments a need was was limited to 5 MW. After larger stops modern requirements. Preventive main- foreseen for improved operator - process in 1960 and 1961 with plant installation tenance keeps the process instrumenta- or man-machine communication as well work to increase heat removal capacity, tion in satisfactory condition. Instru- as to increased computer system relia- the reactor has operated at power levels mentation for experimental equipment bility. Already in August 1973 the Pro- between 10 and 20 MW depending on is of new design and microprocessors ject carried out the first experimental the test fuel inventory and irradiation have been introduced together with operation of the HBWR with a computer requirements set by fuel research pro- extensive use of process computers. operated colour CRT based control grammes. During the second charge The loss of heavy water has been kept room, called the OPCOM system. At the operation period 1962 - 1966 the re- at a low level, considering the experi- same time the Project conducted experi- actor was brought up to its maximum mental character and operation of the mental verification of the .Decentralized, power level of 25 MW. Regular deli- plant. The problems of radioactivity .Modular, Process computer system veries of process steam to the adjacent with respect to plant maintenance have DEMP. The characteristics of the DEMP paper and pulp factory was started in been few and it has been possible to system were high reliability with soft 1964- keep radiation exposures below accepted degradation. In addition to regular maintenance working limits without deviating from New and improved display systems of the plant and its instrumentation, re- normal working procedures. and computer structures were developed placement and renewal of old compo- and tested out both against the HBWR nents and instruments have continuously and reactor simulators. The most fami- taken place. All components under Operating Statistics 1959 -1982 liar are the Nordcom and NCT display pressure are at regular intervals inspected systems and the RESS and CONSUP and approved by the competent autho- Operating time 76,000 hours computer structures. rities. The reactor vessel itself is in very Integrated power 900,000 iMWh satisfactory condition, with no signs Steam delivery 840,000 tons Reliability and Safety of Nuclear Plants of corrosion, and the irradiation em- Heavy water loss ave.0.1 kg/h at pressure is the guiding theme for the fuel perfor- brittlement is not a limiting factor. total 11,000 kg mance and computer application studies, Ultrasonic inspection of welds, nozzles Burn-up 1st charge 15 MWd/t and advanced equipment is available for and bolts indicates no changes in the 2nd charge 10,000 MWd/t this purpose, together with an exper- integrity, and the main gasket has never 3rd charge 10 - 15,000 MWd/t ienced and skilled staff. indicated any leakage. Fuel Performance Studies
Calendar Year
Tyjj c of Invesl iga I io lis 70 71 72 73 74 5 76 77 7S 79 SO 81 S2 83-87
Instrumented fuel performance tests
Diameter rigs. Comparative testing
Power ramp testing He-3 local power control
Densification/swelling tests
LWR diameter rig (rotatable rods)
Power cycling tests. Solid neutron absorbers
BWR/PWR simulations
Gas flow/fission product release tests
LOCA tests. Thermal behaviour
Load follow operation
Gap meter rig. In-core TC connecting plug
BWR/PWR ramping rigs
LOCA tests. Ballooning behaviour
Interim inspection facility
Re-configurated reactor core
Fuel rod testing activity in the Halden Boiling Water Reactor (HBWR)
BACKGROUND
The first attempt to install instruments placed on long term testing of proto- ding. The introduction of the in-core rod in a fuel assembly in the Halden Reactor type fuel and performance studies during profilometer this year was a major step was made in 1963 in support of core dy- irradiation at high fuel heat ratings. forward in fuel rod instrumentation, and namics studies. Instrumentation of indi- As from 1970 the emphasis of the greatly improved the possibilities for vidual fuel rods followed already the fuel research programmes gradually- detailed studies of fuel cladding defor- next year with installation of rod gas shifted to studies of mechanical inter- mations during power operation. Conti- pressure sensors and fuel centre thermo- action between the fuel and the clad- nuous improvements of instruments, ir- couples together with instruments for determination of the power produced in the fuel assembly. In 1965 the first 1000 differential transformer extensometer was successfully applied for measure- ments of fuel rod dimensional changes. The satisfactory performance of these instruments demonstrated the potentials for greatly improving the understanding of reactor fuel behaviour through sys- tematic measurements on instrumented fuel rods during power operation. This led in 1967 to the formulation of the first research programme specifically 1961 1966 1968 1970 1972 1971 1976 1978 1980 1982 aimed at investigations of fuel rod per- formance. The main emphasis was Fuel rod testing activity in the Halden Boiling Water Reactor (HB WR)
10 radiation rig designs, and experimental ductance and fuel centre temperatures technology led to steadily more detailed have bt«:n obtained from a large number and refined observations of fuel rod be- of test irradiations. Fuel centre tempera- haviour under various opeiational con- tures show an almost linear relation to ditions. power and increase significantly with in- During the years since 1963, more creasing gap for a given fuel power den- than 240 instrumented fuel assemblies sity. Typically at 30 kW/m a fuel rod with and test rigs from more than 20 different a 350 pun gap (diametric) between fuel research and industry organizations have and cladding, operates at 400°C higher been irradiated in the I-Ialdcn Reactor. temperature than a 50 (im gap rod. Gap The in-core measurement capability at conductance increases exponentially the Project has been further expanded and with linear power and assumes higher includes specialized rigs and loop systems values than computed assuming pure gas where fuel can be tested under simulated conduction across uniform annular gap. light water conditions, and lately also This is ascribed to hourglass shaping of with prototypical LVVR enrichments. pellets, eccentric position of pellets inside An inspection facility built in the Fuel rod cross section with a central oxide the cladding, irreversible relocation of reactor hall is being extensively used for tungsten/rhenium thermocouple. High power fuel fragments etc. Predictions can be rating has caused structural changes in the cen- interim rod examinations (profilomctry, brought into agreement with measure- rod length and pressure, eddy-current tral part of the UO^ pellet fuel. Extensive fuel cracking is observed. ments by either: including a gap heat etc.), which represent a very valuable transfer rale enhancemence factor (com- supplement to the in-reactor measure- account for in thermal codes. The higher monly called contact area function), or ments. operating temperatures in LWR's in com- assuming a reduced as-fabricated gap The experience from the previous fuel parison to Haldcn rods at comparable width by a certain percentage, and re- research programmes and the in-core heat ratings are a consequence of the taining an annular gap configuration. measurement capabilities developed aL higher coolant temperatures and lower Temperature measurements in rods the Project form the basis for the con- fuel enrichments used in LWR's. pre-filled either with xenon gas, various tinuing fuel performance studies. With Neutron calculations have been per- xenon-helium mixtures, or helium alone the present inventory of versatile, re- formed, and post irradiation measure- have clarified the effects of mixed-gas usable, ancl high quality experimental ments have been analysed to characterize thermal conductivity on rod heat trans- rigs and systems, it is anticipated that the radial power distribution in high en- fer. The data show that gap thermal re- the HBWR will continue to be in the richment rods in Halden. The available sistance increases with increasing frac- forefront of experimental fuel behaviour information indicates that the Bessel tions of xenon and at a rate consistent research for many years to come. functions describe the actual radial dis- with theoretical predictions. At 30 - tribution in HBWR rods very well at 40 kW/m a 230 /xm gap rod filled with beginning of life, and fairly well up to xenon will operate at 500-600°C higher considerable burn-up. FUEL ROD THERMAL BEHAVIOUR centre temperature than a helium filled Data on the effects of gap width and rod. Temperature-power curves for typi- The temperature distribution in light gas composition on gap thermal con- cal xenon filled rods tend to bend over water reactor (LWR) fuels governs fuel at higher heat rates, as opposed to the differential expansion, influences fission straight relationships obtained in helium product migration and release and deter- rods. mines fuel stored energy. A proper under- Data from the gas flow rigs show that standing and accurate characterization of fill gas pressurization slightly improves fuel thermal response is of importance 1500 the heat transport across the gap, but for evaluation of fuel rod performance both under normal and off-normal opera- o 1600 ting conditions. o
Thermocouples in the fuel provide the | 1400 CO bulk of the data base on steady state and QJ transient thermal behaviour generated at E 1200
3 -60 of fractions, fuel densification and swell- E ing, plastic and creep deformation of fuel aV E -80 and cladding, irradiation-induced changes a) h- in radial power distributions and the -100 build-up of fission gas pressure in the in- During irradiation 10.5 MWd/kgUOj terior of fuel rods. 2 3 4 5 6 8 10 15 20 30 50 100 High density, thermally stable UOp Gas pressure, bar pellet fuel exhibits stable temperature Fuel rod heal transfer is improved by filling behaviour early-in-life, while marked tem- gas pressurization. The sensitivity of fuel tem- perature rises are noticed in fuels prone perature to helium gas pressure can be applied to densification. Simultaneous measure- to monitor changes in fuel state (amount of ments by fuel stack extensometers and fragmentation) during irradiation . The figure central oxide thermocouples have shown shojus temperature drop at fuel centreline at Cladding diameter profile that temperature increases in such fuel two bum-up levels as function of filling gas pressure. .-In increase by a factor of 3 is ob- can be reasonably well accounted for by served, indicating extensive fuel cracking. assuming isotropic densification-induced pellet shrinkage (giving gap increase). The the effect saturates (in fresh fuel) above amount of fuel densification and the magnitude of the rise of centre tempe- about 10 bars. The effect is attributed Measured Calculated to a slight increase in helium conducti- rature increase with increasing opera- vity (5-10%) and a reduction in tempe- ting temperature. Moreover, helium- filled rods showed more stable tempe- The temperature distribution through the fuel rature jump distance (boundary pheno- causes thermal expansion induced ridges in menon) values with pressure (inversely ratures than xenon-filled rods, because the latter types of rods are more sensi- the cladding at pellet interfaces. The figure proportional) in light noble gases. shows an in-core recording of the resulting tive to gap variations. In medium and Concentrically positioned pellets in- diameter trace compared to a profile determined large gap rods densification effects are by means of a model based on a multimesh, side the cladding produce the least compensated for by fuel fragmentation finite element method. favourable geometry for effective gap and outward relocation and the actual heat transfer. The average temperature temperature changes are thus reduced. drop across the gap decreases with in- Pre-pressurization make fuel rods less Small gap rods and pressurized rods creasing degree of eccentricity, in par- vulnerable to Lhermal feed-back effects containing stable fuel tend to show ticular in rods containing low-conduc- since the dilution rate due to fission gas stable temperature behaviour also up to tivity gases. In xenon-environments is inversely proportional to filler gas high bum-ups, since moderate fission gas the fuel centre temperature difference pressure. release will have only a minor influence between concentric and eccentric situ- on gap conductance in such designs. ations can be up Lo 150°C at 30 kW/m, The onset, rate and amount of tempe- whereas in rods prefilled with high con- rature changes in nonprepressurized, ductivity gas such as helium, the effect medium and large gap fuel rods will lar- is appreciably reduced and will in prac- gely depend on the fission gas release tice hardly be distinquishable from sto- properties of the fuel. Temperature in- creases in the range from 200 to 300°C have been observed in Halden implying an appreciable degradation of rod heat Diametral gap; 75 m transfer efficiency with burn-up. Measurements of centreline tempera- 1500 ture drop with fill gas pressure, axial gas flow resistance, hot gap width as 32 MWd/kgU02 well as transient temperature response Fresh fuel show furthermore that thermal resist- u 1000 — ance across the fuel in medium/high bum-up rods is increased due to exten- sive fuel cracking and progressive fuel gap closure with exposure.
500 Analysis of thermocouple data from fuel rods subjected to rapid step up of 10 20 30' '40 power level following long term low Spesific power, kW/m power operation has provided informa- Cap, conductance and fuel temperatures are tion about thermal feed-back effects, 10 20 30 40 50 subject to changes during irradiation, due to caused by transient fission gas release. Specific power, kW/m changes in gap width and mixed-gas thermal Relatively moderate power excursions In small gap rod contact conductance makes conductivity in the rods. The above figure may produce fairly large temperature in- shows measured burn-up dependent fuel tempe- gap heat transfer less sensitive to fill gas conta- creases in un-contaminated rods, only mination (fission gas release) atid fuel tempe- rature changes caused by fission gas release and ratures in such rods tend to remain stable with thermal feedback effects in an non-pressurized partly explained by the rise in power. exposure. rod. Transient fission gas release and the re- 12 A 1 I t —r -» the temperatures decayed in an approxi- 1500 Gap meter rig Fillgas: He mately exponential manner with time •f 2000 - Gap: 150/Lim Calculated temeratures Density: 95% constants characteristic of the rod para- \\ (solid pellet model) meters. Small time constants, within the ° 7.37 MWd/kgUO, 4.42 do. 2 range 6-10 seconds, were as expected o • 7.53 do. o * 7.62 do. * •« o O associated with helium filled rods, while u 1500 rods with larger gaps had longer time con- Before power-—T.- ramp .« stants. In the case of xenon filled rods, „ ^ *O** / time constants in the range 15-27 seconds nO / " . •*" were found and the dependence on gap
1000 size was more pronounced. ^ After v o / In most cases the fuel behaviour data • OQ / power ramp » • o y .. oo /\ tend to favour a »cracked pcllet» model V ° / \ in preference to a »solid pelleU model, ^y Start - up especially at higher burn-ups, where also 500 thermocouple dccalibration effects had to be considered. These transient mea- surements provide valuable data against Hellui which to test various models of fuel rod i i t • • heat transfer. 10 20 30 40 50 60 10 20 30 Specific power, kW/m In the analysis of fuel rod transient Time into transient, The figure shows manured thermal response temperature measurements, it is impor- Calculated transient responses of helium and of an non-pressurized rod power ramped to tant to account for the time delay bet- xenon-filled rods and typical thermocouple 50 frlf'/m at a bum-up of 7.4 MWdfligUO - 2 ween changes in fuel temperature and lime delays in such rods. (The power dropped The power increment caused fission gas re- the corresponding response of the fuel suddenly from 100% to 5% at time zero, mid lease and gross degradation of the heat trans- centreline thermocouples. This delay re- coolant temperature remained at 24*0°C during fer of the rod attributable'to thermal feed- the first 30 seconds.) bach effects. The data show that purl of the sults from the heat capacity of the ther- release occurs relatively fast. mocouple itself, and the resistance to heat transfer between the thermocouple Review of literature data and theore- suiting degradation oT the local gap con- and inner fuel surface which is a function tical calculations formed the basis for a ductance contribute significantly to the of local temperature and rod filler gas specially designed irradiation rig for esta- obsei-vcd temperature increase relative to composition. blishing the rate of decalibration of re- the power. The magnitude of the in- All fuel temperature measurements at fractory metal thermocouples under crease will depend on the state of the Halden have been made with tungsten/ HBWR conditions. At present, data are fuel (burn-up, previous power level, available at burn-ups of 14 iVIWd/kg. The degree of grain boundary gas porosity, rhenium high temperature thermocoup- test is continuing and will be used for ctc.), gap width and power ramp charac- les. During the in-pile service these ther- correcting thermocouple measurements teristics. mocouples may exhibit dccalibration or drift. This results in a change of the vol- in high burn-up fuel rods. The results Lage output of a thermocouple such that confirm that the magnitude of dccalibra- 1400 tion depends primarily upon fluence and U the inferred temperature no longer repre- sents the true temperature. The magni- operating temperatures. o 1 200 tude of the drift will depend among others Sphere-pac £ 1000 on operating temperature and neutron a E flux level and energy spectrum. Annular 8 800 Annular - coated
S BOO -
10 20 30 40 Specific power, kW/m
Centreline temperatures during first power ascension versus LHR for different fuel de- signs. A nnular, annular-coated and reference de- signs have comparable gap width and fuel den- sity. Reduced fuel temperatures mean reduced fission gas release and less mechanical inler- action during power ramping.
Transient heat transfer behaviour has also been evaluated by monitoring ther- mocouple responses during linear decrea- sing power and reactor scram experi- ments. Included in the scram tests were 0 1 0 20 30 4 0 50 rods of different design parameters en- Specific power, kW/m compassing combinations of small and large gaps, helium and xenon fill gas, so- Measured steady state (left) and transient (right) temperature responses of four different fuel lid, annular, vipac (powder) and sphere- rods having achieved different stages of burn-ups. Transient response of fuel rods is dependent pack (spherical particles) fuel at a variety upon parameters affecting the rod thermal resistance i.e. fuel diameter, gap width and gas com- of bum-up levels. Following the scram position. Fuel thermal time constants from 5 to 25 seconds have been measured.
13 "0- FUEL ROD MECHANICAL INTER- edges further reduce the tendency for ACTION circumferential ridge formation. The
/ results also indicated that high density Design innovations and operational .20 - // 2 6 10 h fuel produced larger local and overall procedures that will lower localized or !i deformations than lower density pellet i i V. overall stresses in operating fuel rods * • and vipac fuel, in particular at high heat c /i ••* will contribute to increased fuel relia- \ ratings. S .15 Standard j j fa bility, and rather comprehensive experi- Rod elongation measurements have mental studies have been undertaken for 30f i pW^oc also been used to develop an under- characterization of these effects. u standing on how design parameters, Experimental assemblies providing in- 15 .10 JH 4^o, operational procedures and power histo- pi le data on pellet-cladding mechanical /// ries affect mechanical interaction late in interaction and gap closure behaviour, /' / A-—Graphite lubricated life. A clear relation between ramp elon- include several diameter rigs, gas flow V / Cladding ID 10.8 mm gation and increase in permanent axial Density 96% rigs and a specially designed gap meter strain with bum-up was observed. Per- rig. Cladding and fuel stack extenso- Free thermal expansion manent axial strain of fuel rods arc due metcr measurements have yielded exten- to plastic deformations (during ramps sive information on burnup-dependent early-in-life), cladding creep, and irra- interaction and on densification and 0 10 20 30 40 50 dialion-induced material growth. In Specific power, kW/m swelling rales on a varity of fuel designs. many cases, permanent strains were ln-pile rod axial strain measurements During a fast power increase in a fuel rod the several Limes the values expected from from the initial test series on pellct- fuel will stretch the cladding due to differen- irradiation growth alone. Long rods tial thermal expansion, more with smaller clear- exhibited more interaction than short cladding mechanical interaction effccts ances, less with larger. Lubricating the fuel/ (IFA-118 series) showed as also con- cladding interface will reduce the effect. The rods, and rod containing vipac, annular firmed by post irradiation examina- resulting elastic forces in the cladding will and lubricated fuel showed in general tion that the strain cause the fuel to creep plastically with time. smaller elongations than solid pellet The above measurements were obtained tluring fuel rods. The beneficial effects of using - decreased with increasing gap, a power ramp at 17 (MWd/kgU0 ) using clad- 2 dished pellet were confirmed while the - increased with increasing pellet length, ding extenso meters. influence of the pellet length/diameter - decreased in going from flat to dished ratio was ambiquous in medium and to chamfered pellets, and that the Swelling rates of 0.9% AV/V per 1% burn-up were recorded in high density large gap rods. Cladding stresses were - dependence on wall thickness was fuel, lower density fuel showed reduced found to relax slowly in high density small in the range studied (0.40 mm swelling rates, and in some types of fuel fuel and rods irradiated to appreciable versus 0.75 mm). negligible swelling was observed to burn-up, indicating a diminished role Tests initiated in 1973 for study of relatively high exposure. for fuel dcnsification relative to fuel densification/swelling characteristics of Simultaneous measurements of axial creep. pellet fuel showed that the exposure at and diametral deformations on fuel The influence of graphite on pellets, which decreases in fuel column length rods in the one rod diameter rig loaded siloxane on cladding inside surface, and clue to densification cease and the fuel in 1971 showed that the build-up of of pellet density on cladding deforma- column begins to increase in length due axial stresses in the cladding caused tion behaviour was examined in the to fission product swelling is dependent lateral contraction and influenced onset Project first 3-rod diameter rig. The main on the dcnsification stability of the fuel. and development of radial interaction conclusions were that lubrication had during a power ramp. Observed ridge liltlc effect on development of ridges heights were a factor two larger for and overall diameter changes and that 10 20 T I I 1 i dished pellets as compared to flat pellets. the influence of fuel density was smaller 1 Exposures, MWd/kgU02 This difference is caused mainly by the on the radial than on the axial defor- Closure of small (sub-micron) pores difference in the application of the axial mations. In both sets of tests it was ob- ^ Closure of intermediate forces on the pellet ends and not by served that local diametral strains were | (1-10jim) pores hourglass distortion of the pellet geo- considerably larger than the mean axial metry due to thermal stresses. Later tests strain values derived from elongation have shown that chamfering of pellet measurements, and that the radial de-
1 : Zero power, before ramp 2: 51 kVWm, top of ramp 3: 51 kW/m, 1 5 hours later 4: Zero power, after ramp
Apparent rate varies! depending on dmiification Diameter traces
Schematic representation of low teh\,,.ji?,tdre densification-swelling behaviour. Figure cori' structed based on in-pile fuel stack extenso- meter measurements, indicating fuel shrinkage at low exposure followed by solid fission pro- duct swelling. Dimensional changes due to densi- Diameter profile traces obtained of different stages during a high power cycle on a small gap fication-swelling processes in current fuel are rod containing solid high density, flat-ended and chamfered pellets. Data show deformations controlled by strict control of various fuel caused by pellet/cladding mechanical interaction (PCM/), slight relaxation at ramp top, and fabrication variables (micro structure). some permanent local and overall diameter increase upon power reduction.
14 Data on gap closure mechanisms and fuel relocation rates have been deduced from axial gas flow resistance measure- ments in gas flow rigs installed in 1978 and 1980 and in a special rig loaded in
1981 to determine hot gap width in rods Start - up power romp while they are operating. u 60 Specific rod power, 2 kW/m /t In the gas flow rigs, axial pressure .a /! drop data are used to calculate the hy- S! 50 draulic diameter for gas flow inside the I rods. At beginning of life the hydraulic a! diameter corresponds to the as-fabricated 40 fuel-cladding gap, however, as the fuel Comparison of fuel-cladding mechanical inter- pellets crack and relocate, a decrease in 30 Gap action during power ramping for solid and hollow pellet fuel. The hollow pellets cause less the effective gap size for gas flow (hy- 150 200 250 300 350 400' Deflection,/im PCMI-induced deformations, more relaxation draulic diameter) is observed. Rods with and less overall residual hoop strain than do the gaps in the range 100 jum to 230 Aim solid pellets. showed a permanent decrease of about 35% in gap width during the first power formations at pellet end positions were eyclc. Hydraulic diameter tends to de- about twice those at mid-pellet posi- crease gradually with increase in heat tions. load and exposure at a rate which Ramping of BWR type fuel rods, depends upon initial gap size and burn- having diametral clearances of 60, 150 up. At 17 MWd/kgU02 the gap size at and 225 pm to ratings above 50 kVV zero power decreased to 40% of initial showed marked interaction effects and value. As cladding crtepdown is neglid- permanent deformations only in the gible at HBWR conditions the data im- Deflection,jUm small gap rod. The chamfered pellets plies progressive relocation and gap closure as bum-up increases. A method has been developed to measure hoi did not produce ridges. Elongation and gap width in operating fuel rods. The technique diameter data on this rod at high burn- consists basically of squeezing the cladding dia- ups showed slow gap closure rates at metrically over a pellet length whilst simultane- low heat loads, and the amount of stress ously logging the forces required and the dia- relaxation at higher exposures was signi- meter decreaseOnce contact is established between cladding and fuel, lite slope of the ficantly reduced, consistent with trends force versus diameter plot increases (points I observed in other tests. and 2). The lower curve provides slope changes The difference in behaviour between in detail. solid and hollow pellets has been exa- *(see also chapter »Research Facilities and mined in small gap rods with respect to Equipment»). dimensional response to power ramps. The ramp-imposed diametral strains in manent gap decrease was measured Burnup. MWd/kgU02 the hollow pellet rod were reduced by during the start-up power cycle and a 25-30%, relative to the solid pellet rod, Hydraulic diameter at zero power as a function further reduction as irradiation pro- of burn-up and the least-square-fit lines for the ceeded. The gap change rate versus the relaxation rate was higher, and the data from ttvo different fuel rods having two power decreased with burn-up. Fuel ex- residual local and overall diameter different as-fabricated gaps. In the large gap rod changes after the power cycle were the hydraulic diameter (derived from gas flow pansion and contraction takes place in- smaller. The observed differences in di- measurements) has decreased by 60% and in the side the fuel and does not strain the mensional response can be explained by small gap rod by 35% at 13 MWd/kgUO2- The cladding at high heat rating in spite of gap in the latter rod is completely closed at full, small »relocated» gap. An opposite the lower fuel differential expansion in power and additional reduction is thus preven- trend was registered in a low power rod, annular pellets. ted.
