TRIGA® Reactors
TOC-21
TENTH EUROPEAN TRIGA USERS CONFERENCE
PAPERS AND ABSTRACTS
HELD AT ATOMINSTITUT DER ÖSTERREICHISCHEN UNIVERSITÄTEN VIENNA, AUSTRIA
SEPTEMBER 14-16, 1988
A GENERAL ATOMICS NOTICE
Although General Atomics has attempted to compile the material contained in this issuance with accuracy, neither it, its employees, nor its agents can make any warranty or representation, expressed or implied, with respect to the accuracy or completeness of such information or assume any liability with regard to the use of or for damages resulting from the use of any information, apparatus, method, or procedure described in this issuance. i&îlAKWVâL
•W*C"
•:&••• •...
,:«*<* ^
i' •
^Ä;'':":,,v ï . "
, S "*' ^fei^;:; ^;Sp§:
/1ST OF TRIGA USERS' CONFERENCES
1970, February 19-20 Denver, Colorado, USA 1970, August 25-27 Otaniemi, FINLAND 1972, February 21-22 College Station, Texas, USA 1972, September 12-15 Pavia, ITALY 1974, February 25-27 Albuquerque, New Mexico, USA 1974, October 29-31 Neuherber g, near Munich, GERMANY 1976, March 1-3 Salt Lake City, Utah, USA 1976, September 28-30 Vienna, AUSTRIA 1977, February 28-March 1 Tucson, Arizona, USA 1978, February 28-March 2 Corvailis, Oregon, USA 1978, September 4-6 Portoroz, YUGOSLAVIA 1980, March 2-5 San Diego, California, USA 1980, September 16-18 Mainz, GERMANY 1982, March 8-10 Idaho Falls, Idaho, USA 1982, September 15-17 Istanbul, TURKEY 1984, March 12-14 Anaheim, California, USA 1984, August 21-23 Espoo, FINLAND 1986, April 6-9 College Station, Texas, USA 1986, October 7-9 Casaccia, ITAL Y 1988, April 10-13 Washington, D.C., USA 1988, September 14-16, Vienna, AUSTRIA
1 ACKNOWLEDGEMENTS
We wish to express sincere appreciation for the efforts of Dr. Helmuth Bock Chairman of the Conference and his colleagues for their efforts in organizing and conducting an outstanding conference
On behalf of the conference participants, we wish to acknowledge the hospitality extended by the Atominstitut for all of the social events, including the outstanding dinner in the wine cellar.
11 FOREWORD
THE TENTH,EUROPEAN TRIG USERS CONFERENCE WAS HELD IN VIENNA SEPTEMBER 14-16, 1988 UNDER THE SPONSORSHIP OF THE ATOMINSTITUT
The papers which follow in this document are presented in the same order as listed in the Conference Program. All papers which were received for publication (forty-eight) have been included.
m TENTH EUROPEAN TRIGA USERS CONFERENCE
LIST OF PARTICIPANTS
AUSTRIA
IAEA A-2610 ALCALA F. P.O. Box 100, A-1400, WIEN, AUSTRIA
Atominstitut der Osterr. Universitäten BADUREK, G. Shuttelstrasse 115, WIEN, AUSTRIA
Atominstitut der Osterr. Universitäten BOCK, H. Schutteistrasse 115, WIEN, AUSTRIA
Atominstitut der Osterr. Universitäten BUCHTELA. K. Schutteistrasse 115, WIEN, AUSTRIA
IAEA A-2639 BYSZEWSKY, W. P.O. Box 100. A-1400, WIEN, AUSTRIA
Atominstitut der Osterr. Universitäten HAMMER, J. Schutteistrasse 115, WIEN, AUSTRIA
Atominstitut der Osterr. Universitäten KASPAREC, F. Schutteistrasse 115, WIEN. AUSTRIA
Atominstitut der Osterr. Universitäten KIRCHSTEIGER, C. Schutteistrasse 115, WIEN, AUSTRIA
IAEA A-2371 MURANAKA, R. P.O. Box 100, A-1400. WIEN, AUSTRIA
Reaktorinstitut, Steyrergasse 17 NINAUS, W. A-8010 GRAZ. AUSTRIA
IAEA A-2366 ROSENBERG, R. P.O. Box 100, A-1400, WIEN, AUSTRIA
Atominstitut der Osterr. Universitäten SLANETZ, H. Schutteistrasse 115, A-1020, WIEN, AUSTRIA
Atominstitut der Osterr. Universitäten TSCHURLOVITS, M. Schutteistrasse 115, A-1020, WIEN, AUSTRIA
Atominstitut der Oster. Universitäten ZUGAREK, G. Schutteistrasse 115, A-1020, WIEN, AUSTRIA
ENGLAND
GREEN, D. ICI, Phys. and Radioisotop Service Group Chemicals and Polymers Ltd P.O. Box 1, Billingham. CLEVELAND, ENGLAND
IV FINLAND
SALMENHAARA, S. Technical Research Centre of Finland Otakaari 3A, SF-02150 ESPOO15, FINLAND
GHANA
AHAFIA. A. National Nuclear Research Institute Ghana Atomic Energy Com. P.O. Box 80, LEGON-ACCRA, GHANA
GERMANY
KRAUSS. O. Deutsches Krebsforschungszentrum Institut fur Nuklearmedizin Im Neuenheimerfeld 280 D-6900 HEIDELBERG, FEDERAL REPUBLIC OF GERMANY
MAIER-BORSZ W. Deutsches Krebsforschungszentrum Institut fur Nuklearmedizin Im Neuenheimerfeld 280 D-6900 HEIDELBERG, FEDERAL REPUBLIC OF GERMANY
MENKE, H. Institut fur Kernchemie, Universität Mainz Postfach 3980, D-6500 MAINZ FEDERAL REPUBLIC OF GERMANY
ROEGLER. H. Interatom GmbH, D-5060 Bergisch Gladbach l FEDERAL REPUBLIC OF GERMANY
HUNGARY
GYEMESl Z. Institute of Isotopes of the Hungarian Academy of Sciences P.O. Box 77, H-1525BUDAPEST, HUNGARY
VARGA. C. Institute of Isotopes of the Hungarian Academy of Sciences P.O. Box 77, H-1525 BUDAPEST, HUNGARY HALX
Laboratorio Energia, Nucleare Applicata ALLONI, L. Viale Taramelli 10, 1-27100PAVIA, ITALY
Laboratorio Energia, Nucleare Applicata ALTIERI, S. Viale Taramelli 10, 1-27100PAVIA, ITALY
Universita di Pavia, L.E.N.A. CINGOLI. E Viale Taramelli 12, 1-27100 PAVIA, ITALY
E.N.E.A. Dipartimento TIB FESTINESI. A. CASACCIA-Strada Anguillarese km 1+300 ROMA, ITALY
v LOSCHIAVO, V. E.N.E.A. Dipaartimento TIB CASACCIA-Strada Anguillarese km 1+300 ROMA, ITALY
MANZELLA, P. E.N.E.A., Division of Nuclear Safety and Health Protection Via V. Brancati 48,1-00144ROMA-EUR, ITALY
MELONl S. Universita di Pavia, L.E.N.A. Viaie Taramelli 12, 1-27100PAVIA, ITALY
REIS, G. E.N.E.A. Dipartimento TIB CASACCIA-Strada Anguillarese km 1+300 ROMA, ITALY
ROSELLI F. E.N.E.A. Division of Nuclear Safety and Health Protection Via V. Brancati 48, 1-00144ROMA-EUR, ITALY
SANTORO, E. E.N.E.A. Dipartimento TIB CASACCIA-Strada Anguillarese km 1+300 ROMA, ITALY JAPAN
HARASAWA, S. Institute for Atomic Energy Rikkyo University, 2-5-1, Nagaska YOKOSUKA, 240-01, JAPAN
HORIUCHL N Atomic Energy Research Lab. Musashi Institute of Technology Ozenji 971, Asoa-ku KAWASAKI, 215. JAPAN
N0ZAK1 T. Atomic Energy Research Lab. Musashi Institute of Technology Ozenji 971, Asoa-ku, KAWASAKI 215, JAPAN MOROCCO CNESTEN, 15 Rue de Sebou Agdal BENMANSOUR, A. RABAT, MAROC
CNESTEN, 15 Rue de Sebou Agdal HADDOU, A. RABAT, MAROC
CNESTEN, 15 Rue de Sebou Agdal HOUARl A. RABAT, MAROC
S3HIZEBLANJ2 HOLM, H. General Atomics-Europe Dorfstrasse 63, CH-8126 ZUMIKON SWITZERLAND
vx TAIWAN
WANG, T. Institute of Nuclear Engineering Nuclear Reactor Division National Tsing-Hua University Hsinchu 30043, TAIWAN, R.O.C. HJBKEX Istanbul Technical University Faculty of Aeronautics &. Astronautics BARLA, M Maslak 80626, ISTANBUL, TÜRKEI
Istanbul Technical University Institute for Nuclear Engineering BAYULKEN.A. Ayazaga Campus 80626, Maslak, ISTANBUL, TÜRKEI
Institute for Nuclear Energy Technical University, Ayazaga Campus 80626 BILGE, A. Maslak, ISTANBUL, TÜRKEI
Istanbul Technical University Nucleer Guc Radiation Lab CIFTCIOGLU. O. Gumussuyu 80191, ISTANBUL, TÜRKEI
Istanbul Technical University Institute for Nuclear Energy DEMIRALP, R. Maslak 80626, ISTANBUL, TÜRKEI
Istanbul Technical University Institute for Nuclear Energy OZGENER, H. Maslak 80626, ISTANBUL, TÜRKEI
Istanbul Technical University Institute for Nuclear Energy TUGRUL, B. Maslak 80626, ISTANBUL, TÜRKEI
Institute for Nuclear Energy Technical University, Ayazaga Campus 80626 YAVUZ.H. Maslak, ISTANBUL, TÜRKEI
U.S.A. General Atomics P.O. Box 85608, SAN DIEGO, CA 92138, U.S.A.
CHESWORTH, R. General Atomics P.O. Box 85608, SAN DIEGO. CA 92138. U.S.A.
GANLEY, J. General Atomics P.O. Box 85608, SAN DIEGO, CA 92138. U.S.A.
HYDE. W. Gamma-Metrics 5550 Oberlin Drive, SAN DIEGO, CA 92121, U.S.A. MILLER, J.
vu WELSH, C. Ga m ma-Metrics 5550 Oberlin Drive, SAN DIEGO, CA 92121, U.S.A.
YUGOSLAVIA
Josef Stefan Institute DIMIC, V. Jamova 39, YU-61000 LJUBLJANA, JUGOSLAWIEN
Josef Stefan Institute DUS1C, M Jamova 39, YU-61000 LJUBLJANA, JUGOSLAWIEN
Josef Stefan Institute GLUMAC, B. Jamova 39, YU-61000 LJUBLJANA, JUGOSLAWIEN
Inst, for Mathematics and Physics JOVANOVIC, S. Univ. "V. Vlahovic" Cetinjski put b.b. YU-81000 TITOGRAD, JUGOSLAWIEN
KR1ST0F, E. Josef Stefan Institute Jamova 39, YU-61000 LJUBLJANA. JUGOSLAWIEN
MELE, I. Josef Stefan Institute Jamova 39, YU-61000 LJUBLJANA, JUGOSLAWIEN
RANT, J. Josef Stefan Institute Jamova 39, YU-61000 LJUBLJANA, JUGOSLAWIEN
SMÖDIS, B. Josef Stefan Institute Jamova 39, YU-61000 LJUBLJANA, JUGOSLAWIEN
VOJNOVIC, D. Josef Stefan Institute Jamova 39, YU-61000 LJUBLJANA, JUGOSLAWIEN
VUKOTIC, P. Institute for Mathematics and Physics Cetinjski-put b.b. YU-81000 TITOGRAD, JUGOSLAWIEN
vin CONTENTS
SUBMITTED PAPERS
TENTH EUROPEAN TRIGA USERS' CONFERENCE
SESSION I. REACTOR OPERATION AND MAINTENANCE EXPERIENCE
- Utilization of Research Reactors - A Global Perspective: R.G. Muranaka, IAEA, AUSTRIA 1-1
Safety Inspections to TRIGA Reactors: W. Byszwesky; IAEA, AUSTRIA 1-11
- Maintenance of the TRIGA Mark II Reactor in Ljubljana and the Disposal of Radioactive Waste: V. Dimic, G. Pregl and B. Pucelj ; "J. Stefan" Institute, YUGOSLAVIA 1-21
- Operation and Utilizations of Dalat Nuclear Research Reactor: P.Z. Hien; Dalat Nuclear Research Institute, VIETNAM 1-25
- Nine Years of Operation of ITU-TRR TRIGA Mark II Reactor: H. Yavuz, A.R. Bayulken, M.A. Yavuz; Istanbul Technical University, TURKEY 1-31
- Initial Operation and Utilization of the Bangladesh 3-MW TRIGA Reactor: M.A. Mannan, M. Hossain; Bangladesh Atomic Energy Commission, Dhaka, BANGLADESH 1-39
Operational and Research Activities of the Tsing Hua Open Pool Reactor: Tien-Ko Wang, Der-Ling Tseng, Huai-Pu Chou, Minsun Onyang; National Tsing Hua University, Hsinchu, TAIWAN, R.O.C 1-55
Operating Experiences at the Finnish TRIGA Reactor: Seppo Salmenhaara; Technical Research Centre of Finland, Espoo, FINLAND 1-73
Operating Experience and Maintenance at the TRIGA Mark II Lena Reactor: P. Cingoli, S. Altieri, P. Lana, G. Rosti, L. Alloni, S. Meloni; Universita di Pavia, ITALY 1-77
- The Research Reactor TRIGA Heidelberg II: W. Maier-Borst, 0. Krauss; Heidelberg, FEDERAL REPUBLIC GERMANY 1-87
Operation and Maintenance Experiences at the C.R.E Casaccia TRIGA Reactor: A. Festinesi; Casaccia, ITALY .....1-95
Operation Experience with the TRIGA Reactor Vienna: H. Bock, J. Hammer, G. Zugarek; Atominstitut, AUSTRIA .1-113
Ghana's Nuclear Program: A.K. Ahafia; Ghana Atomic Energy Commission, Legon, GHANA . 1-121
LX SESSION II. NEW DEVELOPMENTS AND IMPROVEMENTS OF TRIGA COMPONENTS AND SYSTEMS, INCLUDING INSTRUMENTATION
Concepts for the Renewal of the TRIGA Reactor Instrumentation Vienna: J. Hammer, G. Zugarek, H. Bock, Atominstitut, Vienna, AUSTRIA 2-1
- Replacement of the Cooling System of the TRIGA Mainz Reactor: H. Menke; Universität Mainz, FEDERAL REPUBLIC GERMANY 2-7
- Use of a Digital Reactivity Meter on TRIGA Reactor for Instrumentation and Training Purposes: B. Glumac; "J. Stefan" Institute, Ljubljana, YUGOSLAVIA 2-17
Digital Automatic Control System for the TRIGA Reactor: S. Harasawa, et al; Institute for Atomic Energy, Rikkyo University, Yokosuka, JAPAN 2-27
Operating Data Documentation System for a Research Reactor: F. Kasparec, J. Hammer; Atominstitut, Vienna, AUSTRIA 2-47
SESSION III - FUEL AND FUEL MANAGEMENT
- Reactivity Effects in Mixed TRIGA Cores: M. Ravnik, I. Mele; "J. Stefan" Institute, Ljubljana, YUGOSLAVIA 3-1
- The Effects of SS-304 Cladding on Core Calculation of Istanbul TRIGA Mark II Reactor: S. Gungor; Cukurova Universitesi, Istanbul, TURKEY 3-9
Gamma Spectrometrical Examination of Irradiated Fuel: E. Kristof, G. Pregl; "J. Stefan" Institute, Ljubljana, YUGOSLAVIA 3-25
Fuel Element Replacement and Cooling Water Radioactivity at the Musashi Reactor: T. Nozaki, T. Honda, N. Horiuchi, 0. Aizawa, T. Sato; Atomic Energy Research Laboratory, JAPAN ...3-39
SESSION IV - SAFETY ASPECTS, LICENSING AND RADIATION PROTECTION
Some Aspects of Licensing, Periodic Inspections and Backfitting of TRIGA Reactors: H. Weiss; Technische Universität Graz, AUSTRIA 4-1
- Reactor Physics Analysis of the Musashi Reactor for Licensing Purposes: H. Kadotani, M. Takami, 0. Aizawa, T. Sato; Century Research Center Corporation, JAPAN 4-7
Probabilistic Safety Assessment of the Research Reactor - TRIGA Mark II: M. Osredkar, V. Dimic, M. Dusic, M. Kozuh, D. Vojnovic; "Josef Stefan" Institute, Ljubljana, YUGOSLAVIA 4-17
Probabilistic Safety Assessment of the Vienna TRIGA Reactor: C. Kirchsteiger, H. Bock; Atominstitut, Vienna, AUSTRIA 4-29
x - Recent Radiation Protection Problems at the Vienna TRIGA Reactor: M. Tschurlovits, H. Slanetz; Atominstitut, Vienna, AUSTRIA • *-45
- Radiological Environmental Impact Related to Reactor Accidents: S. Altieri, P. Pedroni, S. Meloni; Universita di Pavia, ITALY , 4-53
Fission Products Distribution for Istanbul in an Hypothetic Reactor Accident: A.R. Bayulken, M.C. Barla; Istanbul Technical University, TURKEY .4-69
- Evaluation of the Effective Dose Equivalent to the Public of Pavia After the Chernobyl Nuclear Accident: S. Altieri, A. Berzero, N. Genova, S. Meloni, G. Rosti; Universita di Pavia, ITALY 4-81
SESSION V - EXPERIMENTS WITH TRIGA REACTORS (Solid state neutron physics & neutron radiography)
The Stationary Neutron Radiography System: A TRIGA-Based Production Neutron Radiography Facility: R. Chesworth, D. Hagmann, General Atomics, San Diego, U.S.A. 5-1
- Modernization Design of Neutron Radiography of ITU TRIGA Mark II Reactor: B. Tugrul, A.N. Bilge: Istanbul Technical University, TURKEY .5-21
- The Possibility of Gamma Ray Sterilization by Using ITU TRIGA Mark II Reactor: A.N. Bilge, B. Tugrul, H. Yavuz; Istanbul Technical University, TURKEY 5-33
- Liquid Waste Processing from TRIGA Spent Fuel Storage Pits: K. Buchtela; Atominstitut, Vienna, AUSTRIA .5-41
Neutron Physics Experiments Exploiting the Pulsing Capability of the Vienna TRIGA Mark II: G. Badurek, Atominstitut, Vienna, AUSTRIA 5-51
- TRIGA out of Core Gamma Irradiation Facility: J. Rant, G. Pregl; "J. Stefan" Institute, Ljubljana, YUGOSLAVIA 5-61
- Analysis of TRIGA Application for Irradiation of Power Reactor Pressure Vessel Specimens: I. Mele, M. Ravnik; "J. Stefan" Institute, Ljubljana, YUGOSLAVIA 5-67
SESSION VI - RADIOCHEMISTRY, RADIOISOTOPE PRODUCTION AND NAA
The Evaluation of the Radioisotope Production for Industrial Uses: A. Bilge, H. Yavuz, B. Tugrul; Istanbul Technical University, TURKEY 6-1
XI Evaluation of Selective Boron Absorption in Liver Tumors: D. Chiaraviglio, F. De Grazia, A. Zonta, S. Altieri, B. Braghieri, F. Fossati, P. Pedroni, T. Pinelli, A. Perotti, S. Specchiarello, G. Perlini, H. Rief; ITALY 6-17
- A New Handling Tool for Irradiated Samples at the Lena Plant: L. Alloni, A. Venturelli; Universita di Pavia, ITALY 6-23
- Determination of Essential and Trace Elements in Milk and Measurement of Short-Lived Nuclides Using FIMS: R. Demiralp, S. Kalayoglu, E. Unseren, F. Grass, H. Bock; Istanbul Technical University, TURKEY 6-37
The Use of the TRIGA Mark II Reactor, Ljubljana, Yugoslavia in the k0 Method of Neutron Activation Analysis: B. Smodis, S. Jovanovic, P. Vukotic, R. Jacimovic, P. Stegnar; YUGOSLAVIA 6-47
- Neutron Flux Characterization of the TRIGA Mark II Reactor, Ljubljana, Yugoslavia, for use in NAA: S. Jovanovic, B. Smodis, R. Jacimovic, P. Vukotic, P. Stegnar; YUGOSLAVIA 6-59
SESSION VII - REACTOR PHYSICS
- Nonlinear Dynamics of the ITU TRIGA Reactor: N.A. Hizal, S. Gencay, 0. Ciftcioglu, B. Can, E. Gungordu, M. Geckinli; Istanbul Technical University, TURKEY 7-1
Reactor Surveillance by Noise Analysis: 0. Ciftcioglu; Istanbul Technical University, TURKEY 7-11
Measurement of New Core Characteristics at the Musashi Reactor: N. Horiuchi, T. Matsumoto, T. Nozaki, 0. Aizawa, T. Sato; Musashi Institute of Technology, JAPAN 7-23
- Criticality Calculations for the TRIGA Mark II Reactor of ITU by the Finite Element and the Finite Difference Methods: H. Atilla Ozgener; Istanbul Technical University, TURKEY 7-A3
An Analytical Approach to the Positive Reactivity Void Coefficient of TRIGA Mark II Reactor: E. Edgu, T. Yarman; TURKEY 7-61
Xll SESSION I Reactor Operation & Maintenance Experience
UTILIZATION OF RESEARCH REACTORS - A GLOBAL PERSPECTIVE
R. G.'Muranaka Physics Section International Atomic Energy Agency
Introduction
This paper will present 1) a worldwide picture of research reactors, operable, shutdown, under construction and planned, 2) statistics on utilization of research reactors including TR.IGA reactors, and 3) some results of a survey conducted during 1988 on the utilization of research reactors in developing Member States in the Asia-Pacific Region.
The statistics are derived from the Agency's computerized Research Reactor Data Base (RRDB). The information was collected by the Agency through questionnaires sent to research reactor operators in 1985 and 1986. All data on research reactors, training reactors, test reactors, prototype reactors and critical assemblies are stored in the RROB. This system contains all the information and data previously published in the Agency's publication, Power and Research Reactors in Member States, as well as updated information.
Reactor operators were asked to confirm only the general information for reacto rs shut down, and to update the information on operational reactor s and reactor s under construction. Nearly all reactor operators returned the questionnaire; however, a few (13) reactor operators did not resp ond. The reactors for which no updated information was received or for which information is not readily available in the open literature are classified as unverified informatio n but for statistical purposes, these reactors were assumed to be opérât i onal. Reactors that have been shut down temporarily for modifica tions or upgrading were also considered as operatio nal .
Global Characterization of Research Reactors
According to the information available to the Agency at che end of June 1 9 d 8 , there were 325 nuclear- research reactors operating in the world. To date, 205 have been shut down, 17 are currently under construction, and 18 new facilities are planned. While developing countries account for less than one-quarter of tue operating facilities., they account for most of the facilities under construction or planned.
1-1 -2-
üperating Reactors
The 325 operating reactors are located in 56 countries Their geographical/political distribution is as follows:
North Amer ica 116 Europe (Western) 94 Asia - Pacific 52 South America 13 Africa 4 Middle East 5 Europe (Eas tern) 17 USSR 24 325 (VG-2)
Countries with the largest number of research reactors are
USA 99 USSR 24 FRG 21 France 20 Japan 18 UK 15 Canada 14 211 (65%)
It should be noted that 45 of the reactors in the USA are Department of Energy reactors. The reactors in the RIIDB range in power from 0 to 25U MW. About 25% of the reactors are rated at less than 1 KW including 46 that are classified as zero power. Forty two percent of the reactors are rated at less than 100 kW. As will be shown later, this class of reactors have the lowest utilization of the group.