The gap meter rig has provided direct 520 evidence on how initial gap changes with 525 irradiation as a result of fuel cracking, re- 15 location and densification. The gap mea- •CI •OoJ suring method consist, basically of 810 squeezing the cladding diametrally over a pellet length whilst simultaneously log- OS ging the force and the diameter decrease. E When contact is established between ra cladding and fuel, the,slope of the force O O versus diameter increases. The diameter z 76 78 80 82 !• The axial gas flow pressure drop measurements Calendar year are used to derive the hydraulic diameter inside decrease at which this occur is then a fuel rods. Hydraulic diameter determinations measure of the gap. Figure shows the number of diameter rigs can be used to monitor the extent of fuel crack- Measurements of gap width versus loaded into the Halden reactor since 1971. Diameter rigs are used extensively for compa- ing and relocation. Figure shows a permanent heat rating at increasing burn-up in the j gap decrease after first power cycle and that rative testing and combined in-pile measure- ; further cracking continue to occur as the bum- first two rods in the rig indicate a power ments (fuel temperature, deformation, fission up increases. (No cladding creep-down is expec- dependence on the onset and rate of a gas pressure etc.) on conventional and modified ted at HBWR conditions.) gap closure. In a high power rod a per- fuel designs. Linear power, kW/m Fuel rod 1: standard Fuel rod 2: OVJ density, stable 46 50 I Fuel rod 3: low density, unstable Specific power
5 10 15 20 25
Burn-up, MWd/kgU02
This figure shows I he cffect of fuel structure on fission gas release. The standard fuel rod has the lowest build-up of fission gas pressure versus burn-up. The law density unstable fuel showed the highest pressure huild-up, while the low density stable fuel exhibited intermediate release.
FISSION GAS RELEASE AND FUEL
DEFECTION STUDIES ° 2000 - -r- -T— - KWU mi. if.i * YANKt CORE VC , IJI -BEUAI.W ind P" llZM°C.-0 5H. (Si The release of fission products from v.. nM COMt el al I; • HAlGfN Pit DA » 1-1*1 the uranium oxide matrix to the rod free a 1500 volume is related to fuel temperature and structure and may, depending on operating conditions, lead to cnd-of-life 1000 Halden fission gas rod pressures cxcccding failure criteria. release threshold Further, liberation of fission gases im- r 500 Deflection pairs the heat transfer from the fuel, 10 20 30 40 since the conductivity of noble fission < Fuel rod average burnup, MWd/kgUO^ Schematic prescntatio". of in-corc gap measure- gases is much lower than that of helium, In-pile fuel rod pressure data show that in the ments on a fuel rot', during u power ramp in a currently used as filler gas and heat initial phase of irradiation the fuel retains the gap meter rig. The ,i>iip is observed to close transfer medium between the fuel and gaseous fission products. The duration of such progressively with power increase. the cladding. an incubation period depends oil operating tem- perature, The figure shows a comparison of the Information on fission gas release threshold temperature curve for fission gas re- i.e. an increase of hot gap with irradia- from commercial fuels are being gene- lease (derived from 1-laldcn 'od pressure measure- tion created by fuel sintering/shrinkage rated from surveillance programmes ments) with open literature data from rods in and no time-dependent change in rate of which include destructive examination which the release was < 1%. gap closure versus heat load. High power of rods to determine end-of-lifc pressure densities produce large temperature dif- and gas composition. Post irradiation tinuous monitoring of rod pressure allows ferences radially across the fuel, which puncturing has been widely applied to the separation of power and burn-up tend to promote extensive fuel cracking Lest fuels. However, considerable advan- effects. The observed evolution of gas and outward relocation of fuel fragments. tages in relation to mechanistic evalua- release may then be related to particular During the last years an increased in- tions are obtained when direct measure- events in the course of the irradiation, for terest has been noted for the test capa- ments of the fractional release can be example a power increase. bilities offered by 3-rod diameter rigs performed inpile by continuously re- The in-pile investigations carried out with combined thermo/mechanical mea- cording the increase of rod internal in the I-IBWR have systematically shown surements and I-Ie-3 local power control. pressure. This technique has been widely that in the early phase of irradiation the From 1975 to 1977 a total of seven rigs used at the Project in connection with fission product release is negligible, even of this type were loaded into the reactor base load irradiations, power ramp and when operating power is relatively high. for different participating organizations. load following operations, and in chan- The extension of this »incubation» period These rigs have been extensively utilized nel transient experiments. before appreciable release is observed, for detailed investigations related to In order to assess the reliability of the depends strongly on the temperature at ramp performance of standard and modi- in-pile pressure measurements systematic which the fuel operates. Such observa- fied fuel designs by simultaneous mea- comparison with post irradiation data tions have been associated with the abili- surements of dimensional changcs and has been carried out, showing very satis- ty of the fuel to retain the gas atoms dif- time-dependent fission gas release on factory agreement between end-of-life fused from the UOg grains in gas bubbles rods incubated to considerable exposures values from '.lie Lwo types of determina- at the grain boundaries. When sufficient in separate base irradiation rigs. tion. The scatter is typically within gas has precipitated into these bubbles, The main objectives of these investi- O.OD MPa at low pressures or 15% at high the grain boundaries become saturated gations are to improve the basic under- pressures, which implies a tolerable un- and bubble interlinkage takes place, thus standing on a number of potential per- certainty of the resulting fission gas re- allowing the escape of fission gas to the formance limiting aspects and condi- lease fraction. rod voids. At reduced fuel temperatures, tions in relation to e.g. power ramp and The major benefit of the in-pile mea- grain boundary predipitation slows down load follow operation, and to generate surements is that,- since they are per- considerably, leading to large incubation qualified data for development/verifi- formed at high frequency, a correct time burn-ups before saturation of the grain cation of analytical performance codes. and burn-up sequence can be established. boundary with such bubbles is completed. In contrast to post irradiation data, where The analysis of well characterized I-IBWR only the end-of-life status is found, con- experiments, in which the pressure mea- • V* t •
60 LTi Specific power 3 40 4 5! i T-. 0 30
1 20 t w / 10 Fuel rod pressure data
3 4 S £ 10 Burnup, MWd/kgUO,
Development of fission gas pressure in two nar- response to a power increase to a higher power row gap fuel rods. Significant fission gas release level. The data scatter because the measure- takes place at 7 MWd/kgUOn in both rods, in ments ivere made at different power levels. surements were accompanied by direct Unstable fission product in-pile recording of the operating tempe- release in a gas flow rig rature at the centre of the Tuel, indicates that when irradiation is carried out below 1200°C, the release remains low, i.e. of the order of one percent, even at burn- 5 -4 ups well exceeding 20 MWd/kg. At these low temperatures, therefore, it is expec- ted that structural changes in the fuel 10 12 14 are minimal and that the release is small A certain fraction of the unstable isotopes pro- at any bum-up of practical and com- duced by fission will be released from the fuel mercial interest. into the free volume. This fraction, or release- An alternative method of determining to-birth ratio, will increase with irradiation, or release behaviour and fuel structural fuel bum-up, as seen in the figure. It illustrates the changes in the fuel structure caused by dif- changes consists of flushing a fuel rod fusion of gases into the grain boundaries for- with a constant internal flow of gas and ming porosities which gradually interlink, thus measuring the 7-activity of the exhaust forming flow paths to the rod open volume.
gas. Spectral analysis allows the deter- mination of the ratio between the re- lease from the fuel and llie formation (release to birth ratio) of various unstable isotopes. This technique is used in two HBWR experiments involving a total of 6 rods. The release to birth ratios obtained for the unstable isotopes of krypton and xenon exhibit an approximate inverse proportionality to the square root of the isotopic dccay constant in accordance with simple diffusion based models. Further, since determinations arc performed on unstable isotopes, the evo-
7'lie composite photograph shows the section -12 -10 -8 through a fuel rod which has been operated at Intdecay constant} conlinuosly high temperature. The high tempe- Release-to-birth ratio (R/B, fission gas release rature results in "columnar" type grain growth rate to fuel void/fission gas production rale) as and in a gas release fraction from these inner function of the decay constant, X (various iso- regions close to 100%, The gas which remains is topes). Results from measurements at two mainly concentrated in bubbles. The enlarged power levels in one of the Project's gas flow rigs. photograph, from a section near the rod centre, The R/B varies with \fk indicating that diffusion clearly shows the various stages of grain edge is the dominating fission gas release mechanism. bubble interlinkage.
17 1 VJ V) 0 •Xj 3* w S-. EE. 3 — • H a- c 0 3 3' 3* X, "t •f. -I p era' ui' : ca 2 " 2J o £ o 5 ' 33 S n 3. HI c •-n O n Q. £ 3 3 n o" < n H. o 3" n < n c — G. 3 §i nB0m
— E 3 Fission gas release, % o o 3 < 2 =' ^ Specific power, kW/m & a. "e 5. 3 a. « 5' 2 V5 0 5 5a M —V> rr = r* K- °o jt. ^ o 3 S. s 3 o 3.T3 4 = " 5- a d' r; c " °1 9 o & I i i 1 & 3 ^ 3 ° 3 S ™ 3 •<" S "" S c p " r? 3 3 3 tn r- 0 O ~ — oT n. H. c 3. 3 ft tj B K a _ s- o 0) o o ; —, P3* i-n 3" —' ^ n 3 cn .s S ? o e. o 3" 3 V n v1/7 u 3 3- < 3 n rr. c/t *- t _, D s i OQ :!. LA. "I 3* n % P o CD 3 > c CA •8 2 •s S O- S'S p 3 S«•>- 38 ' 1 . ? 2 V) a TO 3 C —. V5 o' s- 3 2; a. - n o • - lo w • lo w fiss . ga s releas e * - fue l a. m - o 2 *=> — E-S- O C
hig h P hig h =" 2. K- S.3 s < Failur e Lowe r 3 3 •a 5/1 IX I a 3 n o • UJ ) i •O " I c r. c 3 n « £ -~ 3„ 0 C3 fue l v. 1 n 3 < D- < fue l fiss . ga s releas e 1 fue l temperature s 1 r-Cfc r> a £ cla d o- p S 5- ^n i/i limi t occurenc e 1 r. a. £ 3 - V, era* 7SJ
temperature s I O- cla d i 0 ~ 3" 3' " I 3 1 a bondin g 1 3 cr.c 0 i " P c/> 0 S 5 S ° fo r fl S W w 3 a ^ bondin g I £ ^ 7 f; «« OQ = -3a n 2 a 0 •o f3 p to n 3 rs , (— . Crq 0 rs £ a 3 S5' 3 S-! £ -1 H 2 3 3 O S g • — c/i r* (/) o a. j a o 3! tf • n> 3 ~ rs 5 C —. 3* ° trt & o O S tsi c/: a E. £ c C* — n Q (A V. « • — 0 3 3'
* P
' J" O s i ST* 3. = 52 n 3 K., n ^ 1 s Crq . sr 5 — I© f-S O »t3 — crC. < n crq sX 9 p CL s fj m cr HI o rs 2 "_i 2 ^ O » o 2 3- r. 3 X 0 rt P- 2. 0 ri v* VI < a a ? o ?D a. a. th* « n « 13 5 7 Crq 3 "U n o ** ^ G. o 'c n ji vi r. r, 2 Z - < s- = 2.1 P r- r. n tr 3 r: —i 3. —• s' _. n — ^ —^ t.r ^ ^ c 3 s n C —i -r £ ~ ^ - 3- •. 3 "" 5' ^ — V. VI 0 ^L p C ^ o' P ' ~ 3 S 3 - O 5 2 <-. n Z " P ? « « p- 3- at r. -i 3 0 ~ o — ^ 5 c P V-. 3n D— ~ C( _ r- 6 s. 2 p <3 Z P 3 i. p 2 X" s. C; • —. —C r: 3 S r." •o oq • — O — n 3 " T ^ V. 2 " ° 5 S § 5.K 5 n r 3 S 0 o ? E i 5' 3" 3" r. q S JT ii era ^ - o o' S c 3 3 5 3 i n tr 3 3 3 " S" =' o -a 5T 3 3? 3 3 0 r, o 3 o P H" ^^ ~ rR. 33 TO
3- B a s o. £ * £ o - 3. c 0 3^. ri "3 3 o n a. a. rr V. 2c < 3: 2 -i 5. 3 C ^ r^ — c r. w —. r. 2 o 3- 3 £ r. v. 3 3- 3 r. p s si < a y- § 1 r. = r: b ri 0 d 2. -» c r, c — < o C ' k 3 —; tr. 3 <-. ? S 2 5 C 3 — cr. n qo. r"''C. p 3P £ ° « ra ~ £ — — CfCJ r. y. fi —i % £ — r. ft -r S w r- 3 3 cr v. -I C3 ,- C » o _. . 3 3 v; 3 3 r: c 3 _3.' 3. fs 3 C u: FT n. t 5% vT Fission gas release, % ? ? 2 g C — H c 0 proven to be an efficient vehicle for con- veying information to the Project parti- \ Failure boundary, small gap rods cipants. and a valid method for interpre- y (24 data point) ting and rationalizing the experimental data. The increased interest of Project a o" o J participants in the use of data from the - HlSWR tests has generated the need of a a og producing and maintaining an efficient o< a o° —13- o orf) data processing and transfer system. An extensive computer-based data acquisi- Ramp-^ |--Ramp tion and retrieval system for IIBWR test fuel data has been recently made available, History of rod 1 01 see special section on this subject. This Closed synbols: failed • data bank management system will be 0 no PIE utilized to provide easily useable and transferable data packages on fuel perfor- 1 estimated error except where mance and assure efficient use of the data shown both at Halden and at Signatory organi- 10 15 25 zations. Burnup, MWd/kgU02 Analytical efforts have been expended in Lhc areas of dntn qualification, provi- In order to produce fuel failure* a minimum power level has to !>•?• exceeded while at the same ding a link between the pure dala collec- time a minimum power increment has to be imposed. Figure shows ramp test results in the tion and the formation of qualified data llaldcn reactor with failure boundary in conventional fast ramps. Minimum power rating to produce fuel failure decreases with bum-up. Damage accumulation effects are explored by sets. In a few instances, such as for stu- performing multiple ramp tests with rods of different designs. dies of the dccalibration of fuel thermo- *caused by stress corrosion cracking. couples, the analyses have justified the need of particular ad hoc experiments. Further, analysis capabilities have proven 1 - Elongation drop of 50/um 80 lo be essential for design and lest feasi- 2 - Failure Indication on activity monitor bility studies, especially when concep- E Power __ , 45kW/m / i tually new types of experiments are be- 60 ing proposed. This has been the case for ~
19 -F- HEAT FLUX u 800 -H- HEAT TRANS. COEF. -T- TEMPERATURE 600 Band including 5 thermocouple readings at same level
I 200
012345678 Time into transients, min
Controlled cladding temperature transient in the flow starvation rig. This type of transients are produced to induce ballooning (local spherical expansion of the cladding) in the rods at the same elevation and determine flow blockage in a bundle of pressurized fuel rods after a loss-
REFILL V. r of-roolant accident (LOCA). p——- REFLOOD tirae-. seponds 7$ loi 12^ isi 17^ zoA ing a centre rod. Testing consists of a rapid heat-up to cladding temperatures Typical results obtained during blow-down, heat-up and reflood phase of an experiment performed in the flow starvation rig to compare quench behaviour of nuclear and electrically heated rod approaching 750°C followed by a pro- bundles. Temperatures were measured on the cladding inner surface. Also shown are the heat flux longed period ( ~ 400 s) during which and the heal transfer coefficient at cladding-coolant interface. the temperature is maintained between 750 and 800°C. The rigs carry a number into the Hist series of reference experi- of thermocouples both on cladding and ments. These were conducted on a nu- 1000 ;Rodswith at various location in the channels to pro- V electric heaters clear fuel rod bundle operated at diffe- "A V perly characterize the thermal response rent power levels in a loop capable of of the rods and of the subchannels with- producing coolant blowdown and estab- 800 X in the test section. lishing reflood at desired rate and at a In order lo ensure that coplanar bal- Furc I, rod. ;1i \ i prefixed time into the transient. The fuel looning will lie .icliie-,uii tin .ill rods, a 1 600 : I rods were equipped with cladding thermo- / ^^ / Reflood preliminary thermal-hydraulic experi- couples mounted at the inner surfucc of x L ment has been conducted wiLh the objec- the cladding to avoid disturbances of the O 400 . 1 . 100 200 300 tive of verifying that isothermal condi- quench front during coolant reflood. A Time from coolant flow shut off, s tions could be attained in ihe peak tem- total of 36 experiments yielded high Comparative testing of nuclear rods and SEMI- perature zone and that the temperature value data on the separate effects of rod SCALE electric heaters. The results show that, transient could be controlled by regula- power, reflood rate, time Lo rcflood and due to the absence of fuel-to-cladding gap in ting the inlet flow rate of the rellood rod temperature on Lhc quench behaviour the solid SEMISCALE heaters, quenching is water. This Lest has successfully indicated of the cladding. These tests formed the appreciably delayed in these types of electric that coplanar (at the same level) balloon- heater bundles. terms of reference for the subsequent ing can be produced, since the in-rod and testing with two clecLrically heated rod rod-to-rod cladding temperature differ- bundles, one carrying solid (i.e., no hea- simulator bundle exhibited quench limes ences were satisfactory. At the same ter-cladding gap) SEMISCALE heaters, comparable with ihe nuclear bundle. time, this test was very useful for estab- the other with REBEKA heaters in which The series of tests presently conducted lishing a proven operation sequence and heater fuel - cladding gap is adequately in the flow starvation facility of the has provided a base for improving details simulated. The comparative testing show- HBWR relate to cladding ballooning be- of Lhc Lest rig design. ed that the absence of gap in the SEMI- haviour in pre-pressurized fuel rods with SCALE heaters induces a delay in clad- the objective to compare the behaviour ding quenching, whereas the REBEKA of electrical simulators, commonly used also in ballooning experiments, with actual nuclear fuel rods under identical experimental conditions. Ballooning re- fers to large radial cladding expansions due to internal gas pressure and tempe- rature induced increase in ductility of the cladding material. The objective is to establish whether cladding interaction effects arising from ballooning of nuclear heated rods are adequately simulated when deformation is a consequence of electrical heating, the concern being that "0 50 100 150 200 the rigid heaters may impose smaller Time from coolant flow shut off, s cladding deformations and smaller flow Quenching behaviour at various reflood rates. blockage than nuclear fuel rods. Identi- The experiments were conducted on a fuel rod cal operational sequences are used to test bundle and used as reference for comparative nuclear rods and REBEKA simulators in testing with REBEKA and SEMISCALE type identical clusLers of four rods surround- electric heaters.
20 Computer Control Research
Calendar Year
Survey of Activities 59 61 63 65 Plant Control * Steady State Control\ i. Load; F o 11 owing Control Core Control. Flux Distribution5;* HBWR/r^aS 11 Power Di tribution ControCm LWR'sl ' SCORPIO,;) Operator-Process Communication * r HBWR Operator Console OPCOM Project^ ; ^V,' < Computer based Control Room(/)>» ' , HAMMLAB^-" f WtK, Computer Structure jDEMR,Projects' , - ^.J'-- ^ & f1 , ' Computer Structure Developments v HAMMWARS; System Reliability , J. R ESS Project, 5 - Software Reliability'^" , l, t, ^/Safe^SysterffSurveillancepj » 6 ^eqif^^ethodOu^ Plant Analysis ;vDisturb"ance'AnalysiSi sStatusAnalysisand;Alar |ha'CO;R?« ipn^LineiReji^a^iJjty^Evaluatii :Man.-, Machine Systems';
lExpertjsya^^M
Survey of Activities Since the start of the computer control research programme in 1967 there has been a steady expansion of activities such that today most areas of importance for computer control of nuclear power plants are covered. From being an unreliable tool for solving non-essential tasks, the com- puter is today necessary for economical plant operation, and in the future it will probably perform BACKGROUND HISTORY most control and supervision functions at nuclear power stations.
Automatic data processing has, ever To facilitate handling of data from bration of instrumented fue] assembly since the first dynamics experiments were reactor physics experiments, and later on power sensors. A device named EVA, for carried out on the reactor, been a neces- also from test fuel assemblies, a variety converting up to eight analogue signals sary complement to the reactor instru- of electronic equipment, including ana- simultaneously to digital form and recor- mentation and experimental equipment. logue as well as digital computers, was ding them on paper tape, was construc- Before the age of computers, the recor- installed. The Transfer Function Analyser ted. However, before we got our own ding and analysis of instrument signals controlled sinusoidal reactivity perturba- digital computer, the data had to be pro- was performed by means of analogue sig- tions and computed transfer functions cessed out-of-house. nal conditioning instrumentation, chart between reactivity and selected process It became obvious in 1963 that our di- recorder and hand calculation. variables. ISAC was a device used for re- gital computing requirements were in- Our first analogue computer, the Don- cording and on-line statistical analysis of creasing rapidly. Accordingly, a GIER ner computer (1959), was programmed measurement signals. The BECKMAN/ medium sized computer was installed that to solve various special tasks, including EASE analog computer was used prima- year. This computer was programmed to reactor simulation for operator training, rily for on-line data reduction, determi- solve a Variety of tasks, such as reduction xenon simulation and reactivity measure- nation of hydraulic characteristics of ex- and statistical analysis of experimental ment. perimental fuel channels, and in-pile cali- data and reactor physics calculations.
21 EVA Analog to Digital Converter
I SAC TFA Statistical Analog Transfer Function Computer Analyzer
GIER Digital Computer
Hydrodynamic BECKMAN Test Loop Analog computer Data Processing Facilities 1960—1964 These facilites were quite efficient for carrying out the reactor physics and dynamics experiments. The rapidly growing volume of the fuel research programme, however, soon required more power- ful data processing equipment.
Computer control was taken up as a 1800 process computer was installed, "classical" and "modern" control prin- new research topic in 1967. To some ex- which was also used for logging and ciples. Particularly concerning control of tent this was a logical extension of past monitoring of data from the reaclor and systems with a multitude of interdepen- work with data handling for the reactor for the fuel research programme. Al- dent variables, "modern" control approa- dynamics and fuel research programmes. though of little interest from a control ches revealed themselves as very power- However, most important for the new point of view, the development of soft- ful, and considerable experience with activity was the recognition of the pro- ware for the latter function gave the firsL such methods has been gained over the cess computer as a very important tool experience in process computer applica- years. for optimum operation of complex nu- tions. An alpha-numeric console for opera- clear power plants. Moreover, the HBWR During the early period the computer tor-computer communication was built seemed Lo be well suited for research and was tried for simple conlrol functions, and installed in the reactor control desk, development in this field with ils power such as steady state control of nuclear and before long the operators became reactor plant complexity and unlimited power, primary circuit pressure, steam dependent on the computer for optimum availability for experimcnlalion. An IBiVl- generator water level, etc., using both control of the reactor.