(VC-3)
The age of operating rea ctors is surprisingly high. Most reactor:tors are over 2zu0 years ola.old Theiineir ages range rrofrom newly corammissione d to 42 years old. The average age is 22.4 years with the median age of 24 years. Whil«uxx (VG-4) Constructions The research reactors commissioned since 1980 are listed uelow. It snould be noted that 14 of the 20 listed are in developing countries. 1-2 France ORPHEE 14 ,000 kW 1980 USA FFTF 40,000 kW 1980 Canada SLOWPOKE 20 kW 1981 USSR ARGUS 50 kW 1981 Argent ina RA-6 500 kW 1982 Czechoslovakia LR-0 5 kW 1932 France SCARABEE M 100,000 kW 1982 Korea, R. of AGN-201 0 kW 1982 Malays ia TRIGA-PUSPATI 1 ,000 kW 1982 Libya IRT-1 10,000 kW 1983 Canada SLOWPOKE 20 kW 1984 China MNSR 27 kW 1984 India PURNIMA-II 5 kW 1984 Jamaica SLOWPOKE 20 kW 1984 India DRHUVA 100,000 kW 1985' India FBTR 4 0,000 kW 1985 Bangladesh TRIGA MK II 3 ,000 kW 198Ô Canada SDR 2 ,000 kW 1987 Indones ia RSG-GAS-30 30,000 kW 1987 Iraq TAHMUZ-2 500 kW 1987 Research reactors under construction include the following, with expected criticality dates indicated when known: Canada MAPLE X-10 10,000 kW 1991 Chile LO AGUIRE 10,000 kW 1988 China PPR-PULSIHG ' 1,000 kW 1988 LTUR 5,000 kW 1988 LPR 5,000 kW 1989 Czechoslovakia VR-1P 10 kW 1988 VR-13 100 kW 1990 Algeria • 500 kW ? India KAMINI 100 kW ? Iran ZPR 0 kW ? Japan Upgraded JRR-3 20,000 kW 1989 Korea, R. of KMRR 30,000 kW 1990 Morocco MA-R1 1,000 kW 1990 Peru PER-2 10,000 kW 1988 USSR PIX 100,000 kW ? USA SURS 1,000 kW 1989 TRIGA-II-TEXAS 1,000 kW 1988 Developing councries account for 12 of the 17 construction projects. Planned/Under Consideration The following countries have indicated interest in a new reactor and are in various stages of study, planning or negotiations. clow realistic the expectations are cannot be assessed at this time. 1-3 Argen t ina Sri Lanka Colomb ia Madagas car Cuba Nigeria Ecuador Poland Egypc Saudi Arab ia Ghana Syria Japan USA Korea, DPR Uruguay Shutdown Reactors The RRÜB shows 205 research reactors shutdown. Nearly all of these have been in industrialized countries where the reactor was shutdown because the specific mission of the reactor had been fulfilled or for financial and low utilization considerations. A few, however, may have been shut down for concerns of aging, and replaced with newer designs or for problems of safety bacicf i 11 ing . Utilization - Worldwide In order to categorize research reactor utilization based on information received from reactor operators, the following criteria has been used. It should be noted that this categorization is based only * on annual hours of operation and does not consider extent of use of installed facilities or the quality of utilization. A. Maximum Utilization > 6000 hr/yr B . High Utilization 2000 - 6000 hr/yr i.e. 1-2 shifts/day Good Utilization 1200 - 2000 hr/yr i.e. rvj 1 shift /day Low Utilization 800 - 1200 hr/yr i.e. 1/2 - 3/4 shifts/day E Under Utilization A total of 216 reactor operators provided varying degrees of detail on utilization. This group includes 40 TRIGA reactor operators. The breakdown of the reactors are as follows: Maximum Utilization - 58 reactors (14.8%) including 2 TRIGAs Average Power of reactor = 25.1 MW High Utilization - 58 reactors (26.9%) including 10 TRIGAS Average Power of reactors = 13.5 MW 1-4 Good Utilization - 39 reactors (13.1%) including 11 TRIGAS Average Power of reactors = 4.5 MW Low Utilization - 25 reactors (11.6%) including 6 TRIGAS Average Power of reactors = U.34 MW Under Utilization - 62 reactors (28.7%) including 11 TRIGAS Average Power of reactors = 0.15 MW Comparing TRIGA reactors with the rest of the reactor population, we find that, based on daily operating hours reported, 78 % of TRIGA reactors operate one-shift or more compared to 60% of the rest of the reactors. At the lower end, 12.5% of TRIGAs are underutilized against 29% for the rest. There appears to be a direct correlation between reactor power and use. Most highly used reactors are involved in materials testing, radioisotope production on a large scale, or neutron beam work. In the lowest category, most of the reactors are very low power or critical assemblies designed for specific studies or high power reactor mock-ups. As far as- the TRIGA reactor statistics in this lowest category, most of the reactors are located in American universities and are used almost exclusively for education and training courses, hence only on an as needed basis. A question often asked is the importance or role of a research reactor in a nuclear power program. Without drawing any conclusions, the statistics show that of the 26 Member States with nuclear power programs all operate research reactors. In fact, this group operates 271 research reactors. Six countries with nuclear power plants under construction operate 17 research reactors. Only one country, Cuba, has no research reactor but is negotiating the purchase of two. Asia-Pacific Developing Countries The research reactors in developing member states in the Asia-Pacific region were studied in greater detail through special questionnaires sent to reactor operators. Twenty four questionnaires were sent out and 16 replies were received. Of the 16 replies received, five were not included in the statistics because two were too new to have established utilization programs and three others because of use restricted to education and training. The Indian research reactors are not included in this summary because their data were received too late. However, India could hardly be considered a scientifically or technically developing country. Their reactors, mostly of indigenous design and construction are well utilized and heavily into basic and applied research. The inclusion of their data 1-5 would have presented an overly optimistic picture of utilization in the region. The reactors in the statistical base may be characterized as follows: Age 15.1 yrs median 16-11 yrs . Power 14.1 MW without HFETR = 3.0 med = 2 3 - 2.Ü Hours of operat ion 2524 hrs/yr Neutron Activation Analysis (NAA) NAA is the most popular use of reactor neutrons in the region. Eleven reactors have NAA programs of varying levels. Thioe average nuraoenumbecr o fJ. sampj.esamples anaiyasanalyzedu xcini a year is 3840 with the numbers ranging from 200 to over 12,000. The larigestnumoe r of samples are for geological exploration. The resu.It s of the survec nuirait y ar-,-r* e a-,s „ Pfollow ~ 1 1 ~,,r, s . Field No. of reactors No. of reactors doing NAA in Field with greatest no. in Field Geological ALL 6 Environmental 10 0 Medical/health, etc. 9 2 Agriculture 7 2 Industrial 8 0 Universities 8 0 In-House Research 7 1 Others 6 0 Radioisotope Production Ten reactors produce radioisotopes totalling 164,500 to 273,700 curies including large quantities of Co-60. If Co-60 is excuded, the totals are 900 Ci (1985) to 1250 Ci (1987). The results of the survey are summarized below. Medical (7 reactors) 1987 - 86,190 Ci including 84,000 Ci of Co-60 Radioisotopes include: Mo-99, 1-131, Au-198, P-32, Cr-51, Co-60 Agriculture (7 reactors) - produced tracer level quantities of: P-32, Ca-45, C-14, Rb-86 Indus trial (6 reactors) Co-60, Ir-192, Cr-51, Zn-65, I-131, Br-82, La-140, Ba-131, Po-210, Au-198, Na-24, Be-7 Univers i t ies (4 reactors) Am-241, Cs-134, Co-60, C-14, P-32, Pt-197, Br-82, S-32, K-42, Ca-45, Ag-110m, Na-24, tfo-99, Rb-ö6, Cr-51, Fe-59, Au-198 1-6 In-House Use (4 reactors) Mo-yy, Co-60, Br-82, Ag-110m, Sb-122, ßa-131, P-32, Fe-5y, Cr-51, Cu-64, Au-198, K-42, Na-24, Rb-86, 2n-65 Other Uses (4 reactors) Sn-113, P-32, 5-35, Ca-45 , Cr-51, Br-82, Na-24, Z-65, Cu-64 Basic Research Projects as Reported Ci-5-2 Thermal Neutron Cross-Section Fiss ion Yields Ternary Fission and Cold Fragmentation CN-5 Low Energy Neutron Physics Reactor Materials CN-y Neutron Physics, Nuclear Physics, Radiochemistry, Radiation Damage, Neutron Spectroscopy, Reactor Engineering, Materials IÜ-1 Crystal Structure, Magnetic Structure Texture KR-2 Texture PK-1 Debye-Waller Factors, Phonon Dispersion in Mixed Alkali-halides , Phase Transitions, Polymers VN-1 Low Energy Neutron Physics Reactor Materials In the area of neutron scattering, the region has 12 spectrometers installed at 5 reactors, 3 are under construction or in the design phase and 9 additional are planned. If all are constructed, spectrometers will be installed at 8 reactors. Neutron Radiography There are 6 neutron radiography facilities installed and an additional 4 are planned. The operating facilities average 136 radiographs per year ranging from 57 to 250 annually. Education and Training Fourteen reactors reported activities in this area of utilization. Although most involve reactor operator and other reactor related training, a very large range of fields have been specified. The average for the reactors is 300 students although a more typical number is about 40. One reactor reported 3743 students per year. TRIGA reactors (6) average 624 students including the 3743 students mentioned above. 1-7 Other Uses A large number of varied activities had been reported in this category. In addition to the expected NTD silicon and materials testing, the other activities are listed below by- reactors . C23-2 NTD Silicon Fuel and Materials Testing CN-5 NTD Silicon Fuel and Materials Testing CN-9 Nucleopore Membrane Standard Neutron Field Neutron Filter NTÛ Silicon Radiation Processing Gemstone Coloring CN-11 Shielding Biological Effects of Radiation NTD Silicon Fast Neutron Irradiation of Semi-conductors Nucleopore Membrane Use of Nuclear Heat 10-1 Reactor Kinetics Reactor Parameter Measurements Temperature and Flux Distribution Gamma Dose Rate Heating MY-i Reactor Interfacing System Development Noise Analysis Thermal-hydraulic Research Fuel Management Code Development TH-1 Gemstone Irradiation Reactor Experiments VN-1 Use of Ge-71 as XRF Source Flux Trap as Gamma Irradiator for Polymerization Summary 1. New Research Reactor projects are increasing in developing countries with many cases of first reactors being installed or planned. In industrialized countries, new reactors planned are generally high performance machines for specific purposes. Utilization statistics indicate good usage of research 1-8 reactors. The figures should improve if only general purpose reactors were considered and not include the zero power specific purpose devices. These reactors, 46 in number, have the lowest utilization. Utilization in developing countries show good" increases based on the data received from the Asia-Pacific region. The greatest use of research reactors in this region is in NAA and radioisotope production; NAA mostly for geological exploration and to a lesser extent for medical/biological applications and in agriculture. Radioisotope production is mainly for medicine, agriculture and industrial applications. Utilization in basic research is emerging. In the Asia-Pacific area, newly commissioned/recommissioned reactors in the Philippines, Bangladesh and Indonesia will improve the utilization in that area. The Agency has Technical Cooperation projects in neutron spectrometry (basic and applied research) in Thailand, Malaysia, Bangladesh, and Indonesia. It takes a minimum of about 5 years for a first research reactor in a developing country to establish a good reactor utilization program. The factors that influence this are many, mainly involving development of the necessary infrastructure within the reactor center and external to it. The time required to accomplish this should be recognized by those involved in programs to establish and improve utilization of research reactors. Research reactors can be and are being used to the national benefit by providing services to governmental, health, agricultural, industrial and educational sectors. 1-9 RESEARCH REACTOR UTILIZATION Quescionnaire Respondents Asia-Pacific Region Developing Countries Country Facility Power (kW) Cr it icality Comments Bangladesh TRIGA MK II 3000 1986 newly commissioned China HWRR-II 15000 1958 HFETR 125000 1979 SPR 3500 1964 SPRR-300 3000 1979 TSINGHUA 2800 1964 Indones ia TRIGA BANDUNG 1000 1964 training only PPBMI-BATAN 100 1979 RSG-GAS 30000 1987 newly commissioned Korea, Rep. of TRIGA MK II 250 1962 mostly training TRIGA MK III 2000 1972 AGN 201 0 1982 educational only Malays ia TRIGA PUSPATI 1000 1982 Pakistan PARR 5000 1965 Thailand TRR-1/M1 2000 1962 Viet Nam DALAT RR 500 1963 reconstructed 1983 1-10 WJ TENTH EUROPEAN TRIGA USERS CONFERENCE Atominstitut, Vienna, Austria, September 14 - 16, 1988 SAFETY INSPECTIONS TO TRIGA REACTORS W. Byszewski 1-11 Table of Contents INTRODUCTION OBJECTIVES OF THE INSPECTION AREAS OF THE INSPECTION 3.1. General 3.2. Nuclear Safety 3.3. Radiation Protection PERFORMANCE OF THE MISSION REPORTING POLICY SOME CONCLUSIONS FROM INSPECTIONS RELATED IAEA PUBLICATIONS 1-12 1. INTRODUCTION There are about 27 TRIGA reactors in 20 countries throughout the world outside the USA. Of these about 12 TRIGA's are under the Agency sponsored projects (see Table 1), whereby the Agency is entitled to examine the operational and radiological safety aspects of Agency assisted projects. To meet these needs the IAEA announced in 19 72 the availability of safety advisory missions to research reactors, including TRIGA reactors. In the most comprehensive form of safety mission, a team of experts assesses all areas affecting the ultimate safety of a particular research reactor. These missions are conducted regularly, about once each four years. Each mission comprises IAEA staff and if necessary external consultants. This operational safety advisory programme was created to provide useful assistance and advice from an international perspective to research reactor operators and regulators on how to enhance operational safety and radiation protection on their reactors. Safety missions cover not only the operational safety of reactors themselves, but also the safety of associated experimental loops, isotope laboratories and other experimental facilities. Safety missions are also performed on request in other Member States which are interested in receiving impartial advice and assistance in order to enhance the safety of research reactors. 2. OBJECTIVES OF THE MISSION The objectives of a safety mission are to conduct a comprehensive integrated safety assessment of the research reactor facility and to compare the safety of the reactor with the Agency's safety standards. The evaluation is also aimed at facilitating an exchange of knowledge and experience between the experts and reactor personnel. Safety missions are intended not to be a regulatory-type inspection that checks compliance with national requirements, but to achieve enhanced operational safety through application of effective practices used at other facilities around the world. 1-L3 3.' AREAS OF THE INSPECTION 3.1. • General The mission checks whether the safety at the reactor is subject to review by a regulatory organisation independent of the operating organisation and noting the frequency of regular inspections and any non-compliance observed by the inspectors. It checks whether a reactor safety committee, or an equivalent advisory group, exists to review safety problems arising in reactor operations and in planning of experiments. The mission also checks whether the Safety Analysis Report (SAR) is up to date including accident analysis and whether the SAR addresses current deterministic and probabilistic methodologies for safety assessment. 3.2. Nuclear Safety The nuclear safety review is mainly based on the requirements of the "Safe Operation of Research Reactor and Critical Assemblies" as it appears in the Agency's Safety Series No 35, 1984 edition. The review examines the following areas : 3.2.1. Safety specifications The mission checks whether there are approved safety specifications, including limits and conditions for the conduct of operations and experiments, and corrective actions in case of violation of the safety limits. 3.2.2. Periodic Testing and Inspection The mission checks the surveillance test intervals established for the different reactor systems, the availability of written procedures for testing and inspection and the compliance of test results with safety specifications. 3.2.3. Management The mission verifies the organisational structure of the Research Centre and Reactor Department for clearly defined duties and responsibilites in implementing and controlling facility activities. Particular attention is paid to training and retraining programmes, staff size, qualification and licencing of the operators, programmes for reactor utilization, quality assurance for operation, physical security and housekeeping. 3.2.4. Operating instructions During the inspection the mission checks whether all normal and emergency operating procedures required for guiding the operating personnel are in place, the personnel are trained and retrained in these procedures, and a system has been astablished for regular review of all procedures and for the communication of any revisions to the operating personnel and other holders of the document. 1-14 3.2.5. Records and Reports The mission checks whether ail essential -! records on design and operation related to safety are being maintained. 3.2.6. Maintenance The mission checks primarily the preventive maintenance programme for the reactor, the organisation of maintenance, the equipment available, and the procedures and documentation for maintenance. It also checks whether a system of work permits and approval after the maintenance is completed exists. 3.2.7. Experiments and modification The mission checks the assurance of safety in routine experiments and irradiations and the safety review process for new experiments or modification in the reactor systems. 3.2.8. Physical security The mission checks whether a physical security plan exists for the facility and whether it has been approved by the regulatory body. 3.2.9. Quality Assurance Programme for Operation The mission checks whether the operating organisation has a quality assurance programme duly reviewed and approved by the regulatory body that will govern the quality at all safety related items during the operations. Does the quality assurance programme cover periodic testing and inspection of components and equipment at a determined frequency. 3.2.10. Conduct of Operations The mission conducts a walk-through of the facility and asks questions of reactor personnel on matters relating to the reactor systems and the safety of reactor operation. The mission checks : a) Housekeeping in the facility b) Operational status of components and equipment important to safety c) Leak tightness of containment/confinement d) Proper demarcation of high radiation areas, e.g. experimental facilities, spent fuel storage, etc. e) Following up the approved procedures by the operators and accomplishment of the required checklists and forms. f) Function and calibration of instruments and systems. g) Knowledge of operating limits of the reactor and, of its operational characteristics by the operators. h) Reactor incidents and abnormal occurences with safety significances which may have occurred. i) Number of unplanned scrams per year (for the last three years) j) Future plans for the reactor facility (conversion, increase in power level) 1-15 3.3. Radiation Protection The mission examines the following areas : 3.3.1. Roles and responsabilities of management, the line orgnaisation and the authority of the radiation protection staff with regard to the operating personnel. 3.3.2. Staff selection, training and qualification. 3.3.3. Occupational radiation control, including equipment, facilities and procedures for external and internal dose control, and surveillance activities including dosimetry and monitoring activities. 3.3.4. Public radiation control, including equipment, facilities and procedures for control and monitoring of liquid gaseous and solid waste discharged into environment and environmental monitoring. 3.3.5. Emergency planning and preparedness at on-site and off-site organizations responsible for responding to nuclear accidents and radiological emergencies. 4. PERFORMANCE OF THE MISSION The members of a mission study information provided in advance by the research reactor organisation to familiarize themseivea with the reactor, its main design features, operating characteristic»» and the organisation of the reactor operation. This information is contained in a questionnaire which is filled out by the reactor operator. At the reactor site members of the mission : - examine the safety documentation of the facility, - review the oprational status of the reactor,and observe, if possible, a reactor start-up and shutdown, - discuss technical details with the responsible personnel. At the conclusion the mission's principal findings and recommendations are discussed with the senior management of the operating organisation and representatives of the regulatory authority. IAEA safety inspections to TRIGA reactors are shown in Table 2. 5. REPORTING POLICY After examining in depth the operational safety and the radiation protection aspects of the reactor operation the mission orally convey their findings and recommendations to the relevant authorities (operating organisation and regulatory body) during the final meeting. Shortly afterwards, a final mission report with conclusions is submitted through official channels to the Member State concerned. Further distribution of the report is at the discretion of the requesting Member State. 1-16 6. SOME CONCLUSIONS FROM INSPECTIONS The results of the inspections have shown that in some countries there are problems with radiation protection practices and nuclear safety. Very often the Safety Analysis Report is not updated, regulatory supervision needs clarification and improvement, maintenance procedures should be more formalised and records and reports are not maintained properly. In many cases population density around the facility has increased affecting the validity of the original safety analysis. 7. RELATED IAEA PUBLICATIONS 1. IAEA Safety Series No 35 State Operation of Research Reactors and Critical Assemblies (1984 Edition) 2. IAEA Safety Series No 9 Basis Safety Standards for Radiation Protection (1982 Edition) 3. IAEA Safety Series 74 Safety in Decommissioning of Research Reactors (1986 Edition) 4. IAEA TECDOC - 348 Earthquake Resistant Design of Nuclear Facilities with Limitied Radioactive Inventory (1985 Edition) 5. IAEA TECDOC-400 Probabilistic Safety Assessment for Research Reactors (1987 Edition) 6. IAEA TECDOC-403 Siting of Research Reactors (1987 Edition) 7. IAEA TECDOC-448 Analysis and Upgrade of Instrumentation and Control Systems for the Modernisation of Research Reactors (1988 Edition) 1-17 Table 1. TRIGA REACTORS UNDER THE AGENCY SPONSORED PROJECTS TRIGA MODEL MAXIMUM RATING, 1:W(t< ) STADY STAFF PULSING Bangladesh Mark II 3.000 852.000 Indonesia Mark II 250 250.000 Malaysia Mark II 1.000 1.360.000 Mexico Mark III 1.000 2.000.000 Philipines Convers ion 3.000 650.000 Romania Mark II 14.000 22.000.000 Tailand Mark III 1.000 1.200.000 Turkey Mark II 250 1.200.000 Vietnam Reconstructed 500 - Yugoslavia Mark II 250 - Zaire Mark II 1.000 1.600.000 1-18 Table 2 RECORDS OF IAEA SAFETY INSPECTIONS TO TRIGA REACTORS Dates of Safety Mission Visit 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 Bangladesh Finland Indonesia x X Malaysia x X Mexico X Philipines x X Romania X Thailand x X Turkey X Vietnam X Yugoslavia Zaire 1-19 Maintenance of the TRIGA Mark II Reactor in Ljubljana and the Disposal of Radioactive Waste V.Dimic, G.Pregl and B.Pucelj "J.Stefan" Institute Ljubljana, Yugoslavia September 1, 1988 1 Maintenance of the Reactor The TRIGA Mark 33 reactor of the "J.Stefan" Institute in Ljubljana (Yugoslavia) with a power of 250 kW has been successfully in operation since 1966. During 1986 and 1987 the reactor was in operation for 1070 hours and 1000 hours, with 1018 MWh and 997 MWh of energy production, respectively (Pig. 1). In 1988 the same energy production is expected because of the broad isotope production programme, irradiation of silicon (doping) and neutron activation analysis are still very active projects utilizing the reactor in Ljubljana. During the week the reactor is usually in operation five times overnight. In order to improve the safety of the reactor's operation, the following improvements to the reactor's instrumentation were introduced: • new scrams were introduced; warning of too low a water level in the reactor tank, to high fuel temperature, or when the door of the dry cell is opened; - new, parallel instrumentation for all four channels was put into operation; - new instrumentation for measurement of the activity of the secondary water was installed; - an automatic system to stop the secondary water flow when the reactor is shut down was introduced. No major events which interrupted reactor operation have occurred. 2 Temporary storage of low and intermediate level nu clear waste The low level nuclear waste disposal building at the TRIGA reactor in Ljubljana was built when the reactor went critical and it is rather small (20m2). Therefore it was 1-21 decided in 1986 to build a new one (Fig. 2) with, a bigger area and all the neccessary equipment for the safe temporary storage of low and intermediate level radioactive waste which has been generated at our research reactor, in hospitals and industrial operations in the Republic of Slovenia where about 2 million people are living. The waste site was selected very near to the reactor. The area of the building is 250m2 together with the decontamination area and the office. The volume is 850m3. It is partially underground and is grassed over. The storage site was put into operation in May 1987 and it is used only for solid waste because close near the site is an underground water reservoir. Therefore, no gas or liquid waste can be stored there. Before the operation of the storage facility a survey was carried out to measure the radiation level inside and outside the facility. Inside the the storage facility dose rate was 60/xGy/month which is the natural background. As a consequence of the Chernobyl fallout the outside dose rate is higher (90/j.Gy/month). A regular monitoring programme has been carried out öfter the storage facility became operationaly. The programme consists of the measurements of the dose rate at 7 places inside and at 3 places outside the storage facility. The measurements are performed every week. There are now 84 drams in the storage facility with various radioactive wastes. A selected drum is monitored by a gamma-ray detector before being stored. From this analysis it was found that the following isotopes are in the waste: Mn-54, Co- 60, Zn-65, Sr-90, Cs-137, Eu-152, Ra-226, U and Am-241. A separate section is used to store the defective rotary specimen rack (in special concrete shielding) and some other bulky waste. The integral activity in the facility is estimated to be between 3300 MBq (min) and 23.000 MBq (max). We can conclude that a proper and flexible waste storage site at the reactor is a very important facility device which can be used by other institutions and industry. Radioactive waste is temporarily safely stored only when it is properly managed. And the objectives of radioisotope waste management must be: - to comply with radiological protection principles - to minimise any impact on future generations to the maximum extent practi cable - to preserve the quality of the environment. These objectives can be fulfilled only when the disposal of waste is properly managed by people with good knowledge of radiological protection. Therefore, it was a good idea to collect all radioactive waste in our Republic which can be safely stored at the Reactor Centre in Ljubljana. Of course, the Krsko Nuclear Power Plant and the "Zirovski vrh" Uranium mine have their own storage facilities.. 