THE COMPUTER AGE
The activities in the process computer which in a modified version has been in- control tasks at nuclear power stations control field have undergone a steady ex- stalled in several Swedish buill nuclear have also been studied for several years. pansion and cover a wide range of topics power stations. A revised version of this Tne decentralized modular process com- of considerable importance for the safe system was later developed, intended as puter system DEMP, based on co-opera- and economic operation of nuclear power a main communication system in com- tion of a group of computer resources, is plants. puter controlled plants. A prototype ver- the most well known result of the early Work on operator-process communica- sion was set up in an experimental con- work in this field. The versalility of mo- tion stalled already in 1969. The objec- trol room, with interface to a real-time dern computer equipment and the shift tive was to develop communication sys- nuclear power plant simulator, enabling of emphasis concerning reliability and tems utilizing modern computer based experimental investigations of communi- costs from hardware to software, resulted communication devices, and to study hu- cation procedures etc. under realistic in the later development of CONSUP, a man engineering aspects of such systems. conditions. At present a full-scale simu- modular computer structure with parallel It all started with the well-known OPCOM lator based control room is being estab- processing and built in redundancy for system, which was a communication sys- lished to enable more advanced experi- high computing capacity and increased tem primarily based on colour-graphic ments. reliability for critical functions. The pre- displays. The HBWR was operated for High reliability computer structures sent development in distributed micro- periods of months through this system, suitable for performing supervision and processor systems and communication networks is adding new dimensions to The w'ork on core surveillance and con- core surveillance, the emphasis has been the solution of these problems, and the trol grew out of the fuel irradiation pro- on use of core simulators on-site in order underlying computer structure for the gramme with the need for better control to obtain information about the dynamic new full scale control room will be based of the power in each of the large number behaviour of the core in response to pro- on these principles. Almost any control of test fuel assemblies distributed in the posed control inputs. Such simulators and supervision task at a nuclear power HBWR core. The objective was to develop can for example be used to ascertain that plant could then be performed through efficient methods and systems for super- local fuel heat ratings stay below opera- such a structure. vision and control of the reactor core tional limits during load maneuvers. Du- power distribution during variable load ring the last couple of years, an integrated The computer's capability to perform operation. More recently, the work has core surveillance system calLd SCORPIO intelligent information selection has been entered problem areas related to the dy- has been developed and successfully de- utilized in studies on status analysis and namic behaviour of power reactor cores. monstrated, as a proof for the usefulness alarming, and on interactive disturbance Much effort has been devoted to the of such systems in the operation of nu- analysis. The aim is to extract informa- adaptation of optimal-control methods clear power plants. tion which is relevant in a current opera- for power distribution control. The aim tion situation from the vast amount of has been to obtain maximum core man- information available, and use this in the euverability within constraints derived identification of disturbances, including mainly from the limited ability of the determination of prime cause and predic- fuel to endure power shocks. Several con- tion of the most likely consequences. trol algorithms have been developed and Studies in this field have been hacked up tested in extensive simulation studies. In by experiments with the HBWR, in which the computer selected the information to be conveyed to the operator depending on the actual operation condition. Based on these studies and experiments, a con- cept of alarm handling based on simple process status logic has been developed, and will be extensively tested on the new full scale simulator. A disturbance analy- sis system based on reliability analysis HBWR process interface techniques was installed at an operating nuclear power station for testing purpo- ses, but due to the vast amount of process analysis required to achieve the proper DEMP Process control OPCOM
level of accuracy, these kinds of systems N 2 B have not yet been accepted. JE Reliability analysis techniques do play NORD 1 an important role in the design of nuclear Tfi r f power plant control and supervision sys- tems, and in addition to using it in the —T design of highly reliable computer struc- Inter computer communication bus tures and in status information analysis, it has been used in other projects, e.g. an investigation on the merits of applying Operator communication CONSUP these techniques in the supervision of and simulator plants and component reliability during dvelopment NORD 10.03 plant operation. But the most important H use has been in the design of compute- rized safety systems. It started in the NOROSO NORD 10.79 early seventies with programming and testing of the triplicated fully computer based RESS system, later installed in Terminals CiYBER 74 e3" parallel to the conventional protection Function . ] system at BrUnsbUttel nuclear power sta- computer Decentralized process computer system tion. In more recent times, work has been focused on software reliability problems. The IBAl-2 800 was installed in 196? to simulators has made it possible to investigate In particular, the problems of correct handle the increasing amount of fuel and plant more complex and representative control specifications and the validation of the data, and was also a main tool in demonstrating rooms and supervision systems, and the DEMP/ the operational benefits of computer-based OPCOM systems were replaced by the CONSUP finished program has been given a lot of control and supervision. Other aspects of system, where a modular structure of parallel consideration, and a special specification computerbased control and supervision of the processing function computers could handle language has been developed, along with HBWR in the early seventies were demonstrated most of the control and supervision tasks at a programming aids, the so-called X-SPEX through OPCOM (OPerator-COMmunication by nuclear power station, represented by the programme system. At present an inves- the use of colour CRT displays) and DEMP STUDS real time plant simulator. This sytem (increased reliability and flexibility through has also been used for extensive human factor tigation is being performed aiming at tes- DEcentralized Multi-Prosessing). The reliability experiments. At present, a full scaled control ting the effect of language and formal aspects in computerized protection systems room based on a detailed nucleur power plant specification methods on programme were taken up in the triplicated RESS system. training simulator is being established. correctness. More recently, the use of nuclear power plant
23 PRESENT ACTIVITIES
Operator Interfaces
Colour display based operator interfa- ces require both a flexible colour display system and flexible operational proce- dures for control desk functions. The earlier systems have been based on semi- graphic controllers, but presently a high resolution fully graphic system has been acquired and a picture generating soft- ware system called PICASSO is being de- veloped. For control desk functions, a dynamic function keyboard based on touch screen techniques is being deve- loped. The use of voice input/output is also being considered. PICASSO is based on earlier experi- ence with picture editors developed at the Project and feedback from users of the editors and the surrounding software necessary to display on-line pictures. These editors were based on semigraphic colour CRT's with the limitations of North Sea Oil Production Platform small resolution and problems of display- An overview of the production process of a library, and showing the recommended colours ing trend curves. PICASSO generates mi- North Sea Oil Platform is here shown on a high for the oil/gas/water-parls. Process variables, mic diagrams on fully graphic colour resolution screen. The picture is generated by trend curves and bar graphs can be generated CRT's, and includes meLhods for connec- PICASSO based on a macro-defined symbol anywhere in the picture. ting the picture to the process data base, to make it possible to dynamically update the diagrams according to the process the calls to fit the subroutines usually state. PICASSO consists of four main delivered together with the hardware. If blocks; the symbolic directive compiler, there are functions not existing in hard- the interactive compiler, the picture-to- ware, they have to be simulated in soft- proccss linker and the on-line display ware. To transfer to different types of part. In addition it contains a connection controllers, changes are done in the de- to the data base (data base interface) and vice adapter only, not in the compiler or to a logic compiler. display program. In PICASSO a picture generating lan- The keyboards formerly used in con- guage has been designed. The aim of this trol desk design arc based on push- language is to make it easy to define a buttons for generating the desired infor- picture for people not experienced with mation and for executing control com- computer programming. In fact, a person mands. Each Lime a button is pushed, a Dynamic Process Variable Visualization totally unfamiliar with programming shall fixed output code is generated. From the In a fully graphic colour display system, the be able to generate a picture, connect to function keyboard the operator should three-dimensional behaviour of special process the process, and display it. Special effort be able to control the complete process, parameters can be displayed by using special has been placed on constructing the lan- and for a complex process this may re- display techniques. guage such that it becomes very readable quire a large number of push buttons. and therefore well suited for documenta- Compared with these keyboards, the use chosen. It uses a transparent overlay on a tion. PICASSO does not have parameters of touch sensitive screens offers a lot of CRT involving two sheets. The switches and is more a command language. Some improvements. consist of three parts. One sheeL is glass standard functions for number output, Several technologies have been used in or plexiglass and coated with a transpa- bar graphs and trend curves exist besides the design of a touch sensor. The sensors rent resistive substrate. This sheet is situ- the possibility of using basic primitives. currently available on the market use the ated close to the face of the CRT. A se- If one needs complex calculations to ge- properties of anumberof different physi- cond sheet is plastic with a transparent nerate curves or other types of advanced cal phenomena. A common technique in conductivc layer. The two conductive computer graphics, this can be realized the design is to have a signal generated layers are sliced in stripes and mounted by appending the compiled code to the by the sensor travel through some media. peipendicular Lo each other. A spacer is current picture. The display program then The signal is then detected by the sensor. mounted between the two sheets and docs the output, and device independency If a finger has been placed on or near the keeps the second sheet near, but not is achieved since PICASSO has been de- media, the finger modifies or changes the touching, the first shecl. When a finger signed to meet the demand of device in- characteristics of the signal. The detector or other actuator applies pressure, the dependency. It shall be easy to display measures the changed characteristics of flexible sheet is forced against the sheet pictures and use the interactive compiler the signal and determines the presence of glass or plexiglass. The flexible con- using different types of graphic control- and position of the finger. ducLor makes contact with the fixed con- lers. This is achieved by having standard For further investigations at HP, the ductor and thus closing the switch. When calls lo a devicc adapter which converts "membrane switch" technique has been the actualing member is removed, resili- ency of the flexible sheet causes it to re- turn to the normal position. The voltage of a signal depends on the point of contract, and voltage measuring sensors compute the x-y coordinates of the touching finger.
Touch Sensitive Screens In the actual application a touch sensitive panel of the membrane type will be formed la fit a 12 inch CRT. The touch sensitive screen is divi- ded into 192 touch pads. 12 lines and 16 touch pads per line. The size of each touch pad is 1/2 x 1/2 inch. A microprocessor controls the touch plate continuously and executes the com- munication and control to the CRT. The resolu- tion of the CRT is 64 characters/symbols per Plastic tine by 24 lines, i.e. each touch pad includes 2 Spacer character lines with 4 characters on each line. Glass or plexiglass
Man-Machine Communication
Involvement of psychologists in the
design of control systems in'Nuclear Po- Man • machine system wer Plants is a relatively new activity at Halden. The incentive is that operator Development and evaluation errors and poorly displayed control room information should be avoided. The Halden Project has established a Man-Machine Systems Laboratory, HAMMLAB, in which a team of specia- lists including psychologists and compu- ter experts, are working on development HAMMLAB and evaluation of process information display techniques, which aim to support and assist the operator in decision mak- ing, particularly when the plant is distur- bed or operating under abnormal condi- tions. For example, the HAMMLAB staff have expended a great deal of effort in devising systematic procedures for the experimental validation of newly desig- \ ned control room displays and operator Theoretical aids. background There is an increasing trend towards the use of computers in the control room. In the future, more and more plant control systems will include computers as an integral component within the man- machine interface. Whether such a sys- tem is successful in carrying out any given control task depends upon the proper allocation of functions between the hu- man operator and the machine. More- over, it is important to be aware of the possible impact of the installation of computerized aids in an established con- trol room environment. There must al- ways be proper integration of the com- Aspects of HAMMLAB: puter within the whole control environ- ment. Perhaps the most important requi- The structure of the Halden Man Machine retical. It should be stressed that these basic Systems Laboratory should be considered in aspects are closely interrelated and all equally rement of such a system is that it should terms of three basic interdependent aspects: the important in the successful planning of a re- keep the human operator well informed physical, the methodqlogical, and the theo- search programme. The Experimental Validation of the Critical Function Monitoring System: This new display system, designed to assist operator decision-making during a process dis- turbance, has been the subject of detailed, and systematic, experimental validation by 11 aiden Project. The experimental observations were made on site at the Lo viisa Arw clear Power PI a n t Ill & Q 0 Training Simulator using experienced plant operators. £ B. JHii Halden staff, along ivt'th personnel from the 1 Technical Research Centre of Finland (VTT) 5 affiEf" ^*-'*:; I 1, and the power utility (IVO), are seen hare, at Loviisa, se11big up and testing the equipment in p re para tion for the experimental t via Is.
yywi
about the status of the whole process ning substantial empirical data about its rious levels of process simulation fidelity. and, in particular, about the status of performance. For instance, in the case of the HALO- those tasks which have been delegated to In order to make a useful validation of system, we have obtained valuable results automatic systems. a man-machine system it must be syste- using static representations of alarm dis- There arc two principal reasons for matically tested under representative con- plays. On the other hand, the experimen- carrying out an experimental validation ditions to determine whether it performs tal validation of the Critical Function of a new man-machine system. First, the according lo expectations. But it is not Monitoring System (CFMS) (a display original development and testing normal- sufficient to restrict such a validation to system developed by Combustion Engi- ly takes place in a rather simplified envi- a purely technological investigation; there neering Inc.), necessitated experimenta- ronment where test cases will have been must also be a careful assessment of how tion in a much more realistic setting using carefully selected and thus may not be well the man interacts with the system. experienced PWR operators in the Lo- appropriate to real world operation. Se- The Halden team is currently involved viisa power plant replica training simula- cond, this type of system is usually de- in several evaluation studies. In order to tor. In future projects we plan to exploit signed to cope with rare or unusual events carry out this work we have devised a sys- our own full-scope PWR simulator which so there can be little possibility of obtai- tematic methodology which employs va- is currently under construction.
Highly Reliable Computer Structures
Work has for many years been carried loops and to place the microcomputer in levels as data acquisition, control or task out to investigate the potential of com- the field close to the sensors and execution, operator communication, and puter techniques in control of complex actuators. This has led to a steady trend data storage and retrieval. As a conse- processes. The use of minicomputers towards distributed control systems, quence of this, the multicomputer opened up new application areas of real which relative to a centralized computer system is given a hierarchical structure time control, but the concentration of system, has the advantages of lower which gives the opportunity to investi- all monitoring and control functions of a installation costs and greater security of gate such structures for more general plant in a single computer was associated operation of the plant. However, new control and supervisory systems. with the risk of severe economic conse- problem areas are introduced when trans- Two of the main requirements which quences of downtime resulting from com- ferring daLa between microcomputers in are put on the structure of a com- puter failures. To increase the reliability, different configurations, which can have puterized system for safe operation of a multiminicomputer structures were a strong effect on speed and total system nuclear power plant are that the different developed and successfully demonstrated reliability. control and supervision functions shall on complex industrial processes. Based on the experience obtained in be available with a very high probability, The situation has improved signifi- the areas of decentralized and distributed and that the different functions them- cantly with the appearance of the highly computing, the design of a new system selves shall comply with a multitude of integrated microcomputers as an ideal utilizing the latest technology as micro- different functional requirements which component for carrying out the moni- computers and fiber optics has been should be easily implementable into the toring and control functions required in initialized. As in most complex industrial structure. process applications. By utilizing this plants, the control and supervision func- Strict licensing criteria on some specific new technology it is possible for each tions can be split into functional sub- functions require a simple software unit to control asmall number of process systems representing typical hierarchical structure, easy-to-prove hardware, and
26 only a very loose communication to the rest of the system. The ideal concept would be to establish a hardware struc- Operator communication/colour graphs ture with a "standard" communication system that could be "prelicensed" and accommodate any communication re- To other display equipment quirement set by software modules im- plemented according to their specific functional requirements. The reliability requirements are solved by introducing redundancy, diversity and separation criteria into the hard- ware structure. The functional require- ments are dealt with by adjusting the hardware modules to these require- ments, using multi-processors, array pro- cessors, functional overlapping, etc. The difficulty lies in the integration of the different functional and reliability re- quirements on hardware into a total concept for a computerized system. To serve as such a totally integrated super- vision and control system, the hardware structure will, in addition to the required reliability and flexibility features, also include features like:
- a modular structure for on-line ex- pansion, e.g., during commissioning - standard interfaces for upgrading capabilities, e.g., avoiding obsolation by replacement with newer technology - automatic testing facilities and com- ponent failure and effect analysis in the maintenance and testing pro- cedures - on-line error detection and correction strategies by programme monitoring and data validity checks Process interface In an integrated multifunctional system structure, the data communi- cation strategics will become critical for an overall optimal performance. The different functions will have very diffe- rent requirements on data rates, response Distributed System for Process Supervision and times and interfunction data exchange, Control and the structure proposed for imple- The main functions of the design of a de- mentation must be able to accommodate centralized and distributed computing system different data communication media and structure, and are as essential to safe can be split into a four-level hierarchical strategies dependent on requirements. operation as the functions using the data. structure. 'The levels cover the basic functions The flexibility inherent in the basic hard- High data availability can be assured of data acquisition, data communication, task ware structure should also be reflected execution, operator communication and data through at least two different methods; storage management. in the data communication philosophy, redundancy of critical data communi- Generally, the communication between com- in the sense that the data communication cation paths, including the process and puter modules is arranged such that practically strategy should within wide limits be un- operator interface; and diversity of data any number of modules can be connected in affected by changes in data communi- communication paths, meaning that on parallel on any level in the hierarchy. This arrangement will also be favourable iti view of cation requirements. breakdown of normal data communi- easy maintenance and expansion, and will The question of data availability must cation, an alternative, but not optimal, greatly affect availability and the methods of be handled by the data communication data communication path is established. testing the system. Software Reliability
Along with the increasing use of com- is the possibility to make simultaneous puters in the control and instrumentation test runs of the two programmes. The systems of nuclear plants, it has become identical performance of the two pro- evident that the methods of programme grammes will be a good test of the cor- production and software validation re- rectness of the programme versus the spe- quire more attention. This is especially cifications under normal as well as ano- the case if computers shall be applied in malous conditions. This may reveal am- the protection system of a nuclear plant, biguities in the specifications, since such in which case an absolute confidence in ambiguities arc often connected to ano- the correctness of the implemented pro- malous situations. grammes is required. As a consequence of Various proposed methods for quanti- this, a joint project with the Research tative reliability assessment were applied Centre of Finland (VTT) was started in to the data acquired during the debugging 197 7-78 to study methods for the pro- and testing phase. The PODS Project duction and testing of reliable program- The experience from this project sup- mes. SRD (System Reliability Directorate) will ports the idea that all phases in the de- act as a customer and produce customers speci- The intention of the project was to in- velopment, from the initial specification fication. These wilt be translated by the X- vestigate various methods for producing to the final verification, should be based SPEX system, to evaluate the merits of this sys- and testing highly reliable software, in- on the same underlying structure. tem. There will in this project also be three in- cluding five phases: specification, dual The co-operation with VTT continued dependent programmes, one at Halden and one at VTT based on the X-SPEX specification, and programming, programme analysis, pro- with the development of a formal specifi- one at CEGB, based on their own manufacturers gramme testing, and reliability assess- cation language called X and a computer specification, i?i order to investigate the possi- ment. programme (SPEX) for analyses of the bility for identical errors in two programmes. The specification was produced as a specification which also produces a speci- joint effort, but the programming was fication document. performed independently by a team at An application of X - SPEX on test between VTT, the Halden Project and VTT and a team in Maiden, both making examples, including a majority voting SRD and CEGB in UK. This project, cal- their version of the programme solely on system and an alarm analysis system, was led PODS (Project on Diverse Software), the basis of the specification. Both teams to investigate the applicability of X and is similar to the first project, viz., a com- used a procedure of letting an inspector SPEX and to determine the semantics of mon specification as a basis for separate supervise the programmer's work. the X-language. programming activity with a final back A main advantage of dual programming In 1982 a joint project was initiated to back testing.
Computerized Operation Manual for Reactor Safety Systems
The protective systems of nuclear present the results to the operator in the list all the possible second failure combi- power plants are highly redundant and best way. nations within the reactor safety system can accept certain combinations of failed During 1979 a pilot version of such a along with the respective action, ranging components without the loss of functio- system was completed and tested in the from unlimited repair time to immediate nal capacity. The problem is to determine computer laboratory at the Project. This shutdown of the plant. how failures in the protection system system is based on rules defined a priori, The user interface was implemented a ffcct the overall availability, and which so-called tables of second failures, which by an interactive colour CRT communi- measures should be taken by the operator in case of failures. The structure of a re- actor protection system permits at least the occurrencc of one single failure. De- pending on the failure combination, the occurrence of second failures may be tolerated for a limited time or not at all. In 1977/78 the Halden Project, in co- operation with Osterreichisches For- shungszentrum Seibersdorf, Austria, de- veloped a concept for computer assis- tance in the use of operational manuals for safety technical specifications. A prototype system called RGB (Rechner- These pictures show the overview of the status tails on any system, as depicted in the second gestutztcs Betricbshandbuch), of a com- of all ten safety systems included in the Fors- picture. This request has been made approx. 2 mark CSTS-system, and the details of the sys- hours later, and the system has automatically puterized operational manual was deve- tem with the most serious limitation (323). A counted down the time, allowing 1 clay and 22 loped on the basis of this concept. set of non-operating components has been in- hours of continued operation in this state. Also The objective of the work on surveil- put to the system, and an analysis of the system shown are the circuits of the system which is in- lance of protection systems is to develop status with respect lo the safely technical regu- operable and which additional tests that arc re- a system which can analyse the status of lations has been performed. The time limitation quired. Finally, the operator must acknowledge for operation of each system is, in addition to when these tests have been performed and his the protection system concerning failed the indication of additional test requirements initials. The last acknowledgement is shown on and operating components, in order to (JA = YES) and the relevant references to the the picture. find the proper action to be taken and to written manual. The operator can request de-
*>R cation system. By using this system, the Power Plant. This pilot installation called display, giving the operator a quick and operator could enter information on the CSTS-system (Computerized Safety efficient comprehension of the conse- failed components in the protection sys- Technical Specifications), was implemen- quences of changes in component status. tem and obtain an immediate real-time ted at Forsmark in 1980 and used for an In additon, the CSTS-system makes it evaluation of the total protection system on-site experimental investigation in possible to simulate faults in connection status and the required operator respon- order to test the main ideas of such a with planning of maintenance work. The ses through colour pictures displayed on computerized off-line system and to gain operator can feed in planned status chan- the CRT screen. some useful feedback from practice. The ges in a test mode to investigate possible The next step was to test the concepts operator can enter the changes in status restrictions. Last, but not least, the sys- in an operator nuclear power plant. For of operation for components in the safety tem can be used for training and educa- this purpose a co-operation was estab- system, and the computer will keep re- tion of operation staff. The operators lished with Asea-Atom and Studsvik cord of the component status and inform can check their own decisions against the Energiteknik in Sweden, with the aim to the operator on operation restrictions, results of the computer analysis and train install such a computerized operation time limitations and test requirements. in handling different operational situa- manual in the FORSMARK-1 Nuclear The information is presented on a colour tions.
Alarm Handling
The development of the HALO-con- The operator needs a simplified pre- properly considered. The operator will cept (//andling of/llarms using LOgic) is sentation of the plant status in order to for example perceive changes in pre- based on the assumption that the opera- maintain a clear overview while he needs defined patterns of symbols and colours tor needs support especially during the detailed information to support his more easily than reading alphanumeric first stages of his decision making, in diagnostic work. The HALO alarm text. Efforts have therefore been made focusing his attention into the right presentation concept reflects this duality to establish recognizable patterns in the direction and in collecting information. in the operator's needs by separating the pictures by more extensive use of One of the worst problems in present presentation of the information needed colours and symbols rather than using days' control rooms is that too much for overview from the detailed infor- alphanumeric text to convey the infor- information is presented in a short time mation needed for diagnostic work. mation. during major disturbances and that the It is essential for a good solution that The purpose of the HALO system is essential information for these situations the operator's visual perception capabili- thus to extract relevant alarms out of the is stratified around the control boards. ties and his reasoning strategies are available status information. The means of reaching this are by using logical com- binations of signals to form alarms.