1-22 M Wh r 300 _ Jl 280 _ IL IL fli« 260 _ 240 220 I 200 4 to lu 180 X 160 JIn m 140 _ 1 T, 120 _ rü rÜl In 100_ Ln 80 60 _ 40 _ eISn s 8 £ SR O CM S s 8 s PÏ CS CO 20 to £8 CO so in £ o m CO S Ol in in in u> ER m to o o 3 o> 8 IO I r i "• i • •—i—~~i—"—i—" i ' * i—-—I—"—i—" i i—• ' i • "—r-~—r~*-"—i—"—i—"—x———i——r 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 198i 8 Rg.1 Thermal power generation <^i«rfrirmi?Àà • •:.;;,:$•• IM '-WW- I -1^ Fig. 2 Waste disposal building OPERATION AND UTILIZATIONS OF DALAT NUCLEAR RESEARCH REACTOR P. Z. Hien Dalat Nuclear Research. Institute, Dalat, Vietnam. Abstract The reconstructed Dalat nuclear research reactor was commissioned in March 1984 and up to September 1988 more than 6200 hours of operation at nominal power have been recorded. The major utilizations of the reactor include radioisotope production, activation analysis, nuclear data research and training. A brief review of the utilizations of the reactor is presented. Some aspects of reactor safety are also discussed. Introduction. In Dalat (South Vietnam) the 250 Kw TRIGA MARK II reactor, which was in operation from 1962 to 1968 was completely out of action after the removal of the fuel elements just before the end of the war in 1975. The newly reconstructed 500 Kw reactor of Dalat Nuclear Research Institute is unique of its kind in the world: Soviet - designed core and control system installed in the undismountable infrastructure of the former TRIGA reactor, which includes the reactor tank and shielding, graphite reflector, horizontal beam tubes, thermal column etc ... [1]. The new reactor reached initial criticality in November 1983 and regular operation commenced in March 1984. Up to September 1988 more than 6200 hours of reactor operation at nominal power have been recorded. Reactor structure and characteristics. The fuel elements are of Soviet standard type WR-M2 with an enrichment of 36% U-235 and are clad with Al. The fuel assembly which are 60 cm long consists of a hexagonal shaped outer tube and two cylindrical inner tubes. Inside the core are inserted 7 control rods (2 safety rods, 4 shim rods and 1 regulating rod). Outside and around the reflector are positioned 9 ion chambers for reactor control purpose. All shim and safety rods are made of boron carbide, while the regulating rod is made of stainless steel. The total reactivity worth of shim rods is 12.8S reff 1-25 Fig. la £ Cross-section of the fuel element WR-M2 F ig, •{ V Reactor Core Arrangement 0 Fuel element u/p Beryllium 13J Safety rod "^ Shim red Regulating rod öo Pneumatic tube (13-2) ^ Pneumatic tube Wer. irradiation channel £ch\ Neutren trac 0) 1-26 The reactor core ' is located in the tank structure of the former TRIGA reactor and surrounded with a thin layer of beryllium blocks together with the graphite reflector of the former TRIGA reactor. There are 89 fuel assemblies in the reactor core. Seven unit cells at the center of the core are occupied by a water - beryllium neutron trap, which has an inside diameter of 65 mm. (Fig. I) The core cooling mechanism is basically natural convection as in the former TRIGA reactor, but the driving effect is enhanced by a 2 meter long chimney placed directly above the core. The whole core - chimney assembly is suspended from the top of the reactor tank. The driving effect of the chimney and the increase of the reactor power result in the increase of the dose rate above the water surface (mainly from N_16) as compared to the former TRIGA reactor. Thus a rotary platform made of steel with lead glass windows installed on the reactor tank significantly reduces the dose rate on the top floor of the reactor. Experimental and Irradiation Facilities. Experimental and irradiation facilities consist of the following : - The water - beryllium neutron trap in the center of the core 13 2 ( f = 2.1 x 10 n/cm sec). th The rotary specimen rack assembly with 40 irradiation holes located in the graphite reflector. 12 2 ( f = 3.2 x 10 n/cm sec ) th - A pneumatic transfer system installed near the beryllium reflector. This system is used for sample irradiation in off-line activation analysis. - Another pneumatic transfer system with dry irradiation channel installed in the core for simultaneous analysis of uranium and thorium by delayed neutron counting technique. - A fast pneumatic transfer system installed in the thermal 10 2 column ( f = 4.1 x 10 n/cm sec) for short-lived activation th analysis with pure thermal neutrons. The neutron beam facility at the tangential horizontal 6 2 channel A thermal neutron flux f = 5 x 10 n/cm sec is th available at the beam port for thermal neutron capture studies and transmission experiments. - The radial horizontal channel, which pierces the graphite reflector and terminates adjacent to the fuel will be equiped with neutron filters (Si,S,V,Mn,Fe,Ni). Monochromatic epithermal neutrons will be used for neutron physics studies in cooperation with the Kiev Institute for Nuclear Research. 1-27 Reactor Operation and Utilizations. At present the reactor is operated in two weeks cycle with 75 hours of continuous operation at nominal power.The accumulated operation time from 10/3/1984 to 10/9/1988 is 6219 hours. The reactor has been operated in good condition except for a two- months shut-down period in 1987 for maintenance of the control system and installation of the 1-131 production line. The data presented in Table I show the trend to better utilization of the reactor operation. Three radioisotopes are regularly produced and supplied to hospitals : 1-131, Tc-99m generator and P-32. Table I. Operation and utilization Statistics of Dalat NRR. 1984 1985 1986 1987 1988 (from 10-3-1984) (to 10-9-1988) Operation time 922 1722 1403 1022 1150 (hours) Isotope Production 5.2 13.2 27.7 34.2 45 for medical uses (curies) Irradiation Samples 1200 1700 "3000 "4000 "2000 for NAA At present and even in the near future the capabifcy of radioisotope production is still higher than the domestic demand in in-vivo diagnosis. A variety of radioisotopes are being also produced for the applications of tracer technique in agriculture (mainly P-32 for the investigations of soil-plant relationship ) hydrology and industries. Apart from isotope production and routine activation analysis, other research activities involving the use of the reactor can be mentioned as following : Development of simple Tc-99m extraction generator by using inorganic ion-exchangers titanium and zirconium-molybdate as irradiated targets. This type of generator are being regularly and successfully used in hospitals [2], - Investigation of the reactor-produced excitation sources for X-ray fluorescence analysis. Excitation source Ge-71 is routinely produced and successfully used [3], Determination of composite nuclear constant k (used in o the standardization of multielement activation analysis). A simple method of k -factor determination had been proposed [4] o and new data for short - lived isotopes are being currently 1-28 obtained by using fast pneumatic transfer system installed in the thermal column. - Experimental neutron silicon doping in the neutron trap. - Use of high gamma dose rate in the neutron trap after reactor shut-down for radiation chemistry investigations. - Analysis of biological and enviromental materials. The concentration levels of trace elements in human hair (Cu, Mg, Mn, Zn, Co, Cr, As, Sb, I, Hg) were obtained for the population of HochiMinh city. - Use of thermal neutron beam for elemental analysis by (n, gamma) and transmission techniques. Some aspects of reactor safety. The measured temperature coefficient of the reactivity is -2 negative and equals to 2 x 10 K / C. This value is typical r eft for the Soviet-designed water-cooled heterogeneous research reactor [5]. However this value is associated with the coolant - moderator, while in the case of the former TRIGA reactor the negative temperature coefficient is nearly prompt and associated with the homogeneous fuel - moderator elements. The prompt temperature coefficient of the reactivity is more lower and can -3 be evaluated as " 1 x 10 p / c [5]. The preliminary reactor reff safety analysis has been performed by the Soviet designers, and the attention was paid to the problem of the response of the reactor to the reactivity insertion. Some severe postulated events had been investigated : - Spontaneous withdrawal of shim rod at nominal and zero -5 (10 % P ) power operation (v = 3.4 mm/sec). o - Spontaneous withdrawal of regulating rod at nominal power operation (v = 20 mm/sec). - Downfall of fuel assembly leading to the reactivity insertion of 1.3 ft in 0.3 sec. reff The calculations taking the action of the protection system into account confirmed the safety of the reactor in every cases. Special attention is paid also to the problem of the quality of the tank structure of the former TRIGA reactor, which became operational more than 25 years ago, but it has been kept idle for many years before reconstruction. Although careful inspection for water-tighness performed before reconstruction works on the reactor tank,the graphite reflector and the horizontal channel showed their normal state, however, there is still concern about the possible failure of these components. Routine monitoring for the primary circuit water by low-background gamma spectrometry method shows rather low specific activity of such corrosion 1-29 65 60 51 187 products, as Zn, Co, G, W ... Besides these radionuclides, fission products as 135Xe and 88Kr were also detected at more lower levels. However, no increase of the specific activity of these radionuclides was observed during 4.5 years of reactor operation. With the technical assistance of the IAEA in this year the tank structure and the beam tubes will be inspected with underwater telescope and endoscope. The fuel cladding temperature will also be measured. A thermometric fuel element with 9 thermocouples is now prepared in Soviet Union and the experiments will be performed in early next year. It is hopeful that the results of the above mentioned experiments and inspections will serve a basic for the future plan of operation of the reactor and the possibility of its power upgrading. References [1] Pham Duy Hien. Proc. of first Asian Symposium on Research Reactor Tokyo, Nov. 1986, p.124 [2] Le Van So. IAEA Seminar on Radionuclide Generator Technology Vienna 13 - 17 October, /1986/ [3] P.Z.Hien, N.T.Ann, D.D.Thao, T.Truong. J. Radio and Nucl. Chem., Letters, 118 /3/,217, /1987/ [4] P.Z.Hien, T.K.Mai, T.X.Quang, T.N.Thuy. J. Radio and Nucl. Chem., Letters, 105 /6/,351, /1986/ [5] Archengelski N.V., Dikarev V.C., Egorenko P.M., Rjazanchev E.P., Sov. Atom. Energy 64,/5/,331,/1988/ 1-30 NINE YEARS OF OPERATION OF ITU-TRR TRIGA MARK-II REACTOR Yavuz H. , Bayülken A.R. , Yavuz M.A. Istanbul Technical University Institute for Nuclear Energy ABSTRACT ITU-TRR TRIGA Mark-II reactor in Istanbul with a steady state power of 250 kW and a pulsing capability up to 1200 MW has been operating since March 11,1979 with an energy release of 107.5 MWh and a total of 72 pulses. During this nearly nine years, the reactor was in operation without any major un- desired shut down. One of the major problems was faced when the instrumented fuel element in position 9 of the F ring went totally out of order.Secondly, the cooling tower of the secondary cooling system could not be operated properly during the hot summer days, and also we had a tar leakage problem with the radial beam port connection to the tank. During the regular maintenance work in this summer, the measurements of changes in nuclear and physical parameters of the reactor fuel and dummy elements have also proceeded. I.INTRODUCTION . I.T.U. Institute for Nuclear Energy has gained just nine and a half years of operation and maintenance experience with its TRIGA Mark-II type reactor. Presently, the reactor core contains 67 standard fuel elements, 2 instrumented fuel elements, 3 control rods, 16 graphite dummy elements and the neutron source.This core configuration has not been changed since the final core loading in May 1979 [Fig.l]. The reactor generally has been used for both training and research purposes since March 1979. The main subjects of our research program has been the study of neutron activation analysis, aplications of radioisotopes, neutrography (using the reactor neutron beam tubes) and pulsing. This programme was also presented in more detail at the Seventh European Conference of TRIGA Users (1982-Istanbul) [1]. Presently, Institute has neutron activation analysis and radioisotope aplication laboratories and a new neutron radio graphy facility has been designed which is under construction [2]. 1-31 1.Reactor Operation ITU-TRR TRIGA Mark-II reactor in Istanbul with a steady state power of 250 kW and a pulsing capability up to 1200 MW has been in successful operation since March 11,1979. From March 11,1979 to August 31,1988 the reactor had an energy release of 107.5 MWh and a total of 72 pulses. In this period, the reactor was mainly used for neutron activation analysis, irradiations and neutron radiography. Neutron radiography facility was installed in the tangential beam tube. Since that time, a number of modifications have been implimented to improve neutron radiography applications. As a result of the undesired effects of the facility on the neutrography quality, old facility was completely removed. A new neutrography facility was designed and it is right now under construction [2]. We have also started a study on using our TRIGA for sterilisation purposes, which gave us encourag ing results [3]. The ventilation system was designed and installed as main and emergency ventilation systems, but both systems could not be operated properly. The 67 standard fuel and 16 graphite dummy elements of the core were removed individually and measured underwater for elongations, as well as being inspected visually for physical damage. The elongations are measured with the fuel element measuring tool and the measurement is made in units of mils (1 mil=0.001 inch.) deviation from the length of the standard element. We have compared our measurements with the ones made on the initial loading date March,1979. The elongations' of the elements were found to be less than our measurement error. So, we have concluded that the elongations are at a very low level. The thermocouples of the instrumented fuel element at posi tion F-9, were completely out of order and only one of the three thermocouples of the instrumented fuel element at posi tion B-l can be used. For safe operation of the reactor, the instrumented fuel at position B-l will be moved to position F-10 and a fresh instrumented fuel element will be loaded to position B-l. Due to the inability of the original cooling tower, during summer days, the reactor could not be operated at full power for a long time. To solve this cooling problem, a new cooling tower with a capacity of 275 kW has been purchased. It will be installed parallel with the original unit by the Construc tion Department of the University. 1-32 2.Tar Leakage In The Radial Beam Tube During the inspection of the reactor components in July 1988, the concrete plug was found to be stuck inside the radial beam tube. When the concrete plug had been pulled out, a tar leakage was discovered from the gap between the aluminium tank and the aluminium radial beam tube [Fig.2,Fig.3]. Some of the leaked tar was still dripping inside the beam tube [Fig.4] and the leaked tar had covered the front side of the concrete beam plug [Fig.5]. Leaked tar was examined due to its activity and we found that it had a specific « activity of 3.7 Bq/gr and a specific ß activity of 431.2 Bq/gr. The gamma analysis of the leaked tar with its gamma active contents can be seen in Table 1. Table 1 - Gamma analysis of the leaked tar Specific Specific Isotope Activity Isotope Activity (Bq/gr) (Bq/gr) Zn-65 81.24 W-187 4.00 1-131 50.75 Sb-124 1. 45 Co-SO 35. 30 1-135 1. 28 Co-58 8. 51 K-40 1.04* Kr-88 7.60 Mn-59 1. 00 Fe-59 7.20 Cs-134 0.93 Ba-133 5.36 Bi-207 0. 14 Co-57 4.93 Sn-113 0.11 Cr-51 4.40 Hg-203 0. 10 _,,. .. Xe-133 4. 12 *It might be airborne. EXPERIENCE AND RESULTS After more than nine years of successful operation and uti lisation of the reactor and our experiences have shown that ITU-TRR reactor can be safely operated in the steady-state and the pulsing modes. 1-33 We have "ten fresh fuel elements and adequate selected spare parts in our stock and this year we have ordered some more spares to keep up the reactor running. Some equipment failures happened and some of the components had to be replaced.In this nine years, besides the non-nucle ar events mentioned in this paper, there were also many other minor troubles that are not worth of mentioning. The reactor has provided a valuable assistance to our univer sity, other universities and research centers. 70% of the reactor facility accounts were used for direct education and the rest for research activities. A great number of M.S. and Ph.D. thesis have have been done on or using TRIGA and also nuclear, electrical and physical engineering students from various universities have completed their practical and technical study at our reactor on summer months. We have provided a ${6000 support from The Turkish Atomic- Energy Authority for our " Modernisation of ITU TRIGA Mark-II Reactor In Accordance with the Industrial Needs " project. REFERENCES 1-AYBERS N., YAVUZ H., BAYÜLKEN A.R., "Planning and Imple mentation of Istanbul Technical University TRIGA Research Reactor Program", Seventh European TRIGA Users Conference, Istanbul-Turkey, 15-17 Sep. 1982. 2-TUGRUL B., BILGE A.N., "Modernisation Design of Neutron Radiography of ITU TRIGA Mark-II Reactor", Tenth European TRIGA Users Conference, Wien-Austria, 14-16 Sep. 1983. 3-BILGE A.N., TUSRUL B., YAVUZ H., "The possibility of Gamma Ray Sterilization By Using ITU TRIGA Mark-II Reactor",Tenth European TRIGA Users Conference, Wien-Austria, 14-16 Sep. 1988. 1-34 Pulsing ¥ detector /~~*\ (77) f*\ ^_ Safety O n©Xr^©rs O Channel Qs) Ay X o X M^ W o®l©o©®tr Wide range O detector Safety Channel (TR) Transient Rod fs J Am - Be Neutron Source (SR) Safety Rod (cT) Central Thimble (RR) Regulating Rod Graphite Reflector (JF) Instrumented Fuel (PN) Pneumatic Tube (~°\ Standart Fuel Fig. 1 - Final core loading of IT-U-- TRR 1-35 27)6 Fig.2- Radial Beam Port Assembly (Plan Section-All Dimensions Are In mm.) Fig.3- Radial Beam Tube During Construction 1-36 Fig.4- Radial Beam Port Front Vi ew Fig.5- Radial Beam Port Concrete Plug 1-37 î \ 'INITIAL OPERATION AND UTILIZATION OF THE BANGLADESH 3 MW TRIGA REACTOR By M.A. Mannan and Moazzem Hossain Bangladesh Atomic Energy Commission Dhaka, Bangladesh ***** ABSTRACT A 3 MW TRIGA MK-II pulsing type research reactor fuelled with low enrichment uranium having 19.7% U-235 and 20 wt % Uranium, 0.47% Erbium and Zirconium Hydride, has been installed at the Atomic Energy Research Establishment, savar in the last week of October, 1986. This multi-purpose reactor, capable of both steady-state and pulsing operation, has been put into service in several disciplines since its commissioning and presently in operation without any major problem. The paper describes the initial operating experience and the reactor utilization made in several areas including the operator training conducted for the formation of the initial crew for the reactor. I. INTRODUCTION A 3 MW pulsing type TRIGA MK-II research reactor has been installed at the Atomic Energy Research Establishment (AERE), located at Savar, 40 Km from Dhaka. This is the first nuclear reactor in the country. The reactor will be utilised for research, manpower training and production of radioisotopes for their uses in agriculture, medicine and industry. The fuel loading of the reactor started in the afternoon of September 13, 1986 and the reactor went critical in the early morning of Septemebr 14. The reactor achieved full power (3 MW) level on October 1, 1986 and all the required reactor testings were completed in the last week of October 1986. 1-39 The reactor is fully fuelled by soild, homogeneous mixture of E-U-ZrH alloy containing 19.7% U-235 and 20 Wt % Uranium and Zr-H-j_ Q. The reactor was tested at 3 MW steady-state power and was pulsed with the present core arrangement upto $ 2.00 step reactivity insertion achieving a peak power level of about 852 MW. Table 1 gives the principal design parameters of the reactor. Since commissioning, the reactor has been operating upto the full power level, as and when required, within certain limitations such as, non-availability of sufficient number of licensed operators in the initial crew and non-completion of some research and experimental facilities around the reactor. 2. DESCRIPTION OF THE REACTOR AND DESIGN PARAMETERS 2.1 Reactor shield The reactor shield is a reinforced concrete structure standing 7.9 m above the reactor hall floor. The lower octagonal portion is 6.6 m across the flats. The beamports are installed in the shield structure with tubular, penetrations through the concrete shield and the reactor tank water and they terminate either at the reflector assembly or at the edge of the reactor core. The reactor core and the reflector assembly are located at the bottom of a 2 m diameter aluminium tank, 8.2 m deep. Approximately 6.4 m of déminera 1ised water above the core provides the vertical shield. The radial shielding of the core is provided by a minimum of 2.29 m of concrete having a minimum density of 2.75 g/cm , 45.7 cm of water, 19 cm of graphite and 5 cm of lead. The heavy shield was made using locally available ilmenite and magnetite from beach sand of Cox's Bazar. 1-40 The provision of thermal column has been kept in the lower portion of the shield structure. But at present there is no graphite and the column has been filled with blocks of heavy concrete. 2.2 Reactor Core The reactor core is at the bottom of the reactor tank, which has a 0.63 cm thick wall having an inside diameter of 2 m and a depth of 8.2 m. The reactor core and reflector assembly is a cylinder approximately 1.1 m in diameter and 0.89 m high. The reactor core consists of a lattice of fuel- moderator elements, graphite dummy elements and control rods. The core is surrounded by a graphite reflector and a 5 cm thick lead gamma shield. This entire assembly is bolted to a support stand that rests on the bottom of the reactor tank. The outer wall of the reflector housing extends 0.3 m above the top of the core to ensure retention of sufficient water for after-heat removal in the event of a tank drain accident. Cooling of the core is provided by natural circulation upto 500 kw power level and by forced down flow circulation of tank water for higher powers which is, in turn, cooled and purified in external coolant circuits. In case of loss of cooling water in the reactor tank there is a provision of emergency core cooling system with roof-top back up system. The top grid plate is an anodized alumium plate of 3.17 cm thick. There are 121 holes of 3.82 cm diameter in six hexagonal bands around a central hole for locating the fuel- moderator and graphite dummy elements, the control rods, transient rod guide tube and the pneumatic transfer tube. There are 6 holes of 1.58 cm near the G-ring of the grid 1-41 plate for locating and providing support for the neutron source holder at alternate positions. A hexagonal section can be removed from the centre of the upper grid plate for inserting specimens upto 11.2 cm in diameter. Two other sections are cut out of the upper grid plate, for inserting specimens upto 5.2 cm in diameter. The bottom grid plate is an anodized aluminium plate 3.17 cm thick which supports the entire weight of the core and provides accurate spacing between the fue1-moderator elements. The safety plate of 1.9 cm thick aluminium is provided to preclude the possibility of control rods falling out of the core. The active part of each fuel-moderator element is approximately 3.63 cm in diameter and 33.1 cm long. The fuel is solid, homogeneous mixture of E-U-ZrH alloy containing 20% by weight uranium enriched to about 19.7% U-235 and about 0.47% by weight of erbium. The H/Zr ratio is approximately 1.6. Each element is clad with 0.051 cm thick stainless steel can. Two sections of graphite are inserted in the can, one above and one below the fuel, to serve as top and bottom reflectors for the core. 2.3 • Experimental and Irradiation Facilities The reactor has extensive experimental facilities. It can be used to provide intense fluxes of neutron and gamma for research, training and radioisotope production. The experimental and isotope production facilities of the reactor consists of the following: 1-42 (a) The rotary specimen rack assembly (Lazy Susan) located in the circular well in the reflector assembly. This is the primary facility for the production of radioisotopes in the reactor. (b) Production of very short-lived radioisotopes is accomplished by a pneumatic transfer system located in the G-ring of the core. (c) Another faster pneumatic transfer system is being installed for research purposes. (d) One central experimental tube (Central Thimble) in the middle of the core (A-ring) for in-core irradiation at the region of maximum neutron flux. (e) Three 15.2 cm diameter radial beamports, one of which pierces the graphite reflector and terminates adjacent to the fuel. (f) One 15.2 cm diameter tangential beamport. (g) Other in-core irradiation facilities, such as hexagonal and triangular cutouts etc. 3. INITIAL OPERATING EXPERIENCE Since the commissioning of the reactor in October 198 6, the reactor has been operating at different power levels upto the full power level without any major problem. However, the reactor opertion was limited upto 3 MWH for more than a year due to a restriction imposed by the national Nuclear Safety Committee (NSC). This restriction was imposed since the reactor hall ventilation system was being run manually without automatic 1-43 controls of the fans and dampers as well as due to availability of only one licensed senior reactor operator and one reactor operator initially. This restriction was, however, relaxed upto 12 MWH later by the NSC after the installation of the above automatic controls. The reactor facility is operated and maintained by the engineers and scientists and other technical personnel of the Reactor Operations and Maintenance Unit (ROMU) of the AERE, while the health physics and radiation protection support is being given by the Health Physics and Radiation Protection Division, Institute of Nuclear Science and Technology (INST). Moreover, the Nuclear Chemistry Division of INST has been carrying out reactor water analysis on regular basis (as a parallel check to the measurements given by the reactor in-built instrumentations) as well as performing other detailed analysis of pool water in order to maintain the desired water quality in the reactor. In the absence of any Radiation Protection Act, which is now under consideration of the government for promulgation, as well as due to non-availability of any local codes and standards regarding reactor operations, the reactor is being operated, maintained and all documentations kept as per USNRC practices (to the extent possible) as well as the recommendations of the reactor manufacturer. 