OFF-LINE ON-LINE PROCESS
Input The HALO system is functionally divided in two parts, the off-line and the on-line part.
The main part of the off-line system consists of an editor-function, this programme is a batch- type of programme. The symbolic input is checked for completeness and stored on a file for later use. Documentation of tile input can be produced at any lime in a clear readable format.
In the translation phase the symbolic input data is converted to the internal data structure for Trans- 1 j the on-line system. The internal structure lation 1 . phase 1 I consists of the logical expression for the alarms in a special format together with some attri- 1 butes describing the alarms. Error messages con- | | cerning the syntax is printed in self-explanatory text. 1 Alarm l The on-line programme system consists of four data i main parts. First the registration block where internal raw process data is collecteil. In the prepro- structures cessing step a validation and classification of ••• " 1 • the different signals is performed, consisting of 1 different kinds of checks like limit check, range check, consistency check, redundancy check, 1 etc. The alarm generation block runs through all the different logical expressions for the alarm conditions. If any of these are true, the Colour displays presentation block shows the corresponding alarm on the display.
29 The process of generating alarms can lie divided into two parts:
Generating alarms if some of the automatic functions that should follow a trip are not carried out. Signals indicating that these functions are carried out are suppressed, e.g., the system does not give out a set of indications for the operation of each control rod group but instead gives an alarm if some of the control rods do not go in (called sequence monitoring).
Extracting relevant alarms out of the signals which would be presented as alarms in a conventional system. This is done by suppressing signals which are normal consequences of specific process conditions e.g., low reactor pressure during shut-down. Also multiple signals about one process condition arc suppressed and only one is displayed. HALO overview display. Two urgent group alarms are seen in this other two colours in the picture are blue and picture: One denoting a serious disturbance in grey denoting an operating plant subsystem and As a result of the processing described the main grid (e.g. loud rejection) and the other a shut-off subsystem or an inactive group alarm displaying a steam generator disturbance. Red symbol respectively. above, a relatively large share of signals alarm colour indicates highest urgency. The which would be alarms in a conventional system will be suppressed. Thus a rather drastic reduction in the amount of alarms can be expected (cTr. next sec- tion). The HALO concept has three kinds of displays that the operators can request on different screens in the control room: An overview picture, detailed alarm group pictures and alarm texts. The objective of the overview picture is to give the operator a possibility to obtain with a glance the main status as well as the alarm situation of the pro- cess. In the overview there is a schematic process diagram which is divided into areas representing subsystems in the pro- cess. When one or more alarms in a sub- system are active, the corresponding area in the overview picture is given the actual alarm colour. When there is not any active alarms in a subsystem, the corresponding area in the picture is given a colour, which indicates whether the subsystem is operating or not. As some Proposal for a HALO detail picture that the likely going to fall below the lower alarm level, operator would gel by addressing the steam of the alarms cannot be directly related which is indicated by the dot on the side of the generator symbol in the overview. The reason steam generator symbol. Reaching this level to the main process diagram, they are for the group alarm is a low level in the steam would cause a scram. In addition to the alarms grouped together by their origin and are generator 2. The actual level indication is semi- the operating status of components and a given a spccial symbol in the bottom row blinking in red. Fortunately in this case the number of analogue information are shown in of the overview picture. As a result, each level is already on its way up as shown by the this display in form of cotours/bargraphs and alarm belongs to one alarm group which small arrow pointing upwards. Thus it is not digital indications. is represented byanareaon the overview. The urgency of an alarm depends on the corresponding alarm group symbol The alarm group detail pictures are the seriousness of its conditions and on on the overview screen or the corre- schematic diagrams which can display the time allowed for an eventual opera- sponding alarm text line and depressing a individual alarms in a way similar to the tor counteraction. Urgency is indicated "picture,, push-button. overview. by using different alarm colours and A prototype HALO system has been The alarm text displays arc lists of audible signals. developed and is implemented on the alarms in chronological order. Each such The overview picture or an all-alarms STUDS simulator. A more detailed indication includes time when set, list can thus always give an indication version is presently being planned for the identification code and alarm message in when new alarms are present. A detail new NORS experimental facility. plain language. picture can be requested by addressing
30 Core- Surveillance and Control
Core Surveillance and Control lias been a research item for many years at the Maiden Project. Originally it was the need for improved and more detailed core power control during fuel testing in IIIJWR which triggered the work, but later the dynamic behaviour of power re- actor cores has been studied. Optimal control methods have been adapted for power distribution control In 1 979 the SCORPIO project was started (Surveillance of a reactor CORe by Picture On-line display). SCORPIO is a system which to a large extent com- bines the results from several of the Mai- den Project research activities. Typical examples are the operator-process com- munication work, the data base systems, the core simulator development, and the studies at optimal power distribution Display of core power distribution: Maximum cates the axial nodal position of the power peali. control methods. A main task has there- nodal power is indicated by a colour code on To the right a selected (see yellow cross.1) axial fore been the design of a system that the cross section lo the left. Each number indi- power profile is displayed. combines all these different modules into an integrated unit. The definition of interfaces between modules constitutes a major task. SCORPIO is intended both for PWR and BWR simulation. The system struc- ture is independent of reactor type. De- tails within modules and in the commu- nication between modules are, however, partly reactor dependent. SCORPIO is meant to be an efficient tool for the ope- rator in the control room. The simplified block diagram of SCORPIO can be used to identify the two basic modes of opera- tion, i.e. the core folloio mode and the j> re flic live mode.
.-! predictive simulation can be followed by watching this picture. Together with the speci- fied power trajectory the trend of axial power offset and of minimum margins to safety limits are displayed for each time step.
The man-machine communication system in SCOKPIO aims at being an efficient tool for a reactor operator. Here llaldcn stuff is discussing the design of a picture which is requested for detailed inspection of margins to PCI limits. The main function of the core follow Control system is to produce a good estimate of Closed the core status based on instrument rea- loop dings coupled to a three dimensional control physics model including xenon dynamics T" (CYGNUS). Detailed power distribution and other relevant data are stored in the Interactive communi- -Operator base data base. As soon as the core state data cation are available, the limit check module is activated. Margins to operational limits are accessed through the interactive com- munication system and displayed on a colour screen. In the predictive mode of operation Limit Strategy check generator the user can precalculatc the core beha- viour during a possible future power tran- sient. This capability is of great help in running the reactor in a dynamic situation Core Surveillance System SCORPIO: Simplified where xenon variations often have a block diagram showing the two modes of complex influence on the power distri- operation: core follow system and predictive system. bution. In this way the user can exclude strategics that are unacceptable due to operational constraints since the limit checking module is also activated in pre- Prepared -•^Satisfied >—• control dictive mode. strategy Input to a predictive simulation is done in an easy way via the interactive commu- nication system. Simple input specifica- tion and input check is gained by exten- sive use of function keyboard, a specific input dialog and immediate display of in- put quantities on colour CRT's. To help the operator also a strategy generator can be activated to produce a first sugges- Predictive simulation with SCORPIO: After control settings are detected, the operator specification of a power transient the strategy modifies input at critical time points. A new tion for the controller settings before the generator is activated. Simulation is started and simulation is then initiated and examined. This core simulator CYGNUS is started. results are examined as the simulation goes on. iterative readjustment of a transient is carried If safety limits are exceeded or undesired out until the operator is satisfied.
COMPUTER CONTROL Control room EXPERIMENTAL FACILITY test facility —I— Operator com, — Sim. oper. mulator development A new full scale computer-based a Q Oxv experimental control room facility is presently being established at the Halden Project. It includes the control room test area itself with the necessary facilities for operator communication through colour displays and other display equip- ment, and instructor facilities. It also includes the development and operation of a full scale training simulator called NORS.
The NORS simulator describes a PWR power plant in considerable detail. Most of the auxiliary systems required to simu- late the process under a wide range of HAMMLAB Computer Structure operational conditions are included. This NORD 100 for presentation and operator implies that safety systems like emer- The computer structure used for the HAMM- communication facilities. This structure has the gency cooling and safety injection are LAB simulator-based experimental control potential of satisfying all present computing room includes a^NORD 100/500 for simulator requirements, but a more distributed extension modelled. It is important for the use of development and operation, connected to two of the system is being considered. the simulator that a realistic description
32 of unusual or even unlikely events are The automation and computer aided The operators' control consoles will available as the research in safety has to supervision lead to increased demands on vary in size and shape from those aim at preventing or minimizing the the plant computer system. This is the operated by a single individual to those effects of such events. The simulator system charged with collecting and pro- operated by three- or four-man crews. program runs on a ND 100/500 com- cessing signals and information from the Control room mock-ups can be provided puter. 2 ND 100 make up what is the hard-wired plant instrumentation system. by rear-screen projection system. equivalent of the plant computer system. The use of processed information in Necessary log/event printers will be avail- In NORS they will handle not only the addition to the conventional instrumen- able. The control room area can be usual process information buL they are tation can definitely improve the divided into two areas to perform crew also processing the operator-process com- operators' ability to understand present to crew communication experiments munication. In the first version of the status and predict future behaviour. The between a local control room and a control room there is no conventional logical consequence of this trend is that central control room. instrumentation, all control is carried out the plant computer system must be through process diagrams on VDU's. planned for handling the increased infor- The instructors' area will be separated Large storage capacity arc needed for mation flow. The design and use of a from the control room by one-way glass data collection, particularly during ex- more exlcnsivc plant computer system is or smoked glass that will allow the perimental runs where the number of one of the research and development experimentalist Lo unobtrusively super- variables recorded is large. topics for NORS. vise activity at the operators' consoles. Me also has a control console from which The control room area is the area in he can control the experiment and record which all man-machine experiments will the various performance data. be performed. It contains the central control room, a local control room, and The computer room will house the the instructors' work place with obser- necessary experiment to run the simu- vation gallery. lator and record experimental data. This equipment include processors, disk storage units, system consoles, line printers and communication interfaces.
Experimental Control Room Facilities a: computer room b: operator's work place { ." shift supervisor d: researcher's work place e: local control room (staging and test areaj These facilities and the full-scale simulator f: terminal room will be used in the research, development and g: observers gallery validation of operator support systems and h: conference, education man-machine interfaces, and in the development i: recreation of new training techniques.
33 Fuel Research Facilities
1. Reactor vessel 2. Reactor core 3. Heat exchanger 4. D2 0 subcooler 5. Subcoolerpump 6. Steam drum 7. La IVIont pumps 8. Steam generator 9. Hot well 10. H2 0 subcooler 11. Feed water tank 12. Shield coolant circuit 13. Magnetic jack control rod drive 14. Fuel storage pit
The HBWR reactor has served experimentalists from all over the world faithfully through twenty- four years, and has rendered valuable experience on heavy water power reactor operation, on fuel performance and on modem process control.
• The Halden Project has at its disposal Post irradiation inspection facilities in a selection of proven research tools: Maiden and hot laboratories at Kjeller A range of high performance com- ;. • The HBVVR reactor, particularly well puters and display systems DaO .•;:'; mode- SrP suited for research activities A computerized control room with a rator • Versatile in-core instruments capable large simulator for control experiments and operator training. mode-; of measuring a range of different fuel rator;®-, and fuel assembly parameters during •-v ' 4\:i irradiation • A series of irradiation rig designs Tor performing various types of in-core irradiation experiments of varying The Project has an experienced staff complexity capable of utilizing these research tools • Special high pressure rigs operating in an efficient way to perform: The heavy water reactor has a much larger core under modern light water reactor con- colume than a light water reactor of the same power, because of the longer neutron therma- ditions by means of external control • Research work according to the agreed Ifcing length in heavy water. This leaves corre- loops Project research programmes spondingly more room for experimental equip- • An efficient high capacity data hand- • Special tasks on behalf of participating ment. ling system organizations on mutual agreements.
34 IIBWR AS A FUEL TEST FACILI TY
Instrumented fuel assembly Several features make the Haldcn Boiling Heavy Water Reactor (HBWR) Steam-sampler Fuel pin diameter rig suited for studies of reactor fuel be- (failure detection)
haviour during irradiation: Coolant thermocouples Position indcator Tutbine flow meter Hydraulic • Being a heavy water moderated and Void gauge
cooled reacLor, it has a considerably Shroud larger core than a light waLer reactor Fuel pins of the same power. This gives better Gamma thermometers room for instruments and experi- Neutron thermometers . Coolant T/C mental equipment, and permits ir- Beta current neutron radiation of longer fuel rods. detectors . Fuel pin
• It has a flat lid with one penetration Neulro n for each fuel assembly in-core lo- Differential transformer thermometers cladding extensometer Pin diameter cation, making loading, unloading and gauge head relocation of fuel assemblies and ir- Fission gas pressure gauge radiation rigs relatively simple and Coolant thermocouples
straight forward, and the instrument Turbine flow meter Core can have 45 instrn cable seals can be made integral part Electromagnetic mented fuel assemblies Three-way three way valve or of the rig. of total 60 assemblies valve Pneumatic throttle valve • More than forty-five instrumented fuel assemblies or irradiation rigs, which means abouL three quarters of the core, may be irradiated simul- taneously, thus economizing on driver The Haldcn Project has at its disposal a reactor well suited for irradiation experiments, a range of fuel assemblies. in-corc instruments and versatile irradiation rigs, including some which through external control • The reactor is exclusively available for loops can provide modern light luater reactor operating conditions, and n powerful data handling experimentation. system.
Operating Coin! it ions
• Coolant temperature: 240°C • Coolant pressure: 34.4 ata • Max. thermal neutron flux: ca. 0.5 • 1013n • cm"2 • s'1 • Max. fast neutron flux: ca. 0.5- 1013 n • cm"2 • s-1 • Max gamma flux: ca. 0.5 W
Operating Conditions from 19S3
To be able to provide the neutron flux conditions met in modern light water power reactors the HBWR core has been (^'^^bJKjja^^^^^^B^^^^^^^Hw redesigned during 19S2 and modified accordingly early 19S3. The flux levels thus obtainable are consequently in- creased by a factor 2 to 2.2.
The reactor pit cover. Rotation of the two cir- cular lids permits an access opening in the smaller lid to line up with any position on the reactor vessel lid. Fuel handling can then be performed.
35 In the reactor hall is installed four light water loops connected to high Deprejjurization pressure flasks semipermanently installed to main vent in the reactor core. The loops will main- tain typical modern li\VR or I'WR condi- tions, i.e., 350°C and 70 to 150 bar pressure in the flasks. The loops will also maintain water quality, and are equipped for producing flow starvation or loss-of- coolant conditions in the irradiation flasks for experimental purposes.
Top seal plug
Water outlet
A simplified diagram showing llirec of the light water loops installed in the reactor hall to provide BWR and PWR conditions in the high pressure flasks. The system contains level tanks and com- pressed gas for pressure control circulating pumps, a high pressure feed water pump, water purifi- cation system, radioactivity monitor, and heat exchangers, coolers and heaters as required. Facili- ties for blow down and reflood is provided.
Schematic drawing of a high pressure irradiation flask, semipermanently installed in the HBWR for BWR and PWR experiments. An instru- mented fuel assembly is indicated in the flask.
FUEL HANDLING AND INSPECTION EQUIPMENT
New instrumented fuel assemblies or irradiation rigs are brought to a special hydraulic laboratory at the reactor site. Here eventual rig flow meters are cali- brated, and the whole assembly pressure tested at 40 liar, eventually heated in a hot loop to 240°G, as required. Finally it is thoroughly inspected and eventual locking operations performed. It is then stored in the hydraulic laboratory until loading into the reactor. When an irradiated fuel assembly is un- loaded from the reactor, it is temporarily stored in ponds below the floor level in the reactor hall. Later it can be returned to the reactor for further irradiation, or moved to the fuel bunker building. Here it is stored in a light water pond where it can be inspected through a periscope and finally prepared for transport. Special manipulation and inspection work can also be performed in a handling compartment located in the reactor hall beside the reactor pit. Dry storage cells for irradiated standard fuel assemblies. 36 With the shroud removed, a new instrumented fuel assembly is subject to fi?ial inspection after calibration and pressure testing. In the picture is shown preparations for tightening a group of gas lube connectors. The holes seen in the upper part of the rig are steam outlet openings.
Visual inspection and photography thro ugh a periscope arc performed on an irradiated fuel assembly in a pond in the fuel bunker building.
/I shielded compartment is installed in the reac- tor hall where irradiated fuel assemblies can be handled. The fuel transport flash is seen on top of the compartment. Two inspection and manipulation stations can be seen. A third windoio is also installed for inspection. Fuel rods can be inspected and photogra- phed. They can be removed from the rig and subjected to dimensional and pressure mea- surements, and they can be replaced by other rods, fresh fuel, or rods irradiated in other rigs. The round reactor pit cover is seen in the background. The metallic coil hanging on the rack arc metal braided bellows hoses protec- ting instrument cables and tubes from instru- mented fuel assemblies emerging fro m channels in the floor.
Part of an instrumented fuel assembly with the shroud removed can be seen through the inspec- tion xoindow. The top end of a tubular spiral for Ite-3 neutron flux control is visible. Skilled hands is an important asset of the Halden Project.