3.1 Operating statistics The reactor was operated for one shift only mostly for operator training, neutron flux mapping and spectrum unfolding, soil analysis (through activation analysis) as well as some trial production of radioisotopes such as Tc- 99m and 1-131. However, these operations were limited due to various reasons mentioned earlier including non-availability 1-44 of various experimental facilities around the reactor. Tablerll shows the reactor operating statistics during the period from the date of attainment of first criticality of the reactor upto July 31, 1988. The reactor remained shutdown for about 2 months during June-August 1987 due to an accidental breakage of the inner graphite plug of the beamport (see Section 3.2:). It may be mentioned here that in the third week of June last, 4 Senior Reactor Operators (SRO) and 1 Reactor Operator (RO) were provisionally licensed for the Bangladesh Reactor after qualifying in the RO / SRO examinations conducted by an IAEA expert as per USNRC standards. These operators were trained locally in both theoretical and practical aspects of operation and maintenance in the Bangladesh reactor. It is, therefore, expected that with the above manpower and after the availability of the various experimental facilities (now being developed) around the reactor, the reactor could be utilised extensively soon on a routine basis at least upto two shifts of operations. 3.2 Reportable Occurrence There was only one reportable occurrence, that happened since the commissioning of the reactor. The Nuclear Safety Committee (NSC) was informed about the incident formally on June 25, 1987. A description of the incident is given below: The plugs of the piercing beamport of the research reactor were removed on June 17, 1987 for the purpose of flux mapping at the beamport face. On completion of the above experiment, efforts was made next day (June 18, 1937) to insert first the inner graphite plug with the help of a GA-supplied plug handling shielded cask. The cask (being 1-45 very heavy) was being handled by several operations personnel. But, the inner graphite plug during insertion got broken (sheared off) at the tail-end near the graphite- steel/ lead interface due to sudden side-way jerks of the cask during its alignment process with the central line of the beamport. The broken (front) portion of the graphite beam plug was later removed from the beamport with the help of a long hooked metallic flatbar taking all radiological protections including erecting lead and heavy concrete shielding blocks around the beamport face. The broken graphite portion was safely stored in the active storage room adjacent to the reactor hall while remaining (steel/ lead) portion of the inner plug was kept inside the plug handling cask (till its reinsertion in the beamport later). It may be mentioned here that the broken graphite plug had little radioactivity (about 50 u rem/hr) along most of the length excepting at the tip, where the level was about 5 m rem/hr at the surface. Later, on receipt of authorization from the NSC and as advised by GA, the reactor manufacturer, the reactor was re started on August 10, 1987 after insertion of the remaining portion (steel-lead section) of the broken inner plug and the polyethylene outer plug and installing other components normally existent in the piercing beamport. The reactor was initially operated at 50 watts for the purpose of core excess measurement and the power level was gradually increased to the full power level (3MW) in several steps. The radiation level at full power at the piercing beamport face was found to be about 10 mrem/hr (neutron) and 5 mrem/hr (gamma). It may be mentioned here that during the initial reactor startup in September and October 1986, the radiation level at full power at the piercing beamport face was found to be <<1 mrem/hr (neutron) and <1 mrem/hr 1-46 (gamma), while the highest measured radiation level around the reactor shield face at the return coolant pipe near radial beamport was about 5 mrem/hr (gamma). Therefore, in order to cutdown the radiation level further at the piercing beamport face at full power, external shielding (about 37.5 cm deep) consisting of heavy concrete blocks ( ~ 2.75 gm/cm3 density) was added at the port face. The reactor was again operated at full power on August 11, 1987 and radiation levels were measured at the piercing beamport face. The level was then found to be about 0.3 mrem/hr (neutron) and 5 mrem/hr (gamma). However, the beamport face area had been cordoned off with nylon ropes upto a distance having radiation levels of about 2 mrem/hr (gamma) at full power. Since then, the reactor has been operating, as before, upto the full power level, as and when required, with the above-mentioned arrangement. 4. UTILIZATION OF THE REACTOR The Bangladesh TRIGA reactor has been installed with the following major objectives in mind: (a) to cater the immediate need of Bangladesh in the form of services such as supply of radio-isotopes including radiopharmaceuticals and material analysis (through various techniques such as Neutron Activation Analysis, Neutron Radiography and Neutron Scattering). (b) to develop trained manpower for future nuclear power plants and other facilities through training and research and (c) to perform applied research in the field of nuclear technology and various related branches of engineering. 1-47 Since the commissioning of the reactor until July 31, 1988, there were in total 68 irradiations made in the reactor in various areas namely neutron flux mapping and spectrum unfolding, soil analysis (IAEA-supplied samples as well as local Haripur soils) and trial production of some radioisotopes such as Tc-99m and 1-131. Table III gives a breakup of reactor use time in terms of specific use categories while Table IV shows a breakup of irradiations made in various disciplines till the above date. As mentioned earlier, some of the research and supporting laboratories around the reactor are expected to be completed in the near future. Moreover, necessary preparations for the experimental setups for neutrons radiogrpahy, and neutron scattering are at present being made for utilizing the reactor through IAEA assistance. As regards neturon radiography (NR), various components such as collimator, beam stopper, beam catcher etc. of the experimental set up have already been fabricated locally. Other equipment/components like Bi-filter, radiography cassette, etching bath, enlarger, printer, dryer, photographic films etc. for this purpose have been procured through IAEA. At present, the experimental setup using the tangential beamport is being erected. It is expected that the NR facility would be fully ready for use by the middle of September, 1988. 5. CONCLUSION Since commissioning in the later part of 1986, the Bangladesh research reactor is just passing through the initial phase of reactor operation and utilization as evidenced from the operating statistics outlined above. With the availability of more trained manpower for reactor operations as well as on 1-48 completion of some research and experimental facilities which are being developed around the reactor, it is expected that reactor utilization would be increased substantially in the very near future. 6. REFERENCE (i) Mannan, M.A., Hossain, M. "Characteristics and Facilities of a 3 MW LEU Fueled TRIGA Reactor", Proceedings First Asian Symp. on Research Reactors, Rikkyo University, November 18-21, 1936, PP. 82-93. (ii) Mannan, M.A, Hossain, M. and Whittemore, W.L. "Facilities and Utilization of the Bangladesh Multipurpose 3 MW TRIGA Reactor", Proc. International Symposium of the Utilization of Multi-purpose Research Reactors and Related International Co-operation, Grenoble, France, IAEA-SM-300/033, October 19-23, 1937. 1-49 TABLE PRINCIPAL DESIGN PARAMETERS OF THE 3 MW TRIGA MARK II REACTOR MAXIMUM STEADY-STATE 3000 KW POWER LEVEL MAXIMUM REACTIVITY .$ 2.00 INSERTION DURING PULSING FUEL ELEMENT DESIGN FUEL-MODERATOR MATERIAL U-ZrH (a) URANIUM CONTENT 20 Wt % (b) URANIUM ENRICHMENT 19.7 % U-235 SHAPE CYLINDRICAL OVERALL LENGTH OF FUEL 38 cm (15 IN) OUTSIDE DIAMETER OF FUEL 3.63 cm (1.43 IN) CLADDING MATERIAL STAINLESS STEEL (Type 304) NUMBER OF FUEL ELEMENTS 100 MAXIMUM EXCESS REACTIVITY 7.0% K/K (COLD, CLEAN) REACTIVITY LOSS DUE 2.5% K/K TO EQUILIBRIUM XE NUMBER OF CONTROL RODS SHIM/SAFETY 4 REGULATING 1 SAFETY/TRANSIENT 1 TOTAL REACTIVITY WORTH OF RODS $16.75 REACTOR COOLING FORCED DOWNFLOW OF POOL WATER (a) THE NOMINAL H/Zr RATIO IS 1.60 AND THE MAXIMUM VALUE IS 1.65 (b) BURNABLE POISON : 0.47 WT % ERBIUM. 1-50 TABLE -II REACTOR OPERATING STATISTICS ( As of July 31, 1988 ) Operational From Septenfcer 14 Fran January 1, From January 1 Cumulative from Data for to to to Septenfcer 14,198* LEU core Decenrfcer 31, 1986 Decen-ber 31,1987 July 31, 1988 to July 31, 1988 OPERATING 32.41 104.77 133.73 270.91 HOURS (TOTAL) MEGAWATT HDUR 26.54 68.60 218.46 313.60 MEGAWATT DAY 1.10 2.86 9.10 13.06 GRAM U-235 USED 1.35 3.52 11.20 16.07 HOURS AT FULL POWER 7.12 9.75 56.52 73.39 NUMBER 0? FUEL 100* 0 0 100 ELEMENTS ADDED TO CORE NUMBER OF 0 24 44 68 IRRADIATIONS MADE NO OF FTLSRS MADE 23** 1 5 29 Note: (1) The reactor achieved criticality for the first time on Septeirfcer 14,1986. (2) Final core achieved with 100 fuel elements (including FFCRs) for full paver operation during reactor camùssioning. ** (3) Pulsing operations made during reactor ccmra^sioning. 1-51 TABLE - III REACTOR USE TIME IN TERMS OF SPECIFIC USE CATEGORIES (As of July 31, 1988) Annual values Annual values Annual vlaues Cumulative Reactor use for the period for the period for the period values Category Septerrber 14 January 1 to January 1 to September 14, to December 31, December 31, July 31, 1988 1986 to July 1986 (Hours) 1987 (Hours) (Hours) 31,1988(Hours) TEACHING/ TRAINING 0 63.99 36.32 100.31 (WITHIN BABC) IRRADIATION (WITHIN BAEC) 0 17.33 54.18 71.51 IRRADIATION (OUTSIDE 0 0 0 0 BABC) VISITOR DEM3MSTRATICN 0.8 1.86 2.90 5.56 *REACTCR 122.61 401.59 323.16 347.36 USE TIME (OTHERS) TOTAL REBCTOR 123.41 484.77 416.56 1024.74 USE TIME Note: Consists of reactor usetime due to core excess measurements, console checkout and shutdowns each working day and load/unload of samples. 1-52 TABLE - IV BREAKUP OF IRRADIATIONS MADE IN VARIOUS DISCIPLINES { As of July 31, 1988) Irradiations Fran Sepfcarfcer 14 From January 1 From January 1 Cumulative from made to to to Seotember 14,1986 December 31, 1986 Decerrber 31,1987 July 31, 1988 to July 31, 1988 ISOTOPE 15 PKXÜCTICN ACTIVATION ANALYSIS 34 37 (SOIL ANALYSIS) ACTIVATION ANALYSIS (FLUX MAPPING 15 16 & SPECTRUM UNFOLDING) TOTAL 24 44 68 Note: A total of about 582.87 mCi of Tc-99m and 1-131 <,BS generated as trial productions. 1-53 OPERATIONAL AND RESEARCH ACTIVITIES OF TSING HUA OPEN POOL REACTOR by Tien-Ko Wang, Der-Ling Tseng, Huai-Pu Chou, and Minsun Onyang Department of Nuclear Engineering, National Tsing Hua University, Hsinchu 30043, Taiwan, ROC ABSTRACT Tsing Hua Open Pool Reaction (THOR) is the first nuclear reactor to become operational in Taiwan. It reached its first critical on April 13, 1961. Until now, THOR has been operated successfully for 27 years. The major missions of THOR include radioisotope production, neutron activation analysis, nuclear science and engineering researches, education, and personnel training. The THOR was originally loaded with HEU MTR-type fuels. A gradual fuel replacing program using LEU TRIGA fuel to re place MTR started in 1977. By 1987, THOR was loaded with all TRIGA fuels. This paper gives a brief history of THOR, its current status, the core conversion work, some selected research topics, and its improvement plan. 1-55 Introduction Three of the six research reactors in Taiwan are located at National Tsing Hua University (NTHU). They are, namely, the 1 MW Tsing Hua Open Pool Reactor (THOR), the 10 kW Tsing Hua Argonaut Reactor (THAR), and the zero-power Tsing Hua Mo bil Educational Reactor (THMER). The specifications and per formances of THOR, THAR, and THMER are listed in Table 1. This paper deals with operational experiences and utilizations of THOR only. THOR is the first nuclear reactor to become operational in Taiwan, ROC. It was originally designed by General Electric Co. (USA), was constructed in 1958 by the faculty members of NTHU, and went critical for the first time on April 13, 1961. THOR is an one megawatt (nominal), light water moderated and cooled thermal reactor. The irradiation facilities include a thermal column, 6 beam ports, 2 through ports, 11 irradiation tubes and 2 pneumatic tube irradiation system. The highest neutron glux reaches 1.6 x 101 3 n/cm 2-sec . Services and Utilization Isotope production is one of the major missions of THOR. Twenty-seven kinds of radioisotopes have been produced since 1961 in THOR. Up to now, more than 350 curies of radioisotopes were produced to partially meet demands from users including hospitals, research institutions, universities, and industries. 1-56 Irradiation and activation analysis using THOR have been 9 _ c well developed in the past two decades . Techniques of In strumental Neutron Activation Analysis (INAA), Radiochemical Neutron Activation Analysis (RNAA), and Prompt Gramma-ray Activation Analysis (PGAA) have all been developed using THOR neutrons« The above techniques have been used in the analysis of (1) semiconductor materials ; to determine ppb level of trace elements in Si, GaAs, etc., (2) biological materials: to determine trace elements such as Cd, Hg, As, Mo, Se, Co, Zn, Cu, etc. in various samples, and (3) environmental materials : to determine trace (toxic) elements such as Cd, Hg, and Pb in various environmental and food samples. 