IRRADIATION RIGS
The first instrumented fuel assembly, rigs where the fuel rods could be ex- axial distribution or be built for PWR which was installed in the HBWR in 1962, changed for new ones in the handling or BWR conditions. The rig design may was soon followed by others, and the as- compartment beside the reactor pit. be varied infinitely, however. In the semblies gradually grew in number and These rigs are usually heavily instru- table eleven typical designs are listed, complexity. An instrumented fuel as- mented, and may contain manipulators and six of these are described in more sembly is expensive and lime consuming for moving measuring sensors or fuel detail in the following. to produce, and one therefore started to rods, and they may comprise equipment make reusable assemblies or irradiation for controlling neutron flux level and Irradiation rigs Functions of eleven different test fuel irradiation A bundle of differential transformer cladding Fuel cladding diameter gauges. The differential rig designs are listed below extensometers at the end of a fuel cluster. transformer housing of two of them can be seen and one feeler contact, resting against a fuel rod Q> IRRADIATION RIGS O c c lower end plug. Below are three cladding exten- a> a> a •S » E someiers. Bottom end of a self-powered neu- n« o O) c V 01 C o E» £ CC tron detector is barely visible at the right hand u I C £e n o « a 1! 5 .S o a o ID upper edge. It is embedded in a supporting r a o» 5 tr c _i F Ol 3 •o U c •O n c stringer to shield against electrons from the fuel. a o o it 0) DC CC "3 3 3 3 Measurements: Linear heat rate • Cladding elongation • Cladding diameter profile Fuel stack elongation Internal rod pressure Fuel centre temperature Fuel rod gap width Fission gas release Gas flow resistance Fuel failure • Cladding defects Cladding temperature Manipulations in rig: Power ramping Power oscillation Transporting rods axially Rotating rod Coolant blow off Special experiments: Fuel rod diameter profile Fuel rod gap measurements Gas flow/fission gas release Lossof-coolant BWB/PWR ramping Ramping of LWR segments ; 38 A Sealed electric connectors for centre fuel thermocouples and fission gas pressure trans- ducers. Fuel rod end plugs with thermocouple cable penetration, and pressure transducers are seen near the lower end of the picture. <3 Assembly drawing of an instrumented fuel ir- radiation rig. It contains inlet and outlet tur- bine flow meters and a number of differential transformer cladding extensomcters for two fuel clusters. < Connectors for gas lubes to fuel rods for inter- nal pressurization during irradiation. > Centre fuel thermocouple cables and cladding extensometers at the bottom end of a fuel cluster. 39 Diameter Rig Fuel Loading Three fuel rods, 40 cm long, replace- able. Test Capability • Recording of diameter profiles of three rods simultaneously • Cladding elongation measurements • Fuel end pellet movement recording • Local control of fuel rod linear heat rate and its distribution by means of He-3 pressure in tubular coils. Objectives Provide quantitative information on mechanical fuel/cladding interaction and its dependence on linear heat rate and its distribution, power history, burn-up, and fuel and cladding design and manufac- Diameter recordings from three fuel rods with different fuel/cladding gaps and cladding wall turing parameters. thickness. Notice reference diameters, three at each of the rod. Power Shock Rig Fuel Loading Four fuel rods, 50 cm long, replace- Cladding length changes l able. mm Testing Capability Objectives i/—.K.. . • * **«w' 0 • Moving one fuel rod at a time in or Investigation of how fuel rod failure Power 800 oul of test position by means of hy- caused by overpower shock is depending /—"—•• . . . draulic cylinders on design parameters and previous ir- r ..••J J • Control of neutron flux in fuel rod radiation history. lf W/cm test position by means of He-3 pressure Of particular interest is testing of fuel 0 in tubular coil, thus exposing fuel to rods previously irradiated at 200 - 400 ) Time(hrs)lO heat rates between 60 and 100% W/cm average linear heat rate to burn-up Straining and relaxation of fuel cladding • Measuring cladding length changes of 5 - 20 MWd/kg. during stepwise power increase. Power Cycling Rig Fuel Loading Objectives Eighteen fuel rods, 30 cm long, six in Study effect on fission gas pressure each of three clusters, irreplaceable. and centre fuel temperatures of: Testing Capability • Sudden power increase after prolonged irradiation at reduced power • Exposing fuel to rapid changes in • Rapid cycling between 60 and 100% power between 60 and 100% full power • Mis as u ring centre fuel temperatures • Different types of load follow opera- • Measuring internal fuel rod gas pres- tions. sure. 0 200 100 Linear heat rate (W/cm) 40 1. Fuel rod 19. Fuel column extensometer 2. Fuel rod in operation 20. Fission gas pressure gauge, membrane 3. Fuel rod, stand by 21. Fission gas pressure gauge, bellows 4. Shroud 22. Coolant pressure gauge 5. Forced circulation subcooled inlet 23. Cladding extensometer and failure detector 6. Coolant inlet ports, natural circulation 24. Fuel rod diameter gauge 7. Coolant outlet ports 25. Position indicator 8. 3-way solenoid valve (forced/natural circulation) 26. Water level detector 9. Coolant turbine flow meter 27. D2O hydraulic drive cylinder 10. Coolant thermocouples 28. He-3 tubular coil 11. Flux sensor (Va) 29. Neutron absorbing shield 12. Flux sensor (Co) 30. Lifting rod 13. Fuel failure indicator, steam sampler 31. Fuel rod rotating mechanism 14. Centre fuel thermocouple 32. Pressure flask 15. FliuI perifery thermocouple 33. D2O circulation and pressurizing tube 16. Cladding thermocouple 34. Bellows controlled valve 17. Shroud thermocouple 35. Gas bottles 18. Cladding extensometer Fuel Rod Gap Meter Rig Fuel Loading Objectives q = 46 kW/m TF = ca. 1700°C GF = 0 • One fuel rod, 400 mm long, replace- » Provide information on fuel rod inter- GC = o able nal gap dimensions as function of ir- q = 37 kW/m radiation TF = ca. 1600 °C GF =14 fjw Testing Capacity • Study relocation processes, such as ex- GC 31 urn centricity, densification, swelling and = 27 kW/m • Control neutron flux in fuel rod fuel cracking. TF = ca. 1300 °C GF = 25/im through He-3 pressure in tubular coil GC = 49 nm • Measure internal gaps in rod through q = 2 kW/m TF = ca. 430 °C compressing cladding at various axial GF = 33 Mm locations GC = 88 jum Typical results of gap measurements, q: kW/m; • Record cladding diameter pro files TF: fuel centre temperature, °C; GF: free gap, along two generatrices lim; GC: compressed gap, |im (cladding and • Measure cladding length changes pellet looses contact). Deflection • Measure ccntrc fuel temperature Gas Transport Rig Fuel Loading Study influence of gas composition and pressure on fuel to clad heat con- Four non-replaceable fuel rods. ductance Test Capability Study influence of gap size and surface • On-line control of internal fuel rod gas roughness on gap heat conductance composition and pressure Study influence of fuel temperature • Measure fuel centre temperature and and grain growth on fission gas re- fuel rod pressure lease. • Measure pressure distribution along fuel during gas flow Fuel to claddinggap (hydraulic diameter) versus linear fuel rod heat rate at beginning of life and 5 10 15 Objectives at a given burn-up, as calculated from flow pres- Linear heat rate, kW/m sure drop readings. • Investigate gas transport mechanisms Flow Starvation Rig PWR/BWR Forced Circulation Loop Fuel Loading Design Four to seven fuel rods, 1.5 meter A pressure chamber for on-line opera- long, replaceable. 1000 tion is connected to an external system o for control of pressure, temperature and Test Capability °„ s flow. Z3 • Measuring fuel temperatures, cladding a Operating Conditions temperatures, internal fuel rod gas £ 5m pressure, cladding elongation * • Temperatures up to 350°C • Control pressure, temperature, coolant • Pressure up to 150 atm. flow and coolant level in test section • Flow Objectives Objectives Study fuel rod behaviour under loss- Irradiating fuel rods designed for of-coolant conditions, with special Calculated temperature transients in fuel clad- various types of studies, requiring emphasis on the heat-up and reflood ding and shroud during loss of coolant acci- operating condition typical of operating dent: stages. boiling or pressurized water reactors. 42 Fuel Rod Gap Meter Rig Gas Transport Rig Flow Starvation Rig PVVR/BWR Forced Circulation Loop 1. Fuel rod 19. Fuel column extensometer 2. Fuel rod in operation 20. Fission gas pressure gauge, membrane 3. Fuel rod, stand by 21. Fission gas pressure gauge, bellows 4. Shroud 22. Coolant pressure gauge 5. Forced circulation subcooled inlet 23. Cladding extensometer and failure detector 6. Coolant inlet ports, natural circulation 24. Fuel rod diameter gauge 7. Coolant outlet ports 25. Position indicator 8. 3-way solenoid valve (forced/natural circulation) 26. Water level detector 9. Coolant turbine flow meter 27. D20 hydraulic drive cylinder 10. Coolant thermocouples 28. He-3 tubular coil 11. Flux sensor (Va) 29. Neutron absorbing shield 30. Lifting rod 12. Flux sensor (Co) 31; Fuel rod rotating mechanism 13. Fuel failure indicator, steam sampler 32. Pressure flask 14. Centre fuel thermocouple 33. D 0 circulation and pressurizing tube 15. Fuel perifery thermocouple 2 34. Bellows controlled valve 16. Cladding thermocouple 35. Gas bottles 17. Shroud thermocouple 36. Fuel rod compression mechanism 18. Cladding extensometer 37. Compression sensor 43 IN-CORE INSTRUMENTS Introduction Basis for the fuel performance studies performed in the HBWR is the range of instruments and actuating devices capable of functioning in the reactor core under full power operation. With the exception of cables and thermocouples, and self- powered neutron detectors, all these instruments arc developed by the Project and lnstitutt for Atomencrgi. To meet the in -core environmental conditions the instruments are manu- factured from materials which arc resis- tant io irradiation, elevated temperatures and corrosion, which leaves only special metals and ceramics. Coils are wound from ceramic insulated wire on bobbins which are enamelled or framcspraycd wiLh magnesia. Cables and Tcrmocouplcs The cables and low temperature thermocouples have 1 mm OD lnconcl sheath and aluminia insulation. The thermocouples arc type K (Chromel/ Alumel) with insulated junctions. High temperature thermocouples used for fuel centre temperature measurements up to 2000°C, occationally 2500°C, arc type I (W3%Re/W25%Re) with beryllia insulation in tungsten/25% rhenium sheath of 1.6 mm OD. >1 - — Fuel thermocouple penetration The irradiation induced dccalibration of these thermocouples is subject to re- search at the Project. A special gas tight seal is arranged in- side the fuel rod end plug to prevent fis- sion gases diffusing into the cable insu- lation. Neutron Detectors The selfpowered nemron detectors a: Fuel pellet d: End plug have vanadium absorbers, except for spe- b: W3%Re/W25%Re thermocouple e: Gas tight penetration cial dynamic measurements, when cobalt c: Cladding f: Chromel/alumel thermocoax cable ah-;orbcr detectors are used. 44 Turbine Flow Meter Design: • Rotor with eight blades made from ferritic chrome steel. • Graphite sleeve bearings on Inconel spindle, or sapphire ball bearings in steel races. • Pickup coils with permanent magnet core produce a voltage pulse for each passing rotor blade tip. • The rotor unit is replaceable, regardless irradiation history. Two basic turbine types are normally a: Rotor assembly d: Pick-up used: one with 59 mm throat diameter b: Ball bearing e: Removable cage and flow range 1.5-10 1/s and another c: Shaft f: Housing with 40 mm and 1 - 6.5 1/s correspon- dingly. Other designs of larger dimensions Turbine flow meter are available. Accuracy is ± 1%. Applications: • Fuel channel inlet and outlet flow measurements. • Calorimetric assembly power determination in com- bination with coolant temperature mea- surements, single and two-phase flow. Turbine rotor sizes • Outlet void fraction determination. Gamma Thermometer Design: • Gamma flux is absorbed in a steel cylinder mounted in argon atmos- phere in a pressure container. • Resulting temperature rise is measured by means of Chromel/Alumel thermocouples. • Time constant about 120 s. Application: • Gamma flux measure- Gamma thermometer ments, VV/gr. • Assembly power moni- toring after calibration against caiori- metric measurements (coolant flow and temperature rise). • Assembly pov.er distribution measurements. Remarks: • Very reliable sensor. • Un- disturbed by irradiation. • Less specific local indication thar„ neutron sensors due a: Gamma absorber, stainless steel to the longer range of gamma radiation b: Thermocouple, chromel/alumel in the core. • Strong candidate as apower c: Reference thermocouple distribution sensor in light water reactor cores. Cladding extensometer e Coaxial Differential Transformer Concentric, axial type. Design: • Single primary coil. • Secon- dary coil in two halves electrically ba- lanced. • Coaxial ferritic core disturbs a: Fuel rod end plug d: Ferritic core the balance of the secondary coil and b: Primary coil e: Twin-lead thermocoax cables cause a signal proportional to the core c: Secondary coil offset from central, neutral position. Data: • About 200 mV full scale signal Fuel end pellet position indicator at 50 mA, 400 Hz exitation. • Linear ranges available, mm: ± 2.5, ± 5, ± 6, ± 7.5, and ± 12.5. • Accuracy: better than ± 1%. Applications: • Fuel cladding extenso- meter. • Fuel stack end pellet position sensor. • Part of bellows based pressure sensor. • Pneumatic throttle valve posi- a: Fuel stack end pellet c: Spring tion sensor. • Fuel rod failure indicator. b: Differential transformer d: Ferritic core • Periodic position sensor. 45 Void Gauge Diameter Gauge tM*, Design: • Two sets of concentric Design: • Differential transformer cylindric electrodes, one set earthed, the with primary and secondary coils ar- other connected toasignal cable. • Resis- ranged symmetrically on an E-shaped tance measurement between electrodes. ferritic core. • Ferritic armature ar- • 1000 Hz carrier frequency yields an im- ranged along the face of the E, pivoted pedance signal proportional to void frac- in the middle. • Two feelers on opposite tion up to 60 - 70% void. side of a fuel rod transmits diameter Application: Dynamic void fraction variations to the armature. • A fixture measurements averaged over channel with three spring loaded fingers keeps cross section. the differential transformer in position close to the fuel rod. Data: • Sensitivity: 620 m RMS/mm, linear range ± 0.3 mm. Accuracy ± 1.5%. Application: In-core measurement of Void gauges fuel rod diameter profile. The gauge is moved by means of a hydraulic system, Fuel rod diameter gauge while a position sensor tells its location. Periodic Differential Transformer Position Indicator Design: A core consisting of alternate ferritic and non-ferritic sections passing through a coaxial differential transformer will cause a near sinusoidal varying out- d: Ferritic armature put. a: Primary coil e: Cross spring suspension Range and Accuracy: Combined with b: Secondary coil f: Feelers an integrating electronic system an ac- c: Ferritic bobbin g: Fuel rod curacy of ± 0.1 mm is obtained indepen- dent of measuring range. Periodic position indicator Linear Transformer Position Indicators Design: In a long solenoid coil wound from a 1 mm OD twin lead thermocoax cable one lead is used as a primary and exited with a 50 mA, 400 Hz current. The other wire forms the secondary. a: Non-magnetic core sections c: Differential transformer When a ferritic steel rod of constant cross b: Ferritic core sections section is moved into the coil, a signal is produced in the secondary proportional to the rod insertion distance. Linear transformer position indicator Pressure Gauge, Bellows Type Design: • A bellows is mounted inside a fuel rod. • A ferritic armature is mecha- a: Ferritic core c: Primary nically connected to the bellows. • A b: Coil wound from twin- d: Secondary differential transformer outside the fuel lead thermocoax cable rod cladding senses the position of the armature and thus the deflection of the bellows. Pressure gauge, bellows type Data: • Measuring ranges: 15, 60 and 90 atm. without spring, and 120 atm. with supporting spring. • Accuracy about + 1% of full scale. Application: Pressure measurements inside fuel rods. The bellows/core unit is removed with the fuel rod, while the dif- a: Fuel rod c: Bellows e: Ferritic core ferential transformer can be reused with b: Spring d: Differential transformer f: Twin-lead thermocoax cable a new rod containing a bellows/core unit, or with the same rod, if reinstalled. 46 Pressure Gauge, Membrane Type Design: • A thin membrane forms the boundary between the fluid to be mea- sured and an external gas system where the pressure can be adjusted to balance the unknown pressure. • The membrane rests against an electric contact, which signals when equilibrium causes the a: Fuel rod c: Electric contact e: Tube for pressure membrane to move. • The membrane is b: Membrane d: Signal cable balance gas located in a narrow, carefully shaped space to be able to withstand high pres- Pressure gauge, membrane type sure differences without being damaged. Application: Pressure measurements Moderator Level Gauge inside fuel rods. The gauge cannot be re- Design: • A pressure light titanium float used after unloading of the fuel rod. contains a fcrro-magnctic core • A pri- Data: Two types with measuring ranges mary coil with linearly winding density a: Float and accuracies: 90 atm. ± 1% and 120 surrounds the float • A secondary coil b: Ferromagnetic atm. ± 1% respectively. has constant winding density. Both coils core arc wound from 1 mm OD Inconel c: Primary coil, linearly varying sheathed cable • The position of the density Eddy Current Cladding Defect Sensor magnetic core determines the coupling d: Secondary coil, Design: • A twin coil system surrounds between the coils, and thus the output constant winding density a fuel rod coaxially. • The two coils signal • A transformer is installed Lo e: Reference trans- form one half of an AC bridge which is provide reference signal to permit com- pensation for temperature, frequency farmer completed externally and exited with f: Shroud 16 kHz carrier frequency voltage. • De- and current fluctuations. g: Opening fects in the cladding wall will, when the coil system is passed over them, produce Range: 2000 mm, 210 - 270 mVAC a bridge unbalance, whose amplitude and phase relationship is characteristic for the type of error. Moderator level gauge Fuel Rod Gap Meter Application: In-corc mapping of fuel Application: The gauge is installed cladding defects. The sensor is moved Design: • Two bellows operated with in a rig where the fuel rod can be moved along the rod by means of a hydraulic gas perssurc compress the fuel rod axially permitting measurements at piston or step motor, while a position cladding. • Deflection is measured by various positions. Deflection is plotted sensor tells its location. means of a sensitive differential trans- against gas pressure and analyzed. former. Resolution: ± 4 pm. i Fuel rod a: Fuel rod b: Bellows c: Differential transformer Coils Signal cable Eddy current cladding defect sensor Fuel rod gap meter Three-Way Solenoid Valve ("Calibration Valve") Design: • Stationary ferromagnetic core with two coils, one for each direction of movement. • A cylindric ferritic arma- ture moving on graphite bearings on the e f a b c d e outside of the core. • Two sleeves at- a: Ferritic core d: Ferritic armature g: Forced circulation tached to the armature, one for control- b: Magnet coil (down) e: Graphite bearings flow ling a coaxial port, another to control a c: Magnet coil (up) f: Natural circulation flow series of radial holes in an outer sleeve. Three-way solenoid valve Application: Mounted at the bottom end of an instrumented fuel assembly it circulation operation, or to thesubcooled used during calorimetric calibration of is used to connect the fuel channel to plenum chamber below the core for for- the assembly power sensors, from Which the surrounding moderator for natural ced circulation. The latter situation is its popular name is derived. 47 Pneumatic Throttle Valve TEST FUEL DATA BANK SYSTEM Design: • A bellows pressurized from an external helium system moves the A test fuel data bank management sys- on NORD-lO/NORD-lOO in such a way valve. • A differential transformer senses tem has been developed based on many that information is sorted according to the valve position. • A coaxial port is years of experience with an IBM-1800 the experiment it belongs to. In this normally closed or fully open. With the computer for data collection and retrie- manner a test fuel data bank library is axial port closed, radial ports can be val. Data in this context comprices both constructed based on magnetic tapes. adjusted at any position between closed signals from in-core irradiation rigs and and fully open. The system can produce diagrams and from the plant, as well as other forms of tables as well as magnetic tape where the Application: Mounted at the bottom information. The data bank management data arc stored in a specified organized end of an instrumented rig, it permits system is developed for NORD-IO/ form. calorimetric power measurements when NORD-IOO. the channel is connected to thesubcooled To aid the user an extensive dialogue plenum chamber below tiie core. With In this system, the use of interactive facility is included as well as extensive the radial ports open, the valve can be terminals and large random access files internal data management, in order to used to throttle the channel inlet for has greatly improved the access to data protect the data and maintain the inte- burn-out studies. in contrast to the system earlier used on grity of the library. IBM-1800. This computer is, however, The test fuel data bank management still used for reading in-core and plant system has been put to use for evaluation signals and building a magnetic tape of experimental results from a variety library as well as performing some data of instrumental fuel assemblies as well as presentation. for experiments performed with special Tapes from this library are reprosessed irradiation rigs. Peripherals (plotters, primers.etc) I To out-of- house users NORD -10 Multi-user • -v. ••. , > Interactive.fuel assembly, Mag. V oriented files J \ tape J Random access Data library Post I Design irradiation 1 Other data examination! informa- data 1 tion IBM 1800 Single user 1 ^ : Batch signal oriented files •v tape J Sequential access L J Mag. V 5 a: Natural circulation inlet b: Forced circulation inlet Oata library c: Natural circulation valve plug d: Forced circulation valve plug S e: Bellow I In-core rig and plant signals f: Helium supply tube j g: Flow meter j i 1 Pneumatic throttle valve Test fuel data bank system. 48 Fuel Performance Studies 1983-88 Calender year Type of investigation Instrumented fuel performance tests Diameter rigs. Comparative testing Power ramp testing He-3 local power control Densification/swelling tests LWR diameter rig (rotatable rods) Power cycling tests. Solid neutron absorbers BWR/PWR simulations Gas flow/fission product release tests LOCA tests. Thermal behaviour Load follow operation Gap meter rig. In-core TC connecting plug BWR/PWR ramping rigs LOCA tests. Ballooning behaviour Interim inspection facility Re-configurated reactor core Ramping, power reactor segments Fuel rod re-instrumentation technique Load-follow operation. (Advanced thermal reactor fuel) Gas mixing experiments Automatic frequency control, re-fabricated power reactor segments Waterside corrosion experiments KHlMSf Fuel rod testing activity in the Halden Boiling Water Reactor (HBWR) INTRODUCTION reactor and connectcd to external sys- FUEL ROD THERMAL BEHAVIOUR tems of high pressure loops with strict Improved understanding of fuel relia- control of water chemistry and thermal Fuel temperature measurements con- bility and safety has continued to be the and hydraulic parameters. Different irra- tinue to be a major experimental activity main motivation for the research work diation rigs Tor use in these high pressure and a large number of centre oxide performed at the Halden Project. The flasks have been developed. This has re- thermocouples have been installed in fuel studies included standard fuel designs as sulted in a range of experiments and rods in the HBWR during the last five well as modified design features aimed studies related to the behaviour of fuel years. at improving fuel performance up to under modern BWR and PWR conditions. High temperature fuel thermocouples high burn-up levels, and at different decalibrate due to neutron irradiation power levels and power level change induced transitions in the thermocouple sequences. The research has comprised materials. As a combined result of post- short term studies of specific problems irradiation recalibration of HBWR-irra- as well as studies aimed at effects of diated thermocouples and evaluation of extended burn-up. investigations reported in literature, a The HBWR reactor core was modified best estimate decalibration correlation early in 1983 to establish a central region has been established, which account for with neutron flux twice the previous parameters such as heat rating, enrich- level. This has permitted irradiation of ment, end pellet design, etc. low enriched fuel segments from com- The temperature behaviour of fuel mercial power reactors at typical power rods made from thermally stable fuel levels. and with small fuel to cladding gap is Presentation of the best estimate for W-Re The research programme has required very little affected by fission gas release. thermocouple decalibration as function of further development of the Project's .ex- thermal neutron fluence, based on experience In helium-filled rods with medium sized perimental capability. A number of high at the Halden Project and evaluation of investi- gaps, fission gas release causes appre- pressure flasks have been installed in the gations reported in the literature. ciable thermal feedback because of con- 49 2000+- • 30 kW/m a. 20kW/m 1600 I i ' i * I I t 8 1200 Time from argon Insertion (days) Argon is introduced at the bottom end of a helium-filled fuel rod in the gas flow rig. The change in fuel centre temperature at constant S 800 • " • - power shown aj a function of time in the figure demonstrates a slow rate of axial mixing of I'M argon with the original gas in the fuel rod. S. «XH tamination of helium in the gap and fuel e void volumes formed by densifisation. T T 12 16 20 24 ~3T Temperature data in the gas-flow rigs Pin mean burnup, MWD/kgLI02 have shown an increasing sensitivity to Development of centre fuel temperature with burn-up in a fuel rod with stable fuel and an as- the helium pressure with increasing burn- fabricated (75 microns) narrow fuel to cladding gap. The slight increase in temperature with expo- up. This is assumed to be caused by cir- sure is due to degradation of fuel conductivity and release of fission gas. cumferential cracking of the fuel. When a rod is filled with argon, it has been FISSION PRODUCT RELEASE found that the temperature will decrease with burn-up for a given power level. and the influence of pre-pressurising the Studies have demonstrated a clear In fuel rods which failed during opera- fuel rod with helium. It is found that the correlation between fuel centre line tion, the fuel temperature would in- mixing of gas in the pellet to cladding temperature and fission gas release in crease immediately upon fuel rod failure. gap with that in the plenum develops the burn-up range up to 28 MWd/kgU02, This could be related to the effect of very slowly, up to several hundred hours, from near zero below 1000°C to around steam on the thermal conductance across while pressure equilibrium is normally 30% approaching 2000°C. the fuel to cladding gap. The tempera- reached in a few minutes. It was found that annular pellet fuel ture was observed to increase by approxi- It is also found that fission gas release rods had significantly lower fission gas mately 200°C at 30 kW/m. Long term and degradation of uranium oxide release than solid ones, other parameters study of this phenomenon was prevented thermal conductivity may result in being identifical. This can be correlated by tWe early failu'e of the centreline centre temperature increases with burn- with the lower temperature of the thermocouples, probably due to its up from 100°C in small gap rods to hollow pellet fuel. reaction with the steam. Future work on 350°C in large gap rods at 30 kW/m. At The gas flow rigs are important devices the subject will require sealed thermo- extended burn-up the heat transfer tends for studying fission gas release. The couple junctions. to improve in larger gap rods, probably studies confirmed that the following Considerable efforts have been direc- due to swelling and relocation. release mechanisms are operating: ted towards the study of the develop- Pellet surface roughness seems to have 40- ment of heat transfer in fuel rods and its negligible effect on the heat transfer as dependence on fission gas release/mixing, Burnup Range of compared to other parameters. 9-28 MWd/UGj 20001 , • • i — . • • • 30 kW/m 4 a a 20kW/m 1600- /'/sfoW* zJ&Z^"! * • !'«•» ?W»„: / -vv. f;W • 1200 f : i i^kiV, icSsfif mr * 800 «Ot 1000 1500 2000 2500 16 20 is il' T Weighted measured lifetime temperature, deg. C Pin mean burnup. MWD/kgU02 Development of centre fuel temperature with burnup in a fuel rod with stable fuel and an as- The figure indicates how strongly the release of fabricated medium (230 microns) fuel to cladding gap. Thermal feedback effects can be seen fission gas depends on the fuel temperature in at low burnup due to fission gas release followed by a gradual improvement of heat transfer the burnup range 9-28 MWd/kgUO2- properties as irradiation proceeds due to gap closure. 50 2000 1600- gtr '» 1200- . . corresponds to W E ' power reduction 800- 2000- 1600- 10 - It can be seen how fission gas is released in a bursts at power reductions during a load follow n Q. 1200- " E experiment. Cas communication is improved by ^^^Trt^rr^—4s I the opening up of the pellet to cladding gap at 800- „ ..^nb, •...,- p. g^^r&Tv; •• p.<.;.-; ,•-; • jo reduced rating. 5 10 1 55 2200 2255 3300 35 Rod average burnup, MWd/kgU02 - diffusion release from exposed sur- faces and grain boundaries The fission gas release is higher in solid peilet fuel than in hollow pellet at normal power ratings, - recoil release from external surfaces most likely due to the higher temperature in the solid fuel. The difference tends to decrease upon power ramping to higher heat rates. - release from fissioning of sputtered uranium. Onset of interlinkage and increased 1MDf release was found to agree well with the From 21.947 MWd/kgUOj • Temperature To 27.701 MWd/kgUOj -Rating empirical fission gas release threshold. BOO At high power levels, 35 MWd/kg UO2, it is observed that released fission products originate from about 20% of the available surface area. During power reductions the products are released from the isolated 80% of the surface area in bursts. Experiments in the gas flow rigs I where various gas combinations can be introduced as required, also revealed that mixing of gases in the pellet to cladding 900 gap is diffusion controlled. This could be observed through its influence on the centre fuel temperature as a conse- 800 Jus i2 ill im in Li)i> It iu Ju ..,2m. ,2b i&a i/2 ire Is quence of the resulting changes in the Assembly overage burn-up (MWd/kgUO,) piotl«d tsss/ovn M£t heat transfer. The fuel temperature increase observed after a large drop in power following a period at ele- On occations burst releases were ob- vated power is caused by a burst of relaesed fission gas. The following decay is caused by served to cause large temperature in- mixing of the fission gas with the helium in the rod, largely through diffusion. creases followed by a decay towards pre- release values. This was attributed to diffusive mixing of the burst gas with -4- -4- -4- the helium in the rod plenum. 8SI TO, SOOjin gap, 6jm grain, i7 kW/m, Rod CB 25- 951 TO, 200um gap. 17»ir« groin, iT Rod CO 95% TO, 360pm gap. I7(in grain. 47 kW/n, Red CH 81% TO, 200(im gap, 17jin grain, 63 kW/n, Rod AB 93% TO. 200fim gap, 17pn grain, 53 kW/n, Rod AD PELLET CLADDING MECHANICAL 20- 95% TD. 200(ira gap, 17jm grain. 53 kW/n, Rod AH INTERACTION A review of pellet to cladding inter- 15-- action effects from various experiments has shown that the as-fabricated pellet geometry and pellet to cladding gap £ ID- width will retain its effect on the mecha- nical interaction to relatively high burn- ups. Hoop strain and ridge formation is S' most pronounced with dished, pellets, and smallest with chamfered pellets, while flat ended pellets have an inter- 100 200 300 400 500 600 700 800 900 mediate effect. Hollow pellets prw&ice Elapsed time at peak power, sec less cladding strain than solid ones. Differences in transient fission gas release as function of design and operational parameters. Factors which also affects the inter- Fission gas release during short-time power excursions increases with increasing rating, in- action between fuel and cladding are the creasing gap and decreasing fuel density and grain size. Measurements show c square root of power change rate and the power level time dependence, attributed to diffusion mechanisms. 51 prior to a power increase. Shutdown size, and power level. A further analysis handling oi the fuel is also of importance. TF*3-552 SCRAM ON 86/2/8 of the rod pressure data for the six Low ramp rates during power increases PWR-type fuel rods showed that parts will reduce the stress levels in the clad- of the fission gas released during the ding. transient came out in a "burst", followed The degree of mechanical interaction by diffusion controlled release. The tends to increase with burn-up in fuel derived diffusion coefficient was con- rods with stable fuel and normal gaps, siderably larger than the commonly and strain relaxation at constant power assumed value, indicating that other is limited. Due to irradiation-induced mechanisms than diffusion may also hardening of the cladding large elastic be active. deformation of the cladding can take place. Changes in cladding dimensions until power in fuel rods with firm pellet EXTENDED BURN-UP OPERATION to cladding contact are typically 0.15 - 0.20% per 10 kW/m. Investigation on the behaviour of fuel During early stages of irradiation the at extended burn-up has been focussed decrease in internal gap is associated on the thermal properties of the fuel, with solid and gaseous swelling as well fission gas release, fuel swelling and as with pellet fragment relocation. The mechanical interaction between fuel and gaps close earlier with increasing burn-up 0 5 10 15 20 • cladding, and waterside corrosion of the time into transient, seconds in hollow than in solid fuel pellets. cladding. BWR fuel segments, preirradiated in a Fuel temperature response of a PWR rod to a The experiments have comprised a US power reactor to burn-ups in the reactor scram at moderate exposure (15 AlWd/ very wide range of parameters regarding range 10 - 15 MWd/kgU, have been ex- kgUC>2)• Dynamic temperature response de- fuel properties, mechanical dimensions posed to power ramps in order to add to pends on design variables (diameter, gap etc.) and shapes, and operating conditions. the qualification of HBWR as ramp test and bumup. In-core measurements have been in good facility. The results were in good agree- agreement with post-irradiation examina- ment with data for similar rods ramped ding deformation increased with burn-up, tion. The experiments also comprise re- elsewhere. while the magnitude of the strain de- instrumentation of a series of irradiated creased with increasing number of power fuel rods with pressure sensors to study cycles. fission gas release at high burn-ups. CYCLIC AND TRANSIENT OPERA- PYVR-type rods, base irradiated at 20 Thirteen fuel rods have been re-instru- TION kW/m to 30 MWd/kgU02, were sub- mented with bellows pressure trans- jected to rapid ramp increases to 50 ducers for characterising fission gas In a special irradiation rig where the kVV/m for study of transient fission release behaviour at extended burn-ups. coolant flow could be throttled, the gas release. Pressure changes were mea- The test matrix included different gap effects of loss-oT-cooIant has been sured after each of five bumps increasing widths, fuel density/structure, and built- studied. Both electrically heated rods in time from 15 to 510 seconds. The in mixed-gas thermal conductivities. Re- and fuel rods were exposed to condi- fission gas release correlated with fuel lease rate as a function of linear heat tions leading to ballooning of the clad- to cladding gap, fuel density, fuel grain load increased appreciably with increas- ding. Rod to rod interaction effects were also studied. The results were con- 12.30 sistent with findings in other tests. Several scram tests were performed to study the thermal response of a variety Power cycling •' 826 cycles 12.29 of fuel rods. It was found that the fuel thermal time constant would increase Ramping with increasing burn-up, as can be ex- \ pected from deteriorated fuel and gap 12.28 conductivity. Calculations using a tran- sient temperature code developed at the High power Project were generally in good agreement with the measurements. Noise analysis 1 12.27 was used to determine rod temperature o response at different ratings and to distin- T> guish between fuel rod and thermo- Low power' couple time constants at different power 12.26 levels. I Scram J Vj A comparative study of the effect of power cycling has been performed under LWR conditions. Rods were cycled up to 12.25 _L _L I i _1_ _L 800 times between .30 and 45 kYV/m 48 96 144 192 240 288 336 384 432 480 with measurements of dimensional changes and fission gas release. PIE and Ramp/power cycling time, hr gas release measurements analysis The figure demonstrates houi the strain in a fuel cladding can increase during a power ramp, and showed little difference between power how deformations zuill gradually decrease when the rod is exposed to repeated power cycles. A cycled and non-cycled rods. The clad- diameter gauge is used to carry out these measurements. 52 ins> gap width and dccrcasin segments prcirradiated in a power reactor " 260 started in 1987 for investigating any 280 300 320 340 36C effects of power cycling between surface Inlet temperature, °C 260 boiling and single-phase cooling condi- Local cladding temperature and coolant tem- tions on the waterside corrosion of the perature as a function of inlet coolant tempera- Zircaloy cladding. ture in the waterside corrosion rig. Departure from one-phase cooling is noted above 280°C. 2.0' Nucleate boiling brings the cladding outside sur- face temperature slightly above saturation at Circumferential distribution of oxide thickness 1.5' prevailing pressure (asymptotic level). on a fuel cladding. 1.0 vs...... 0.5- i 0- —-EN D . i • -0.5 £ a. A, _ 831 TO Fu«l J -1.0- 0.7* Ap at 6 MWd/kgUOj i/> V • "V Q> J\ J V|- j C v ; • \J \ : -1.5 Jo* -2.0 v 'J" ' ' i ' -2.5 ^ tO ON * «T I— * tVV"'* o ,:/ f •3.0 V (D •o J -3.5-! oX 1: 5 10 15 20 25 30 Burnup, MWd/kgU02 Low density fuel will densify carly-in-life v- Fuel active region during irradiation to a given level. High density i ; . . .. . , 1 • ...... 1 . . , . 1 . . . . 1 . . . , 1 . . . . 1 . . . . i . . . . < . •'.••• 1 fuel does not density. The figure demonstrates rate of densification and swelling as function of (Bottom) Position on PN038. a.u. (Top) burnup for rods containing fuel -of different dersitv and structure. Oxide thickness on a Zircaloy clad fuel rod measured with eddy current proximity probe, Computerized Man-Machine Communication Introduction Rcscarch at the I'laldcn Project on the application of computers has evolved from the use of computers for process control into a number of various other applications. These activities fall into two main areas. First, there is Man-Machine Inter- action Research which focuses directly on the role of the control room operator. This research focuses on the Halden Man- Machine Systems Laboratory HAMM- LAB. Secondly, there is the development of Computerized Operator Support Systems which comprize a variety of applications. The cxpcrience with regard to both operator and support system capabilities and limitations will in the future be uti- lized to realize the fully computerized Integrated Surveillance And Control Sys- The Halden Man-Machine Systems Laboratory HAMMLAB. tem ISACS in HAMMLAB. an experimental evaluation of a sophisti- ric guidelines that are used for refinement Man-Machine Interaction Research cated safety parameter display system - of the design itself. The results are 1) a In recent years man-machine inter- the Success Path Monitoring System specific experimental evaluation and 2) a action research at the Project has concen- (SPMS) - has been conducted in collabo- proven methodology which can be app- trated on the role of the human operator ration with Combustion Engineering Inc. lied to experimental evaluation in gene- as advanced computer based technology The system provides on-line procedural ral. is increasingly implemented in the con- advice in the case of a serious, unantici- Although an ideal way to evaluate trol room. Whether innovation in the pated plant event. Thus, it complements systems may well to to measure real ope- control room does, in fact, bring about the alarm handling system. Earlier experi- rators' performance in the actual control the benefits predicted by designers, mental work with the Critical Function room situation, this is not always fea- has provided a major topic for laboratory Monitoring System, the fore-runner of sible, because appropriate incidents are investigation during this period. Indeed, SPMS, was carried out by the Halden fortunately rare and occur unpredictably. it was primarily for this reason that Project at the Loviisa Power Plant in For the purpose of system evaluation, HAMMLAB was established with the Finland in 1983. our approach is thus to simulate real con- NORS fullscope PWR simulator, together The experimental technology for the ditions as accurately as possible, within with comprehensive observation facilities evaluation of advanced man-machine an experimental environment such as the for systematic investigation of operator interface concepts has thus been further Project's full-scale PWR simulator. performance, as its principal tool. The developed. Improvements in the overall This facility, therefore, constitutes the primary intention was to use the simu- performance of the total man-machine focal point for man-machine interaction lator as a vehicle for the installation and system is the objective of the various research at Halden. It can now be stated, development of new information display systems being developed to aid process based on the experience gained from concepts. To this end, a range of tech- operators. experiments carried out so far, that this niques for experimentation were devised As experimental evaluation of opera- laboratory has proven to be a valuable which have been exploited in a series of tor aids is a relatively new discipline, tool in the development and evaluation experimental studies involving the valida- there are few standards or proven experi- of operator support systems for the nu- tion of complete systems and assess- mental methods which can be applied. clear plant control room. ment of the feasibility of new operator The. Project is engaged in defining support systems. measurable objectives for operator aid The Alarm Handling System HALO Research effort has been directed to- systems to enable experimental evalua- wards alarm systems and in particular, tion of specific commercial operator The HALO system (Handling Alarms comparisons of conventional alarm sys- aids. This approach includes feedback using LOgic) has been under develop- tems with modern computer-based sys- from the application and testing of a ment at the Halden Project for a number tems with advanced displays. Recently, particular experimental design, to gene- of years. In 1984 it was installed in the 54 Project's own full-scope PWR simulator RCS HEAT REMOVAL and an experiment was carried out com- SUCCESS PATHS / ' " i 1 i paring performance resulting from usage I \htvv IrviT AVAIL 1 BITE* of one of three computer based alarm mm feed soon | 1 systems. The results of this study revealed MIX FES) | " READY | YES little by way of systematic operator per- SAFETY INJECTION | HEAD* | yes formance differences. However, it was SAFETY EEL VALVE 1 EEfllY 1 YES apparent that, at times, operators found the three level HALO system somewhat •> <— YP12S006 cumbersome. It was therefore decidcd to -> 4 * YP12S008 ART DttP present alarm information on only two liO-BBJ 1IX X K levels. The top level is a dedicated, de- J1 -93 . -- "" tailed overview. All process alarms are filtered on-line before display, removing redundant information. Alarm data are then presented on the overview in two modes: (1) as appropriate colour changes to alpha-numcric text, bar-charts and the *B IE 13 « 2 process mimic; and (2) as a direct refe- LEVEL 3.2 3.2 3.2 3.2 rence to the more detailed process for- n~n PRESS. mats. On the second level, alarm para- meters are again represented within the operating formats by appropriate colour /I Success Path Monitoring System (SPMS) second level format m alarm slate. changes. A further experiment compared per- formance with the enhanced HALO sys- this by providing an on-line assessment cidents at PWR plants. Data collected in- tem against a conventional tile-based of both the status of critical safety cluded video recordings, event logs, para- system. Alarms were detected rapidly functions and the status of appropriate meter plots and detailed notes taken by and the information on the overview led success paths for correcting any threat to the experimenters. to swift and efficient selection of rele- the critical functions. SPMS was origi- vant process formats for the disturbed nally developed under the sponsorship sections of plant. Operators commented of the Electric Power Research Institute favourably on the overview display and System for Early Fault Detection as the final phase of the EPRI Distur- found the detailed alarm information bance Analysis and Surveillance System The work on early fault detection sys- clearly displayed and easy to use. In con- (DASS) development project. tems started at the Halden Project in trast, operators in the conventional A prototype version of SPMS has 1985 for the following reasons: annunciator tile condition experienced been implemented on the Halden Pro- Traditional process alarm systems had difficulties in relating alarms to process ject's PWR simulator. The principal goal been found to have a limited capability formats, and the lack of embedded of this implementation is to make avail- in coping efficiently with dynamic pro- alarms within these formats also contri- able a system which could be validated cess situations. One disadvantage with buted to a negative effect on perfor- by means of a man-machine evaluation fixed alarm limits is that after the occur- mance. experiment at Halden. The incentive to rence of a failure it may take a long time A number of general conclusions of conduct such an experiment stems from before the alarms are triggered. Therefore value to control room designers can be a desire to demonstrate the effectiveness an early warning system has been deve- drawn from this study: (1) The feasibi- of the system and to identify ways in loped. It is based upon running small lity of installing a computer based alarm which it might be improved. The evalua- process section models in parallel with system on a scale comparable to a real tion experiment was therefore considered the process. The model outputs are then power plant has been amply demon- as an integral part of the development compared with respective plant measure- strated. (2) If a computer based alarm process and, as such, will indicate the ment. system is to be installed, it should not direction for future work. Fault detection based on a number of be limited to providing textual alarm The basic objective of the experi- reference models has been demonstrated lists: Operators in these experiments ex- ment was to test the SPMS in a realis- on the feedwater system of the Project's pressed a clear preference for symbols tic situation and thereby to assess PWR simulator NORS. There are two and colour in alarm presentation. (3) In whether it performed in accordance with main features that characterize this ap- both mixed and computer based control design expectations. The experiment proach. First of all, it provides the possi- rooms, the alarm system should be an follows on from the series of experimen- bility of detecting faults before a tradi- integral part of the computer based tal evaluations already carried out by the tional alarm system is triggered, even in operating system. There was no experi- Halden Project. A number of reactor dynamic situations. Secondly, by a pro- mental human factor evidence to show a operators have been observed in the ex- per decomposition scheme, the problem need for separate tile panels. perimental control room coping with area can be confined to the particular complex transient scenarios both with, reference models covering the faulty pro- The Success Path Monitoring System and without SPMS. Each session lasted cess part. A two level display hierarchy SPMS about three hours, during which the ope- has been chosen, where the warnings are The Success Path Monitoring System rator was closely observed coping with given with colour symbols in a top level (SPMS) is an advanced computer based the process during a severe transient picture (including time history) with a operator aid which is intended to en- scenario which was deliberately chosen global overview guiding the operator to hance the operator's ability to handle to exploit the features of the SPMS and the lower level detailed displays contain- plant disturbances effictively. It achieves had aspects in common with reported in- ing much more information including 55 The Core Surveillance System SCORPIO In the last halT decade the continued Dfeviat ion *Erom Expected efforts in developing and demonstrating an advanced core surveillance system have brought results and improved con- tact with nuclear power plants. The capa- bilities for predictive simulations and core monitoring have now been utilized for pressurized water reactors at several power stations in the member countries. The real breakthrough for the Halden Project's core surveillance system came when Ringhalsverkct in Sweden decided to acquire SCORPIO. The Predictive System was installed there for one PWR in 1934 and for the two other PWRs in 1985. In 1985, a demonstration of the Predictive System was held in the USA (near Hartford, Conn.) in co-operation with Norsk Data and Combustion Engi- neering. Vattenfall had, at an early stage, indi- cated that they were also interested in obtaining the Core Monitoring (Core Follow) Part of SCORPIO. In December Early Fault Detection (EFD) overview display. Deviation from expected behaviour. Two faults 1987 the installation of the full SCOR- in high pressure pre-heaters detected and resolved. PIO system for one PWR took place in Ringhals. Only a month earlier a Pre- dictive System modified for the fault localization and possible explana- lopment of advanced electronic graphic Catawba 2 Pressurized Water Reactor tion in a message field. systems for use in real-time applications, had been installed at Duke's head- Reference models arc now being eva- i.e. process control and research-/training quarters in Charlotte, North Carolina. luated with real data from the Loviisa simulators. These systems have consti- Here the system will be evaluated over Nuclear Power Plant in Finland, in order tuted an important tool in the Project's an eight month period. to ascertain how well they follow normal man-machine research activities and con- The purpose of SCORPIO is to meet process dynamics. tributed significantly to the Halden Pro- the need expressed by utilities to im- ject's position in the forefront in this prove the knowledge of core status research. The outcome has been a num- both in steady state and transient opera- System for Plant Diagnosis ber of practical industrial applications ting situations. The Predictive System The Diagnosis System using Know- within the nuclear industry as well as is an efficient tool for calculating the ledge Engineering Technique (DISKET) in other areas. Further, the display soft- behaviour of the core in xenon transient developed by Japan Atomic Energy Re- ware package, PICASSO, has been avail- situations for up to 48 hours. It enables search Institute is specially designed for able for use by the signatories. controller strategy generation for desired description and recognition of nuclear power plant accidents. Characteristic changes of important parameters during a transient are store as finger prints in a knowledge base. By structuring the acci- dent hypotheses in three levels of detail, the search space is limited, while still keeping the feature of analyzing parallel problems. At the Halden Project DISKET has been implemented on an ND-570 computer, and a knowledge base for diagnosis of malfunctions introduced on the NORS PYVR simulator is being built up. A proposal for presenting the results of a diagnosis to operators has been made where a ranking of the hypotheses and geographical location of the problem area are shown. The rules supporting a hypothesis are presented as a window in the same picture for detailed explanation. The Process Operator Station CAMPS The Halden Project has. for the last twenty years been engaged in the deve- The operator's workstation for core surveillance with SCORPIO. 56 power variations. The Monitoring Sys- tem gives the reactor operator detailed information on the margins to operating limits and the core power distribution. At Ringhals, gamma-thermometers have been installed in addition to the standard installation, although this is not a re- quirement of SCORPIO. The high resolution colour graphic work station CAMPS designed at the Halden Project is now used for the visualization oT SCORPIO's information to the operator. A programmable touch panel function keyboard, a standard alphanumeric keyboard and a tracker ball are used for operator input. Training SimulatorsforOff-shorc Oil/Gas Production Platforms The mathematical modelling and sim- ulation activities originally tied to pro- cess studies at the HBWR, and later ex- panded to studies of other power plants Operation of the GULLFAKS training simulator. Lhrough development of operator support systems, like the core surveillance sys- tem SCORPIO, have resulted in a high IFE. The simulator was delivered to heim and Narvik of the Norwegian Power competence level in the simulation STATOIL in September 1985 after a 2 Pool. field at the Institute for Energy Tech- years development period during which Such a simulator can be a powerful nology (IFE). This competence, to- a total of about 60,000 manhours were tool in order to obtain a production gether with the experience gained in spent on manufacturing the simulator. schedule, a maintenance performance development and operation of the The Oseberg A simulator was delivered plan, a reserve allocation and a network Halden Man-Machine Laboratory in- to Norsk Hydro in September 1987 configuration that can, at any time, cluding the advanced control room and after an 18 month development period. meet the demand for security, quality the full-scale PWR-simulator, has served This project was carried out as a joint and economy. Also in this project the as a basis for large bilateral projects venture between IFE and Norcontrol experience at IFE on operator communi- carried out by IFE for the Norwegian Simulation, and IFE carried out about cation, simulator techniques and soft- off-shore industry. The delivery of the 60% of the total development effort of ware structures provides a strong basis. full-scale, replica training simulators for approximately 70,000 manhours. During the summer of 1988 a total the North Sea oil/gas production plat- Through these projects, IFE's know- system design specification will be com- forms "Gullfaks A" (operated by STAT- how in the simulator field has been pleted. On the basis of this specification, OIL) and "Oseberg A" (operated by greatly enhanced, and this will benefit the Norwegian Power Pool intends to Norsk Hydro) represents the hitherto further work in this area, both within order delivery of SASIM, and in that highest achievements by IFE in the the joint Halden Project programme and event, the total SASIM project (covering simulation field both technologically in future bilateral contract work for system specification and realization) will and in terms of volume of work. organizations in Signatory countries. be the largest bilateral project achieved Both simulators are characterized by by IFE. extensive process models, replica con- The Power System Simulator SASIM trol rooms and advanced instructor Control Room Design systems. The response of control systems In the last couple of years a new pro- and alarm/shutdown systems is identical ject, SASIM (Samkjoringens Kraftsystem* The pioneering work on control room to the corresponding systems on the simulator - Power System Simulator), design with computer based instrumenta- platforms. This is attained through de- between the Norwegian Power Pool tion initiated in the end of the 1960's tailed emulation of the process control (Samkjoringen), IFE and the Electricity and continued in the computer labora- systems installed on site. Both simulators Research Institute (EFI) has been gain- tories during the 1970's has, in recent are used for training of operators and ing momentum at IFE. years, been complemented with design other platform personnel, and the sta- The aim of this co-operation is to work for industrial applications. Sys- bility and reliable performance of the design and develop a power system sim- tems developed in the computer labora- Lraining simulators gives credit to the ulator for operational planning in a tories, have thus been tested at various technical level and workmanship of IFE system dominated by hydro-electric plants for long periods of time. The ex- staff. Both projects have involved a power plants. This involves modelling of perience gained has been most valuable sizeable number of staff with software a 400 bus network, some 550 power for the design of new control rooms and and hardware expertise and required an stations and 800 water reservoirs and control room retrofitting. effective system for project planning catchments. Also economic models of Industrial applications have been de- and monitoring. firm and spot power markets are covered. signed and implemented both in Norway The Gullfaks A simulator was con- SASIM will be implemented as a distri- and other Project member countries, in structed as a joint venture between buted computer system to serve the four industries as diverse as nuclear power Kongsberg Vapenfabrikk and Norsk Data/ control centres in Oslo, Bergen, Trond- generation, oil production and refine- 57 ment, electrical power distribution, chemical industries, maritime applica- tions and a satellite tracking station. Within the multiple aspects of control room design, work has been centred around control room layout and environ- ment, control desk design and the adap- tion of the operator/process interface to operator tasks. Methods have been deve- loped to elicit and handle sources of complaints and bottlenecks in control rooms due for retrofitting and to miti- gate, if not eliminate, them in new de- signs. Merged with the on-going ad- vanced research in the Project labora- tories on new generations of operator/ process interfaces and work stations, the profound pool of experience and background knowledge now possessed by the Project will meet challenges in control room design well into the next The Symbolics computers are well suited for development of advanced operator support century. systems like the COPMA system. The Computerized Procedure Manual An important feature of the system is of the software implemented in safety System COPMA the possibility to read actual status or related computer systems. The Halden values from the process. The system Project has, therefore, been actively Written procedures provide an impor- evaluates and proposes the next proce- working in the field of software reliabi- tant reference point for operators in dure execution, which can, however, be lity for the last decade. One particular many process plant control rooms. over-ridden by the operator, if necessary. topic has been to study the possibility The scope of the project on compu- Hence, it is always the operator who ini- of enhancing reliability by the use of terized procedures is limited to the de- tiates the proceeding. The operator gets software diversity. This work began as velopment of a computerized system feedback when conditions are fulfilled or a joint project with the Technical that can handle todays' written proce- targets reached. Research Center of Finland (VTT), and dures in the context of an advanced Integration of this tool with the later continued with the PODS (Project computerized and CRT-based control other computerized support systems in On Diverse Software) project, in colla- room. It is not concerned with the use was an important design objective. boration with the Safety and Reliability actual construction/design process of This is solved by providing the oppor- Directorate (SRD)/the United Kingdom procedures. tunity to access relevant display formats Atomic Energy Authority (UKAEA) The motivation for initiating the re- directly from the procedure step display. and the Central Electricity Research search program was that many organiza- Components and automatics can also be Laboratories (CERL)/the Central Electri- tions have identified a number of prob- operated through the system if the pro- city Generating Board (CEGB) in Eng- lems concerning present day manual cedure step requires such manipulation. land. VTT in Finland and the Nippon procedure systems. These include, for The main system features are: multi- Atomic Industry Group Co. Ltd. (NAIG) example, identification of the correct window and multi-process man-machine in Japan. Results from this work were procedure, time to locate procedures, interface, use of AI techniques, such as utilized in a new project with the same time consuming data gathering when the Symbolics languages Prolog and Lisp, partners, the STEM project, which aims following procedures, difficulties in ob- and Lisp machines, such as the Symbolics at investigating a variety of methods for serving whether the initiated action has 3640 and the Texas Instrument Explorer. testing and other forms of evaluating the desired effect, problems with mul- The protocol TCP/IP is used for com- software. tiple disturbances, problems when mov- munication between COPMA and nume- A clear finding from the work in this ing from one procedure to another, etc. rical process computers. area is that ambiguities and errors in Our aim with the computerized pro- In ordei^ to further explore these the specification are a common source of cedure system is to assist operators in issues a joint program between ENEA, program faults. The Project has therefore, central control rooms in identifying Rome, and the Halden Reactor Project is in co-operation with VTT, developed a the relevant procedure, as well as in exe- currently in progress. system, X-SPEX, which assists in the cuting that procedure, once chosen. We At ENEA laboratories, the COPMA specification of real time computer pro- also believe that providing quick access system will be linked to a PWR nuclear gram systems. to stored procedural information and power plant engineering simulator where Safety related computer programs relieving operators of trivial tasks in the research aim is a structured analysis must also be licenced by the authorities. connection with executing procedures of a sample of operating procedures. The Project has therefore been develop- such as collecting data, waiting for ing programs which act as tools for the responses and doing response checks, are analysis of" safety related programs in important additional advantages. the form in which they are actually Software Reliability COPMA is an on-line computerized stored. This work is being carricd out in system that enables operators to retrieve The introduction of computers for the SOSAT (Software SAfcty Tools) procedures from a procedure database the control of nuclear plants also has project, which is a joint project, with and follow one, or several procedures in safety-related implications. A particular Tt)V Norddcutschland and GRS parallel, through an interactive session. problem is to guarantee the correctness Garching. 58 Fuel Research Facilities Helium Heat exchanger Central tube for cabling Pressure flask Triangular tube for cabling Tube for water down flow Fuel rods . Main circulation pump !. Flow regulating valve I. Electric heater Main crcuit cooler i. Regenerative cooler i. Purification circuit cooler '. Filter I. Ion exchanger The light water high pressure test facilities in HBWR need a complicated outer flow circuit in I. Helium pressurix order to simulate not only the temperature and pressure, but also the water chemistry typical I. Helium pressurizing tank of boiling or pressurized water reactors. The figure shows schematically the jnain flow and puri- D. Water supply pump fication circuit typical for a high pressure light water loop. HIGH PRESSURE LIGHT WATER LOOPS IN HBWR The possibility of fuel testing under - pellet/clad mechanical interaction stu- tion, the other for fuel/clad mechani- pressure and temperature conditions dies cal interaction studies under BWR con- typical of pressurized and boiling water - fission gas release mechanism ditions. reactors has existed in the Halden Reac- - cladding waterside corrosion on PWR Loop 6 for waterside corrosion studies tor for more than ten years. The in-core fuel rods under controlled surface on PWR fuel rods subjected to surface instrumentation originally developed for boiling and one-phase coolant flow boiling/or one-phase coolant condi- the Halden Reactor has been used conditions. tions. successfully for testing conditions up to To-day four high pressure loop 155 bar and 340°C. All loop systems are to-day operated systems are in operation: Since 1974 the high pressure light with a lithium addition of approximate- water test facilities have been used for - Loop 1 for studies of fission gas ly 1.35 ppm in order to control the water comprehensive testing programmes release on reinstrumcntcd fuel rods pH at approximately 10.3 (25°C). This which include: under BWR conditions. gives a water pH which is suitable for the - Loop 4 for fission gas release and PWR corrosion experiments in Loop 6 - base irradiation of BWR and'PWR fuel fuel/clad mechanical interaction stu- and it will also avoid deposition of crud rods to high burn-up dies on BWR fuel rods operated in on the BWR fuel rods tested in the other - daily load-follow operation hourly and daily load-follow and three loop systems. - power ramping and automatic frequ- automatic frequency control experi- Except for the loop water treatment ency control operation ments. plant which is situated in the Olavs Hall, - Loss of Coolant Accident (LOCA) - Loop 5 with two separate in-pile the entire loop system is situated inside experiments pressure flasks, one for base irradia- the reactor hall. The loop will therefore 59 have to be remotely operated an.d con- trolled from the Experimental Control Room. The loop control panel indicates the operating pressure and pressure drops, flow rate of main circuit and purification circuit, coolant temperatures, water conductivity etc. All the loop pumps, the remote controlled values as well as special operation such as fuel rod and diameter gauge axial movement is performed from the Control Panel. Normally the loop operational condi- tions, i.e. flow, temperature and pressure are controlled automatically within very narrow limits by means of a Programm- able Logic Controller (PLC). This Con- troller is also used to operate the helium- 3 pressurizing system for local flux control. However, it is also possible to disconnect the PLC for a manual control from the Control Panel. Part of the Experimental Control Room showing from left the control panels for Loop No. 1, PWR Corrosion Loop No. 6, on-line plotter, S-pen recorder. Loop No. 5 control panel and PWR WATERSIDE CORROSION the panel for I'l.C automatic control of loop operation. TEST LOOP The PWR waterside corrosion test facility in the Halden Reactor (Loop No. 6) has been in operation since April 87. After a period of loop and test rig functional testing and to demonstrate that the required water chemistry could be obtained, the loop has been used for studies of the difference in corrosion behaviour of fuel rods subjected to surface boiling and one-phase flow. The present experiments in Loop 6 require a strict control of the coolant conditions and fuel surface tempera- tures. The water pH is kept between 10.2 and 10.3 (at 25°C) and the hydro- gen addition is controlled between 3.5 - 4.0 ppm for suppression of radiolytic oxygen. The water chemistry in the corrosion test loop is controlled by daily water sampling and analysis. These include measurement of lithium, iron, chromi- um, nickel, cobalt and other elements by means of atomic absorbtion spec- trometry, and measurement of hydro- gen, oxygen, water pH and total crud Operating the loops with strict water chemistry specifications requires more sophisticated content. methods of analysis. Here the newly installed atomic absorption spectrometer is used for mea- In the fall of 1988 the loop will be surement of water lithium content. modified with a new purification circuit and will have a separate water treatment plant for operation with _ a lithium/ ler loop coolant conditions are automa- tic operation of the main flow regulating boron content up to 10/1200 ppm tically controlled within very narrow valve, and the coolant temperature is respectively. limits, also during power cycling. kept constant by automatic control For loop pressure control the PLC of the loop electric heater. AUTOMATIC CONTROL OF EXPERI- actuates a set of values to increase the The PLC is used also for helium-3 coil MENTAL CONDITIONS helium pressure in the pressurizer. In pressurizing for local flux control of case'- of a loop leakage or during water individual test rigs. Hence daily load- The operation of four high pressure sampling the PLC will automatically follow operation as well'as the automatic test loops, each with specified flow, start the feed pump to maintain a frequency control experiments which temperature and flow has led to the constant water level in the prcssurizer. require power fluxtuations of ± 10% per installation of Programmable Logic Con- The coolant flow in the test rig is minute over a period of several weeks troller (PLC). By means of the control- maintained at a pre-set value by automa- have been performed by means of PLC. 60 The new helium-3 system for local flux control in operation since 1987, uses a compressor for pressurizing the coils surrounding the fuel rods. With this new system a much higher pressurizing capacity compared to the system pre- viously used is obtained. In order to control the rate of pres- sure change in the coils a set of magnetic valves is operated to direct helium 3 through nozzles with different flow restrictions. The system is therefore suitable for automatic control by means of the Programmable Logic Controller. The new system has been used for daily load-follow experiments as well as for hourly flux variations. During these experiments the flux is varied from 50 - 100% in three minutes and is repeated every hour for periods of several weeks. For the Automatic Frequency Control (AFC) Experiments the new Helium-3 Printed circuits are part of the programmable logic control system. Here a printed circuit is being installed for controlling the Automatic Frequency Control (AFC) experiments system has been used to vary the flux by ± 8% per minute, also for periods of several weeks duration. During these BBsl^'SaSSe experiments the helium 3 is evacuated aaBEffjansE from an upper coil while the lower coil is being pressurized and visa versa. The new system for helium-3 local flux control can be used for irradiation experiments in the Haldcn Reactor con- ditions as well as in the light water loops. It has proved to be very reliable, and it can be programmed to control a great variety of flux variation both regarding Recordings from an Automatic Frequency Control (AFC) experiment: • 1/2: Neutron frequency and amplitude over long detector signals for upper/lower fuel rod u 3/4: He-3 coil pressure, upper/lower fuel periods of time. rod • 5: Reactor power • 6/7: Upper/lower fuel rod diameter IN-CORE INSTRUMENTS During the period 1983-1988 con- siderable effort has been spent on im- proving the existing range of in-core instruments. Metals, ceramics and enam- el are the only materials which can be used for in-core devices in the HBWR because of the radiation environment and the high temperature water. This has worked well in the normal heavy water environment and irradiation flux met in the HBWR. However, with the introduction of the new high pressure and temperature loops designed to satisfy the conditions found in modern water reactors, also including higher flux levels, these devices had to meet tougher demands. This has resulted in replacing the enamel with plasma sprayed alumina. The Sauereisen used Calibration. for immobilizing coil wires has been re- placed with a more neutral cement, and efficient drying out of sensors before sensors. A careful aging process has been a condenser discharge technique has final seal welding. In addition they are introduced to reduce creep during opera- been developed for welding coil wires to filled with helium to permit a complete tion. However, one problem, which is instrument cable leads. leak testing of all welds. difficult to overcome, is creep due Special equipment and techniques Considerable effort has been spent on to stress relaxation caused by neutron have also been implemented for more improving the bellows based pressure irradiation, which results in a signal off- 61 RBOCIW j cwitftr Es cm shroud J Stationary fuel rod I *'j®" ||gBnk TWL* tube Fuel rod B: Welding of instrumented end plug to the fuel rod Electrode flostc Moveable fust rod /n A J ju y mttttttt?;;; ^ Instrumented end plug Cross section of the waterside corrosion with the two test channels, pressure vessel and instrumentation. C: Drilling hole from instrumented end plug into the fuel rod j set. However, it is found that a pressure i Drill sensor can be recalibrated in-core by ftib m;/;;//;///////.. study its response during heating up of i the reactor after a shutdown to low tem- i perature. By applying the gas law the sensor offset can be determined, assum Connecting an end plug containing a pressure sensor to an irradiated fuel rod. ing that its sensitivity is unchanged. A technique has been developed and Ara, JAER1. The level gauge rig contains equipment implemented, whereby a a heater cable and four metal sheathed fission gas pressure sensor can be attached cables containing a scries of alternating to an irradiated fuel rod without releas- thermocouple junctions. The number ing any of its fission gas to the environ- and location of junctions are arranged ments. A new, sealed end plug contain- according to a binary system permitting ing a pressure sensor bellows and a drill indication of sixteen different liquid is welded to the end plug of the old rod. levels. Then a torque is applied to the drill The Project is also studying other through an external magnet system, and ways of applying the heated thermo- a hole is drilled into the fuel rod. couple principle to level indication. A level gauge based on heated thermo- Equipment comprizing an eddy current couples has been manufactured for the proximity probe has been acquired for Dodewaard reactor. It is based on a de- the purpose of cladding surface oxide sign named Bicoth, developed by Mr. K. thickness measurements. The measure- ments are performed in a compartment Computer CIBM-S/1) outside the reactor. IRRADIATION RIGS Most of the rigs used for fuel irradia- tion in the Maiden Reactor has been re- designed and adapted Tor use in the high pressure flasks introduced for fuel stud- ies under modem BWR and PWR environ- gas mixtures permitting extensive stud- mental conditions. ies of heat transfer, gas transport and The range of irradiation rigs available mixing, and fission gas release for HBWR conditions comprize: - Gap meter rig, for measuring fuel/ - Base irradiation rig, for base irradia- cladding gap tion at controlled power rate, while and rigs for modern BWR and PWR con- performing deformation, temperature ditions: and fission gas release studies - Power ramp rig, for experiments at - Multi-purpose rig, for long term stud- various power conditions, including ies of thermal and fission gas release ramps to study fuel/cladding inter- properties action and fission gas release - Diameter rig, for fuel/cladding inter- - Base irradiation rig, for base irradia- action studies under controlled power tion at controlled power combined conditions with fission gas release studies, and - Overpower rig, for special power ramp fuel temperature and cladding elon- and cycling experiments combined gation measurements with thermo/mechanical measurements - Waterside corrosion rig, to study clad- - Gas transport rig, where fuel rods can ding corrosion at controlled power Oxide layer thickness measuring equipment. be pressurized and flushed with given and water chemistry conditions. TEST FUEL DATA ACQUISITION AND PROCESSING Computerised data acquisition started ND-500/CX. A twin CPU ND-5902 sys- within minutes as well as locus on details in 1967 with the Project's first full-scale tem has been installed in 19SS to pro- during, e.g.. power ramps. process control computer, an IBM 1800. vide additional computing power. The TFDB system is running on the During the first years the measured data Also in 198S, the faithful IBM-1S00 Fuel Division's ND-530/CX computer could be kept for 48 hours and were has been replaced b\ an IBM Series/1 and has been further developed in close more permanently recorded as printouts process control computer running a co-operation with its users. It provides only. Regular and reliable long-term data completely new data acquisition and functions such as storage on magnetic tapes began in the reactor supervision system. Work is - regular updates of the 1FA data files last quarter of 1971. Since then an ave- going on to complement and improve with new data from the reactor rage of 800 signals has been logged every this system with software and hardware - survey plots of all signals fifteen minutes and accumulatedon tapes developed by the Project for process - elaborate data screening and manipu- containing the data ol'a quart crofa year. supervision and control. lation functions together with last In addition, several thousand CAL1B runs For a long time, the principal source graphics presentation (scanning with intermediate frequencies for data analysis was the "Test-Fuei-Data- - printed data reports (TFDR) to follow certain experiments for a limi- Rcport" (TFDR), a printed record of the - export tapes hir data transmission to ted amount of time) and more than Len regular fifteen minutes logging. Obvious- participating organisations thousand FAST-SCAN runs (high frequ- ly, this largely manual treatment could At present, some 850 megabytes of ency logging during, e.g., deliberate reac- not keep step with the daily data flow. A fast acccss disk storage are reserved lor tor scrams or rod diameter scanning) first version of a data management sys- the data-bank system. In addition, the have been recorded. tem known as "Tcst-Kucl-Data-Bank" data of IFAs not being worked with are The Project has always sought to use (TFDB) was finished in spring 1985 and kept on tapes. More than one hundred modern, reliable hardware for its com- has become an indispensable tool for the 1FA loadings have been treated so far by puter based tasks. Data processing and data analyst. Its graphics presentation the TFDB, and many old data have been analysis was in 1984 transferred from a module VISION can provide overview re-examined. Norsk Data ND-10/50 computer to an plots of several years' worth of data Data Acquisition IBM-1800 -1988 IBM-S/1 1988- OOOOOOOO OOOOOOOO Jjftfr Reoctor Power History * wuUU P .from Dec. 87 to. Feb. 03 C20 1* j Data Processing > T Z 1 s 1 ND-530/CX < VISION rPDB 450 ,» n j j. ND-5902/CX GO 2 I Si io 3 • 14 l Tin-.e.v.. OOOOOOOO < ^OB 290 H< OOOOOOOO D i TFDR TFDB 130 _l graphic terminals 111 D taser printer li. line printer I (- tape drives (J) EXPORT LU TAPE 63 REPRESENTATIVES TO TI1E 1IALDEN BOARD OF MANAGEMENT AND THE HALDEN PROGRAMME GROUP, 1958 - 1988 In considering the name-list below it should be borne in mind that many hundreds of scientists and engineers from the member countries have actively assisted the Board and Group by attending their meetings and by supporting the technical work of the Project through participation in the annual enlarged Group meetings and through contributions to the numerous specialist workshops arranged under the auspices of the Programme Group. HALDEN BOARD OF MANAGEMENT USAEC/USNRC OECO • Nuclear Energy Agency EURATOM THE NETHERLANDS 58-62 Joseph R. Quinn 58-64 Jules Gu HALDEN PROGRAMME GROUP OECD Nuclear Energy Agency FED. REP. OF GERMANY NORWAY USAEC/USNRC Senior Advicers to the Group: 65-69 Wolfgang Braun 58-59 Lars H. Prytz 61-64 Dale Babcock 70-87 Stefan Krawczynski 59-63 Emit Janscn 65-67 William R. Voigt 58-61 Einar Sicland 70-85 Dietrich Buncmann 63-66 Stcinar Aas 74-82 William V.Johnston 59-66 Leslie W. Boxer . 85- Werner Bastl (GRS) 66-68 Hcnrik Ager-Hanssen 81-82 Raymond DiSalvo 73-77 Klaus Stadie 85-87 Manfred Capelle (Utility) 69- Steinar Aas 82-85 James A. Norberg Assistant Advicers: 85-87 Horst Hcckcrmann (Utility) 69-72 J on Berg 85-87 Daniel B.Jones 85- Werner Alcite (KWU) 73-78 Jan Magnus Dodcrlcin 86-87 William G. Kennedy 66-71 Peter Oliver 85- Heinz Knaab (KWU) 78-84 Kjell O. Solberg 87- Leo Bcltracchi 71-73 JaequCb Roycn 1985 JonTveit 73-76 Nico DcBaer 85- Olc Gjcrde Combustion Engineering 76-77 Peter Oliver FINLAND 73- Robert Duncan 77-81 Michael Stephens 59-65 Svcn-Olof Hultin SWEDEN 81-84 David F. Bessette 60-66 Osmo Ranta 85-86 Ralf Landry 66-79 Olavi Vapaavuori 58-6] Gunnar Ilolte Electric Power Research Institute 86-87 W.L. Riebold 73- Jarl Forst(!n 62-66 Peter Margen 66-81 Hilding Mogard 74-78 Edward Zcbroski 1987 JohnCaisley 81- Bjorn Wahlstrom 78-85 J.T. Adrian Roberts 87- Ralph Caruso 68-81 Pchr E. Blombcrg 81-85 Christian Griislund 78-84 Alexander B. Long ITALY 85-87 Uno HSkansson 85-88 David Franklin AUSTRIA 64-82 Alberto Pcdrctti 85- Kerstin Dahlgrcn 88- Rosa L. Yang 67-69 Giancamillo Ambrosini 87- Jan Mattson 58-66 Michael J. Higatsberger General Electric 67-72 Hcinrich Schmidl 82-87 Memmo Di Bartolomeo 76-81 Helmut Roggenbauer 83- Carlo Lcpscky 71-74 Mark Lyons 88- Antonio G. Fcdcrico SWITZERLAND 74-80 Richard A. Proebstle DENMARK 80-83 J. Samuel Armijo JAPAN 58-60 Gilbert Psarofaghis 83- Herman S. Roscnbaum 58-64 Paul L. Olgaard 61-67 Jean-Liic Meylan 64-66 Bjarne W. Micheelscn 68-76 Junichi.Miida 67-70 Otto Schaub 67- Niels Hansen • 76-77 Atsuyoshi Morishima 88- Lars'Lading 77-84 Michio Ichikawa UNITED KINGDOM 84-86 Tazuaki Yanagisawa EURATOM 86-88 Satoru Kawasaki 58-59 Compton A. Rennie 86- Atsuo Kohsaka 59-60 B. Terence Price 58-60 Roeiof Houwink 88- Teruo Furuta 61-65 John E.R. Holmes 58-62 Sergio Bcrtoletti 65-71 David O. Pickman 60-64 Maarten Bogaart 66-71 Douglas English THE NETHERLANDS 79-88 J. Stanley Waddington 64-67 Wilhelmus W. Nijs 81- Derek S. Hiorns 67-87 Robert Swanenburg dc Vcyc 88- J. Anthony Turnbutl 84- Nico Bruens 64 RESEARCH STAFF ASSIGNED BY MEMBER ORGANIZATIONS 7 he list below includes, with very few exceptions, research staff who have worked with the Project for periods of 18 months or more. Space does not permit us to include the hundreds of shorter term guest scientists and trainees which have been assigned to the Project by its member organizations for exchange of experiences and ideas. AUSTRIA 67-71 Heinz GUnther Walger ITALY 69-69 Ebcrhard Frenkel 59-61 Heinrich Schmidt 59-60 Giancamillo A. Ambrosini AEG SGAE CNEN 69-69 Peter W. Schabacher 60-63 Kurt Schcnk 60-67 Valerio Tosi AEG SGAE CNEN 69-70 Gerhard Dodt 63-64 Alois Hold 65-67 Tullio Bozzoni Siemens SGAE CNEN 70-70 Peter Petersen 64-66 Fritz I. Topfner 67-68 Francesco Doria GKSS Graz University CNEN 70-71 Klaus Frtthner 64-67 Armin Rumpold 67-68 Sergio Griffoni GKSS SGAE CNEN 70-71 Peter Grziwa 64-67 Kurt Schcnk 67-68 Bruno Rimini KFA-JUlich SGAE CNEN 70-71 Jiirgcn Horber 64-70 Friedrich Griess 67-69 Giulio VaUi NUKEM SGAE CNEN 71-71 Eckart Schwieger 66-68 Christine Schittenhdm 70-73 Vittorio Albergamo GKSS SGAE CNEN 71-72 Ernst Robinson 68-74 Helmut Roggenbauer 75-7 7 Carlo Vitanza GKSS SGAE AGIP 71-72 Walter Seifritz 77-79 Rudolf Haubert 77-78 Teresio Busi Tcchnische Universitat Hannover SGAE ENEL 79-79 Peter Fasko 71-73 Johannes Cberall AEG 77-78 Ugo Graziani SGAE NUCLITAL 79-79 Josef H. Zaunei 72-73 Hansjochen Jagersbergcr KFA-Jfflich 79-81 DecioZorini SGAE CNEN 50-80 Hans Pfau 72-73 Dieter Kaspat SGAE NIS 30-81 Franz Dworzak 72-73 Heinrich Otto Siewers SGAE GKSS 73-74 Henning Martens JAPAN DENMARK KFA-JUlich 74-75 Jens Milllcr-Roos 66-70 Makoto Ishizuka GKSS JAERI 59-60 Carl F. Hojerup 67-69 Hiromitsu Tada RisB 74-75 Ivar Ruyter KFA-Julich SUMITOMO 50-61 Paul la Cour Christenscn 69-71 KatsuyukiAra Riso 74-78 Ernst Robinson GKSS JAERI 51-63 Aksel Olsen 69-71 Kazuo Sato Riso 75-75 Rainer Eberlc Tcchnische Universitat Hannover JAERI 64-68 Uffr Scot Jorgenscn 71-74 Toshiaki Tobioka Regneccntralen 75-7C Walter Hanke ' KFA-Julich JAERI 57-69 Jan Boning 72-72 Toyoji Wada Helsingor SV 75-79 Rainer Grumbach GRS Toshiba 67-69 Carsten Dige Gronberg 72-73 Michiya Serizawa Riso 80-81 Wolfgang Wiescnack Technischc Universitat Hannover Hitachi 70-74 Gilbert Fayl 72-74 AkioToraishi Riso JAERI 75-77 Jdrgcn Aukdal 72-72 Yoshihiko Iwano Riso FINLAND Toshiba 80-83 John Friisjenscn 73-76 Michio Ichikawa Riso 59*61 Olavi Vapaavuori JAERI 82-85 Eric Hollnagel AEC 74-76 Junya Shimazaki Riso 60-62 Aulis A. Hellstcn JAERI IVO 76-77 Makoto Ohsawa EURATOM 61-65 Tapio Eurola JAERI EKONO 76-77 Takeo Onchi 59-62 Hendrik Buis 62-64 Olavi Vapaavuori CRIEPI Luxembourg AEC 76-78 Masaaki Uchida 60-61 Henri Lucas 66-67 llkka Mikkola JAERI Francc EKONO 76-79 RitsuoOguma 60-62 Jacques H. Frcyccnon 68-69 Mikko Hurme JAERI France NOKIA 76- Yoji Minagawa 61-63 Lesli Stevens 70-73 Martti O. Ncvalainen 78-79 Makoto Tsuifci Belgium EKONO NAIG 74-76 Jorma A. Karppinen 78-79 Takao Yagi VTT PNC FED. REP. OF GERMANY 75-81 Kari Vilpponen 78-80 Motoyasu Kinoshita VTT CRIEPI 59-62 Heinz IV.A. Braun 77-80 llkka J. Leikkonen 78-80 Kazuaki Yanagisr.wa AEG VTT JAERI 62-65 Heinz H. Vollmer 79-87 Jaako Lahli 79-81 Hidctake Takahashi Technische Hochschulc Stuttgart VTT JAERI 63-73 Rainer Grumbach 80-82 Risto Sairanen 79-81 Hiroshi Tanaka GRS VTT CRIEPI G5-66 Heinz Giinther Walger 80-82 PerttiJ. Visuri 80-82 Tetsuo Nakajima Siemens TVO JAERI 65-67 RudiSieber 81-87 llkka J. Leikkonen 81-81 Tsuncmi Kakuta NUKEM VTT , JAERI 66-67 Manfred Gogola 82-83 Eero Patrakka 81-82 Hisatake Okamoto AEG TVO CRIEPI 66-67 Dieter Risse 83-84 Kari J. Porkholm 81-82 Kiyoshi Tamayamii TI'IV IVO PNC 67-68 HelmuthJ.F. Plitz 84-85 Risto Sairanen 82-85 Jinichi Nakamura AEG VTT JAERI 67-69 Wunibald Deller 84-86 Ari Kautto 82-83 Sadanori Yoshimura NUKEM VTT NAIG 67-69 Michael Gartner 82-84 Seiichi Yoshimura Siemens CRIEPI 82-84 Kimio Hayashi 76-78 Bcrnt Fagcrstrom 68-70 Harvey A. Taylor, Jr. JAERI AB Atomcncrgi USAEC (BNW) 84-85 Katsuihiro Kamimura 78-80 Gunnar Isaksson 71-73 William J. Quapp PNG KTH, Stockholm NFS 84-86 Hidaloshi Amano 79-81 Bernt L. Karlsson 72-75 Victor Hazel JAERI AB Atomcncrgi General Electric 85-87 Tomoyuki Abe 81-82 Mette Holmgren 73-76 Geoffrey H. Chaldcr PNC. FOA Combustion Eng. 85-87 Masao Yokobayashi 84-86 Conny Holmslrdm 74-77 James A. Christcnscn JAERI Studsvik Energitcknik AB USAEC (BNW) 75-76 Gary Thomas EPRI 75-78 Trevor C. Rowland General Electric SWITZERLAND THE NETHERLANDS 76-77 Michael Miller EPRI 58-63 Robert J. Swanenburg de Veye 58-60 Otto LBsclier 76-78 David Franklin RCN Suitzcr Brothers Ltd. EPRI 59-62 Johannes Asyee 59-62 Jean Paul Buclin 76-80 Thomas E. Hollowcll RCN Energie Nucleairc SA Combustion Eng. 60-63 Friedrich N.A. Habcrtnann 60 61 Aldo Sutter 77-80 Richard W. Miller RCN Atom Electra USNRC 61-62 Victor Radcmakers 60-62 Alexander KUng 79-80 John M. Christenson Royal Dutch Navy Sultzer Brothers Ltd. Univ. of Cincinnati 61-62 Tony PJ.M. Stolz 63-65 Ygal Fishman 80-82 Anthony Appclhans Royal Dutch Navy Sultzer Brothers Ltd. USNRC (EG&G) 68-70 Fritz Haferl 62-62 Gerhard A. dc Boer E.T.H. 79-83 David O. Sheppard Delft University GE 62-63 Pieter Slijp 82-84 Gary L. Hunt Royal Dutch Navy USNRC (EG&G) 62-65 Antonius Brouwcrs UNITED KINGDOM 83-86 Bartlcy F. Conroy Royal Dutch Navy GE 63-65 Hcndrik van Lecuwen 58-60 Colin L. Brown 84-86 Mitchcl E. Cunningham Royal Dutch Navy UKAEA Harwell USNRC (PNL) 64-66 Pcicr Vinkhuyzen 58-60 Gerald Ingram 84-86 David Gertman RCN UKAEA Winfrith USNRC. (EG&G) 65-67 Herman N.Jagcr 60-62 Robert W. Bowring Royal Dutch Navy UKAEA Harwell 65-69 Jan H. Post 60-62 James Davidson Royal Dutch Navy BRAZIL (CBTN) Nuclear Power Group 74-75 Sergio R.B. dc Carvalho 67-68 Robert den Boeft 60-63 Philip H. Delves 75-75 Sergio A. Majdalani Royal Dutch Navy UKAEA Risley 67-70 Elbertus B.M. Majoor 63-63 Denis Appleton RCN UKAEA Risley CANADA (AECL) 68-69 Johannes B. Barnas 63-64 James Hannaford 63-64 George M. Allison Royal Dutch Navy UKAEA Risley 68-71 Johannes Bessemer 63-65 Arthur Oebbagc Royal Dutch Navy UKAEA Winfrith 70-72 Teunis Flamcling 64-65 Brian Knight NORWAY Royal Dutch Navy UKAEA Harwell 70-73 Johannes A.H.M. van Ncs 64-66 David L. Ward 58-59 Odd Anders Sjaastad Com pri mo UKAEA Harwell 58-60 Eivind Engcbrctscn 70-73 Age Stoffclsma 65-67 Peter J. Riley 58-61 Knut Bryhn-lngebrctsen Royal Dutch Navy UKAEA Winfrith 58-61 Kjelll'. Lien 70-77 Klaasjoon 67-69 William S.A. Black 58-61 Torstcin Moshuus RCN UKAEA Winfrith 58-62 FinnEnger 71-76 Rob M. Vcrsluis 79-81 Ja nes A. Turnbull 58-62 Olav Robert Kasa 7 £-74 Johannes C. Tjcmmes CEGB Berkeley 58-6fi fJenrik Agcr-Hansscn Comprimo '80-81 Timothy J. Haste 58-66 Roar Rose 75-76 Petrus J. van Kouwen UKAEA Springfields 58-69 EinarJamne RCN 8C-81 Brian J. Holmes 58-75 Torolf Wullum 80-87 Hero Ulfcrt Staal Nuclear Power Company 59-60 Anders Dragseth EC.N 80-82 Robert J.P. Cribb 59-61 Mons Lyng 81-81 Roland van Doornc UKAEA Win frith 59-62 Nils Arnc Standal KEMA 82-85 Rodney J. White 59-62 Roir Kr. Skaardal 87-87 Willem Jan Oosterkamp CEGB Berkeley 59-70 Jorgcn A. Firing KEMA 84-85 Mario:. Oakden 59- RagnarStiand UKAEA Springfields 60-61 Jan llelge Museby 85 -86 Ian Palmer 60-62 Alf Bratteboe SWEDEN BNFL Springfields 60-62 EilivDalen 86-87 Craig S. Reiersen 60-65 Kjell G. Jahren 58-60 Pchr Blombcrg CEGB 60-67 John Georg Sicvcrts AB Atomencrgi 86-87 Annette Vcrlc 61-67 KjcllM. Nesct 59-60 G. Bernandcr Aston University 61-68 Ovc Kjcll Bakkcn ASEA 86-88 Peter M.Crosby 61-76 Jan-Erik Lunde 59-60 Nils Rydcll BNFL Springfields 62-66 EmilJansen AB Atomenergi 62-66 Ragnar Solhcim 60-62 Evelyn K. Sokolowski 62-74 Erik Rolstad ASEA 62-66 Olav G. 0ye 60-63 Fredrik Akerhielm 63-64 Terje B. Christcnsen AB Atomcnergi U. S. A. 63-65 Bjorn Asphaug 61-62 Jan G.H. Falck 63-69 John Inglish Wood AB Atomcnergi 58-59 James A. DcShong 63-70 Gudolf Kjx-riicim 61-62 Tord Hcllsten USAEC (Argonne Nat. Lab.) 64-65 Ivar Devoid AB Atomcnergi 58-59 Joseph A. I-leck.Jr. 64-66 Aksel Bruun 61-65 lngemar Hallstrom USAEC (Brookhavcn Nat. Lab.) 64-68 Reidar lnnvrer 62-64 Arne J.W. Andersson 59-60 Leonard W. Fromm 64-69 Kare Romslo AB Atomcnergi USAEC (Argonne Nat. Lab.) 65-74 Thor H. Korp3s 63-64 Carl Georg Moberg 61-64 Robert D. Smith 65-78 Bjorn Blomsnes x ASEA- USAEC (Savannah River Lab.) 65-82 Jon Kjell Trcngcrcid 67-69 Rune Josefsson 63-64 Per H. Trebler 66-67 Thor Soriic AB Atomcncrgi USAEC (Aerojet - General) 66-69 Stcinar Aas 67-70 Kim Ekberg 63-65 Lane Bailey 67-68 Ilclgc Moen AB Atomenergi USAEC (Savannah River Lab.) 68-69 Arve Andrej.cn 70-72 RairEspcfiilt 63-65 Herbert J.Olson 69-72 Jan Magnus Doderiein AB Atomcnergi USAEC (Savannah River Lab.) 69-72 Svcin Erik Wennemo-Ilanssen 73-75 Svcn Maimskog 65-68 Robert Winn Keaten AB Atomcnergi USAEC 66 Transport & Maintenance Staff 69-80 MarlinJ. Bnikas 60-65 Roger Lochen 69-80 Kilre Netland 60-67 Bjorn Ovre Ostby 58-82 AineHolm 70-72 Jan Kr. Kjcldstad 60-63 Jan Hcgi?n Jacobsen 58-82 Ragnvald Torp 70-77 Gunnc Ryntvcit 60-71 Roy Ringmann 59-62 Kjcll Rustand 70-83 Thov Palmgren 60-72 Martin Ystebo 60-71 AintOlscn 72-73 Arne Baggerud 61-64 Olaf F. Bjornstad 61-63 Asbjom Smcdsrod 72-73 Tore Lindmo 51-64 Rolf Holt-Jacobsen 61-65 Amc Kr. Cylvik 72-73 Rune K. Undeland 61-67 Sveinung Rasmusscn 61-69 Hurtmann Erikscn 72-82 Jorgen A. Firing 61-65 Kjell J. Svcndsen 63-72 Paul Sandum 75-78 Tom Stokka 61-86 Odd Hatlevold 67-88 KSre W. Johansen 76-85 Odd Falmyr 61-86 Hans Julsrud 76-81 Ragnar G. Stokkc 62-65 Oystcin Kolberg 79-81 Kai Klingstrom 63-67 Guri Ekblad TECHNICAL SUPPORT STAFF 79-82 Rune Nlocn 63-66 Knut Martinussen 79-86 llans Kurt Karlsen 66-70 Bjorn Roger Svanberg Reactor Plant Staff 66-78 Oddvar Lunde 80-81 Jon Svccn Haugcn 58-62 Jan Uno Harlem 80-85 SveinNovik 67-70 Kjcll Engebret5en 67-79 Ovc Kvalheim 58-62 Scverin Sydncs 80-85 Gunnar Skjcrvc 58-72 Arthur Aamodt 69-74 Jan HSkon Hansen 80-86 Kjcll Arnescn 58-76 Rolf Gjersfie 70-74 Olav Fossland 82-85 Johnny Petersen 58-80 Knut Ek 75-81 Annc-Mctte Vilpponen 83-86 Gunnar Moltcbcrg 58-83 John Jakobscn 79-82 Arvid Hove 59-61 Bjarnc Svcndscn yoru^gian industry Secondees 79-84 Torfinn Dybvik 59-77 Kjcll Guivik 82-84 Solveig Lavhaug 60-70 Ottar Petterscn 67-69 Fritz Brinck 84-88 Wilhelm Aascr 60-71 Ivan Johannesen Noratom 86-87 Kjcll B. Martinsen 67 -69 K5ic Netland 60-87 Age A.Johanscn Noratom 61-83 Per Y. Dahl 70-73 Bcrnt Slcttvold Administrative Officers 62-67 Wilhelm Ekblad Norsk Data 62-71 Alt H. Pederscn 71-73 Gunnar Berge 58-75 Erling Sandum 62-72 Hans Kr. Jorgensen Kongsberg Vapenfabrikk 59-69 Ragnar Ostby 62-74 Arbjorn Nilsen 71-75 Endrc Endrcscn 61-62 Nils K.Sclte Kongsberg V&penfabrikk 62-83 Anders Pctteroc 73-76 Karc Lochscn 61-74 Harry Holm 63-65 Bcla Grill Norsk Data 71-82 Peter H.Hcim 63-67 Roar Svendscn 77-78 Martin Sundlcy 63-73 Walt her Dahlc Norsk Data 67-87 SverrcJohanscn 77-81 ThorolfOlIi Canteen Staff 69-74 Harald Biserud Norsk Data 70-73 Kjcll Gressholt 78-79 Geir Martinsen 59-64 Aase Anderson Norsk Data 60-63 Gislaug Larsson 74-80 Tore Eisler 78-79 Sveinung Tubaas 62-68 Hclga Hansen 74-87 Sigurdjohansen Norsk Data 63-71 Signe Hewitt 77-85 Morten Spolin Nilsen 79-80 Bjorn Knudscn 79-87 Jan-Erik Hansen 68-79 Margrcthe S. Johansen Noratom 84-87 Roar J. Kjolerbakken 79-81 Erik Edsbcrg 73-85 Anna Signcbden NTH 83-84 Audun Holm 80-81 Terje Noring Norsk Data Laboratory & Design Assistants 84-86 John Midtsacter 59-66 Age Myrene Slatoil ADMINISTRATIVE SUPPORT STAFF 60-63 Annc-Britl Erikscn 84-86 Sigmund Nevdal 60-64 Asmund Hov Statoil Secretaries 62-67 Oistcin Olscn 85-87 Odd Bjonncss 62-72 Ingcr Sorlie Norcontrol 58-61 Bjorg Ager-IIanssen 58-62 Ragnhild Mittet 63-66 Per Kjcrulf Hansen 85-87 Erik Gran 63-67 Erik Foss Norcontrol 59-62 Lisc Krosby 59-62 Helen Wullum 63-67 Kate Vigil is Wold-Olscn 59-81 Kari Finholl 64-67 Lisbelh Gylvik 60-62 Jorun Asserson 64-67 Marit Stoverstcn 60-62 Wcnchc Arnessen 66-70 ?er Mustorp PROFESSIONAL STAFF 60-70 Bjorg Rive 66-71 Grethe Holmcn 60-76 Benedicte Vestad 67-78 LiseTorp Mechanical and Electrical Engineers 62-64 Randi Solberg 69-74 Erling Torp 70-75 GretheJ.Kruse 58-61 Arne LoFstad 62-65 Grethe Strand 71-74 Anne Elise Orvang 58-61 Oivind L. Nilscn 62-69 Kari Romslo 72-75 Trine Elvestrand Johanscn 58-61 Svcin Erik Rodcland 63-65 Ellen Nygaard 72-76 Terje Gudmundsen 58-62 Inge Marino Bagge 63-66 Anne Kari Johansen 73-78 Wenchc Lisc Berg 58-62 Aage Kristianscn 66-69 Marit Brynhildscn 66-69 Siri Kaufmann 75-84 TorunnJohnsen 58-64 Knut VV. Nicss 76-80 Doicen Eikaas 59-62 Hallvard VidsjS 69-73 Marit S. Boker6d Hansen 71-77 Larna Kvalheim 77-84 Unni Orum Jonassen 59-69 Tor Woien 77-86 Bjfirg Knoph Holmen 59-84 SveiTe Didrikscn 76-80 Grcthe Groth 59-87 Willy Sorcnsen 60-63 Olc Kr. Boc PROJECT MANAGEMENT General Management Reactor Management Research Management Project Managers Reactor Operations Reactor Physics and 58-61 Kjcll Pcttcr Lien Process Control Research 58-62 Olav Robert K5sa 62-66 Emiljansen 61-61 Olavi Vapaavuori (F) 58-65 Henrik Ager-Hanssen 66 69 Steinar Aas 61-68 Einarjamne 65-72 Jan Erik Lunde 69-72 Jan Magnus DBdcrlein 68-75 Torolf Wullum 72-73 Rainer Grumbach (G) 72-76 Jan Erik Lunde 75-82 Jorgen A. Firing 74-78 Ernst Robinson (G) 77-79 Raincr Grumbach (G) Dpy. Project Managers Reactor Engineering (Dpy. Heads of Reactor Operations} Instrumentation Development 58-60 Kjell Petter Lien and Fuel Experimental Research 61-65 Henrik Ager-Hanssen 58-61 Einar Jamnc 67-76 Torbjorn J.H. Vik 58-60 Colin L. Brown (U.K.) 58-60 Nils Rydell (S) 76-82 Jorgen A. Firing 61-62 Tord Hellstcn (S) 61-64 Robert D. Smith (U.S.) 62-78 Ragnar Strand 64-65 Herbert Olson (U.S.) 79-82 Tor Hemes . 65-70 Jorgen A. Firing Project Management Staff Organisation as of May 1988 TJ. Vik - Project Manager Graduate Research Staff 66 at the OECD Haldan Reactor Project A. Hanevik • Dpy. Project Manager Other Professional Staff 60 Supporting Staff (many ears) IS K.D. Knudscn • Assist. Pr.Mgi., Fuel Expcrim. Grand total 204 H. Smidt Olsen - Assist. Pr.Mgr., Reactor Safety J. Berger - Secretary PROCESS CONTROL RESEARCH FUEL PERFORMANCE RESEARCH REACTOR OPERATIONS AND ENGINEERING Man-Machine Systems Research Fuel Experiments and Evaluation T. Hemes • Chief of Operations H. Smidt Olsen (Acting) Division Head E. Kolstad • Division Head P. Gunnctud • Assist. Div. Hcad/Exp. Coord. TJ. Bjorlo - Dpy. Division Head H. Offcnberg/G. Bskkelie - Secretaries G. Karlsen • Secretary G. Bjerkely - Secretary In-Raactor Experiments Execution Reactor Engineering Integrated Surveillanca and Control Sytteim K. Svanholm 0. Finholt NORS/HAMMLAB K. Haugjct Y.Minagawa (J) Experimental £. Fladeby Electrical Engineering J.M. Aasgaard Operation B.D. Nythe 0. Evjen S. Sather Electronics A. Tande C-O. Fait H. Rekvin Laboratory H. Brithcn N.T. Fdrdestrdmmen A. Wahlstrom J.Engebretsen S. Holm Data Management A. Langseth L.I. Kristianscn W. Wiesenack (G) System Development J.K. Jensen Mechanical Engineering 1. Kvalem K.Karlsen O.Chr. Berntjen J. Kvalem T. Lovhaug II. Lindskog T. XJrjud (Installation & Maintenance) B. Swanberg A. Stenvald H. Devoid Data Processing 8c S. Larsen Operator SuppDrt Systems A.B. Jacobsen Documentation R. Aksclscn 0. Berg H. Haukedal Data Evaluation A. Jensen R-E. Grini S. Granata (I) O. Jensen S.Hval J. Killeen (U JC.) T. Johansen U.S.Jorgensen M. Smith (UX.) J. Martinsen M. Lilja(F) PA. Tempest (11 JC.) A. Mark B.Meyer V.Tosi T. Andersen 0. Naess T.TsukadaO) (Fuel Handling) T.SusudoO) S. Uematsu (J) J. Hauge Computerisation of Procedural E. Kjelvik F. Owre P. Hofgaard • Contr. Mgr./Progr. Schedules V. Lovaas H. San&uelsson J.S. Larscn Instrument Development and Radiation Protection & Chemistry EJ. Lund U.Ottersen S. Nilsen Experimental System Dciign T. Sivertsen 0. Vitanza - Division Head If. Strangstadstuen Radiation Protection H. Valisuo (F) A.B. Andersen H. Offenberg/G. Baekkelie - Secretaries Man-Machine Interaction E, Skattum E. Marshall (U.K.) Instrument Development & Qualification K.Solberg B. Aarsct R.Oyan S. Baker (U.K.) A. Haaland E. Foshaug L.Smith (U.K.) N.W. Hogberg K. Torgersen T.Johnsen K. Fjellestad Reactor Chemistry Software Reliability A. Kruithof L. Lie G. Dahll T. Stien Calibration & O.Wikstiil V. Pettersen Control J.E. Sjoberg Reector Control T. W6ien R. Aamot - Dpy. Chief of Operations Control Room Engineering and Design & Fabrication R. Valscth K. Bjorkheim Shift Supervisors Simulator Systems R. Agerup M. Ovreeide - Division Head 1. IV. Johansen Design Office S-E. Christiansen F. Pettersen - Assist. Div. Head F.H. Pedersen P. Grcnager T.O. Andersen S. Ldwengreen S. Kaufmann - Secretary P. Dalene A. Melin G. Holmcn S. Naerum Control Room Computer Engineering K. Bisseberg F. Pettersen R. Stalder Instrument Workshop. Operators T. Eng T.H. Bjcrge T. Hvedmg Work Station R. Andrcassen E. Johannessen Senior H/W Eng. Development A. Cascales B. Martinsen R. Stork&s 0. Kolleglrd A. Olsen Senior S/W Eng. V. Staal T. Thowsen A. Hornaes J.Tinglef J. Bauger G. Pettersen E.R. Dahl A. Teigcn K. Dahl T. Olsen Laboratory F. Evensen T. Lindskog Operation & H.O. Evensen F.T. Konradsen Maintenance J. Fredheim J.Necb Administrative Support Functions E. Hansen F. Sundal F. Johansen J.0. Hoi Control Room T. Kristoffersen D. Magnusson Studies H.Milde Purchasing/ J.E. Lyseng B.B.Thomassen 0. Gimie Transport S. Strom G. Ohra M.Jensen O.G. Thorsen T. Bertelsen J.S. Mjolnerod J.E. C.arl5son L. Bokerod R. Suther Training Simulator Development OM. Soiheim E. Stokke 0. Svarod Property 0. Eriksen Maintenance A.S. Sotie -Techn. Mgr. M. Engcbretscn R. Bohn T. Pedersen S. Novik Aa. Arncssen/ Receptionists A. Valseth G. Gravningen C-V. Sundling Control Systems B. Christiansen/ Canteen J. Augustin E.Naess Staff P. Hafstad A.L. Fredriksen/ J. I.xrsrn J.Nygird K..A. Ovreeide B.E. Olsen Personnel & M. Pehrsen Instructor Systems T. Danielsen Financc Office K.T. Hansen A. Stang ' Medical Office U.Weyer B. Magnusscn Non-Norwegian Staff Legend: P. Kriseiansen Computer Systems T. Heide Layout Office (D) Denmark : Oparetion Planning Simulator . . L. Brevig (F) Finland J. van Ncs- R. Aronscn Office Cleaning