1 Conversion to TRIGA core The THOR was originally loaded with MTR plate-type HEU (93% enrichment) fuels. Later on, considering the availabili ties, the MTR fuels were gradually replaced by the TRIGA rod- type LEU (20% enrichment) fuels. As a result, from 1977 to 1987, the THOR had been operated on a mixed MTR/TRIGA core. Finally, in August 1987, all the remaining 33 MTR fuel assemblies were removed at one time and the core was loaded using all (four-rod cluster) TRIGA fuels. Figures 1 and 2 show respectively the THOR core configurations before and after the core convertion in August 1987. The power was calibrated by calorimetric procedure. The 1-57 experiments were performed in the high-power pool. The low- power pool was seperated by the high-power pool using an in- solated gate during the experiment (Fig. 3). Up to 15 thermal couples were used for coolant temperature measuremtn. The changes of water temperature with time were recorded and the reactor power was determined using the equation P=mC dT/dt where m is the water mass and C the specific heat. To increase the accuracy of this power calibration, two submergible pumps were used for better heat transfer. Furthermore, the coolant in the high-power pool was divided into 4 layers (layer I to IV, Fig. 3) and reactor power was calculated based on the aver age temperature slope (dT/dt) and the water mass of each layer. Table 2 illustrates the data obtained in one of the experiments. Some On-going Research Topics 1. Fuel burnup determination by iterative approach Due to the lack of detailed operational history, it is ex tremely difficult to determine the THOR fuel burnup values by calculations alone. Experimentally, it is most common to use -| n / TOT the Cs to Cs activity ratio to determine the burnup value. In addition to using the long half-lived Cs/ Cs ratio, another method using the short half-lived La is also deve loped at NTHU. The La activity was accumulated through re- irradiation of the spent fuel using THOR neutrons . A computer program was first written to calculate the ac- 1-58 1 Q/ i o "7 cumulation and decay of Cs, Cs and fuel burnup as well, based on the assumption of ~ 40 hr./week operational history of THOR with an initial guess of neutron flux. After the first iteration, the calculated Cs/ Cs ratio is compared with the measured value and the magnitude of neutron flux is modi fied accordingly for a next iteration« The iteration process goes on until the calculated Cs/ Cs values converge to the measured values. The burnup values are thus determined in the final iteration. Another computer program was written to calculate the 1 / C\ 0 *5 ^ La activity based on an initial guess of U density and the measured neutron flux (at the time of fuel re-irradiation in THOR). Similar iteration procedure is used until a conver- 235 ged value of U density (and thus the burnup value) is re ached. Four group cross sections generated through WIMS/CITATION code package were used in the above burnup calculations. Figures 4 and 5 show the burnup values, of two of the THOR spent MTR fuel, estimated based on 34Cs/ Cs ratio and La activity, respectively. The agreement between the two methods is quite good. The most distinct advantage of this method is that the information on detailed operation is not necessary. Especially when La is used.no operation history is needed. 2. THOR automatic power-regulating system The THOR power regulation is controlled by a servo blade connected to a two-phase servo shunt motor with tacho meter. The up-and down-movement of the servo blade and its moving 1-59 speed are determined by the difference between the (fission chamber) detected power and the pre-set (required) power. This power-regulating system can successfully handle the narrow-range power change and the Xe build-up. While in the wide-range power change, the phenomena of overshooting and undershooting have been observed (Fig. 6). To solve the problem of overshooting and undershooting, a microcomputer based power-regulating system is designed . The new system is controlled based on the comparison between an "available time" Ta and a "required time" Tr. The on-line test of the new system shows good control of power regulation (Fig.7). It is also found that if wide range power regulation is to be applied, the current THOR servo blade has to be replaced by a blade with higher reactivity worth in order to compensate for the negative temperature reactivity feed back. The soft-ware flow chart and hard-ware block diagram are shown in Figs. 8 and 9 respectively. 3. THOR neutron life-time and sub-criticality measurement This work is to use autoregressive moving average model (ARMA) to identify THOR neutron-noise transfer function. This can further be used in the measurements of neutron life-time g and sub-criticality . Based on ARMA method, the THOR neutron life-time was found to be ~94 ^sec (Table 3), while the re actor is in critical state. When the reactor was operaterated in sub-critical states A and B (Table 4), the sub-criticality was estimated using stable-period method,source-multiplication method, and neutron noise-analysis method, respectively. The 1-60 results are listed in Table 5. Studies are still going on for the improvement of the application of this noise-analysis tech nique. Scheduled System Improvement 1. A third pneumatic tube is to be added into THOR in year 1988. 2. An additional vertical irradiation tube is to be added into THOR core center in year 1988. 3. A GA designed microprocessor based reactor control system is to be installed at THOR in year 1989. OCT 4. A ^JiCf neutron-source-driven subcritical assembly using TRIGA LEU fuels is to be installed in year 1989. 5. Conversion of the THOR thermal column into a deep-penetra tion shielding-experiment assembly is scheduled in year 1989-1990. 6. A power pulsing mechanism is planned to add into THOR in year 1990. 7. THOR power upgrates to 3MW/5MW in year 1991. References 1. "Annual Report-1987", Nuclear Science and Technology Deve lopment Center, National Tsing Hus University, Taiwan, ROC (1987). 2. S. J. Yeh, et al, J. Chin. Chem., 31, 131 (1984). 3. C. Chung, et al, Int. J. Appl. Radiât. Isot., 36, 357(1985). 1-61 4. P. C. Liu and P. S. Weng, J. Rad. Nucl. Chem. L., 95, 81 (1985). 5. C. L. Tseng, et al, J. Anal. Chem., 321, 141 (1985). 6. D. C. Hsu, Master Thesis, Depart, of Nucl. Engr., National Tsing Hua University (June, 1988). 7. H. L. Wang, Master Thesis, Dept. of Nucl. Engr., National Tsing Hua University (June, 1988). 8. L. B. Liao, Master Thesis, Dept. of Nucl. Engr., National Tsing Hua University (June, 1988). 1-62 : TRIGA STD. A-D : vertical tube ure 1. THOR core configuration before core conversion 1-63 A H F : filler G : graphite S : neutron source A-D : vertical tube (movable) TRIGA (g) STD. 34 rods (£) L.L. 64 rods Figure 2. THOR core configuration after core conversion 1-64 I i 1 t x : thermal couple Figure 3. Layout of THOR power-calibration instrument arrangement 1-65 15 25 35 45 POSITION (CM) Figure 4. THOR F1-12 fuel axial burnup distribution 60 55 A BURNUP BY CS-134/CS-137 • BURNUP BY IA140 50 - 45 - 40 - 1^ —, , , , p- 10 50 60 20 30 40 POSITION (CM) Figure 5. THOR F1-24 fuel axial burnup distribution 1-66 Figure 6. THOR power regulating curve 3 M î 1 4 5 6 T (100 sec) Figure 7. THOR power regulating curve using new system 1-67 Start Hardware Initialization Software Initialization Power Signal = T£U (nf/n(t)) Input Tr Calculate Reactivity ri£: target power z : dynamic period Calculate max. contrôlable Reactor Period 'o c: reactivity Calculate eff. decay const. X&: Ta & Tr Calculate Rod motion Rod Diraction output Figure 8. Soft-ware flow chart of THOR new Servo control syetem 1-68 Fission Wide Range Analog to DMA Micro —>• CRT chamber > Log & Linear Digital Computer Amplifier Converter Direction Rod Position Rod Regulating Rod Driving J: Position Rod in System Reactor Core Figure 9, Hard-ware block diagram of THOR new servo control system Table 1 Specification and performances of THOR, THAR, and THMER Reactor THOR"" THAR THMER Construction began 1958 MAY 1973 OCT. 1973 APR. First criticality 1961 APR. 1974 APR. 1975 NOV. Thermal output 1,000 kW 10 kW 0.1 W Maximum £^ 1.6 x 1013 5 x 1011 5 x 106 (n h/cm2-s) t (at 1,000 kW) Nuclear fuel 20%235U-Rods(TRIGA) 20%235U-Plate 20%235U-Disks Type 90%ZOJU-Plates(MTR) Core dimension 18" x 15" x 24" 24MD x 24" 9.4"D x 11" Neutron Moderator Lighter water Light water Polyethylene Neutron reflector Graphite Graphite Graphite Reactor shielding H£0 + Concrete Concrete Pb + H3BO3 Irradiation tube x 11 tube x 2 beam port x 4 Facility beam port x 7 beam port x 1 Thermal column x 1 Reactor codent H20 H20 n/a Typical operation 50 hours/week 30 hours/week Upon request Full power operation 1800 hours/year 80 hours/year* 40 hours/year Cumulative operation 1600 MW^Day 30 kW-Day 50 W-Day Major Education Education Education Activity Training Training Training Research Research Service Reactor designer GE (USA) ANL (USA) NTHU (ROC) "At reduced power level of 0.1 kW -1- "Before conversion to all TRIGA core -1. "Since Aug., 1987, TH0R often operates at 1200kW. 1-70 Table 2 THOR power calibration data at nominal power 300 kW (based on fission chamber #1) before core conversion after core conversion H2O volume temp, slope power temp, slope power (°C/sec) (kW) (°C/sec) (kW) 47642 0.07034 234.58 0.07584 252.9 15581 0.04327 47.19 0.059 64.35 3936 0.05385 14.84 0.06966 19.19 8192 0.03641 20.88 0.06966 39.95 75651 317.49 376.39* Power indication on control panel were modified accordingly, 1-71 Table 3 Effective neutron life time determined by noise analysis reactor data blvarlate univariate condition length neutron lifetime neutron lifetime critical 1024 95.1 usEC 93.5 usEC critical 4096 93.6 usEC 96.5 usEC Table 4 Critical and subcritical states for experiments Blade Critical State A State B # 1 • 14.28 14.28 14.28 # 2 14.28 14.28 14.28 # 3 14.28 14.28 14.28 servo 20.00 16.00 12.00 Table 5 Subcriticality measured by three methods method state A state B stable period -0.2040 dollar -0.5018 dollar source multiplication -0.1444 dollar -0.3756 dollar noise analysis -0.1937 dollar -0.4205 dollar 1-72 V / / / OPERATING EXPERIENCES AT THE FINNISH TRIGA REACTOR Seppo Salmerihaara Technical Research Centre of Finland Reactor Laboratory SF-02150 Espoo Finland 1 INTRODUCTION The Finnish Triga reactor has been in operation for 26 years since March 1962. There are still 57 original Al-clad fuel elements in the core. So far we have had only two fuel cladding failures in 1981 and 1988. The first one was an Al-clad element and the second one a SS-clad. The lew rate of fuel cladding failures has made it possible to use continuously also the Al-clad fuel elements. Although some conventional irradiations of certain type have been repeated successfully tens of times, new and unexpected incidents can still take place. As an example an event of a leaking irradiation capsule is described below. 2 LEAKAGE OF AN IRRADIATION CAPSULE The reactor was in operation at the normal 250 kW power level. The radiation monitor measuring the primary water activity gave an alarm. The gamma spectrum of the primary water was checked with a scintillation detector and a multichannel analyzer'which belong to the reactor instru mentation. A new peak, approximately 666 keV was visible on the display of the multichannel analyzer but could not be identified. The reactor was shut down. A water sample was measured with a separate Ge-detector and multichannel analyzer and several 80Br and 82Br peaks were observed. According to the log-book there was a 20 g KBr sample in the central thimble to be irradiated. The sample was removed from the irradiation position. It showed out that the main part of the sample's activity was missing. The missing part of the activity was obviously still in the central thimble. The central thimble was cleaned by sucking water out of it. As a result of rinsing the tube there was 40 1 water containing about 40 GBq KBr activity. 1-73 The conductivity of the reactor water had risen from the normal 1.5 nS/cm, measured before the event, to 2.5 joS/cm after the leakage. Next morning the conductivity was again in the normal value. The radiation level in the water treatment room containing the ion exchange resin rose from the normal 10 \iSvfh to 200 pSv/h. The distance between the detector and the demineralizer is about 5 m. After one week the radiation level was again near the normal. The reactor was kept out of operation two days. After that the reactor could be taken again in normal operation. The reason for the leakage of the KBr sample was that the sample and the aluminium capsule had been already irradiated one week earlier but not used at that time. The sample and the capsule were stored in a lead shielding where seme drops of water came together with the capsule. In these circumstances due to the corrosion several pinholes were formed in the aluminium capsule, which caused the leakage of potassium bromide. To avoid similar corrosion damages a new statement was added to the standing orders of the laboratory: It is not allowed to load an already- earlier irradiated aluminium capsule into the reactor. A new sample and capsule must be prepared instead. 3 LEAKING FUEL ELEMENT After several hours' normal operation the activity meter of primary water gave an alarm and at the same time the aerosol monitor shewed higher aerosol activity than normal. The gamma spectrum of the aerosol- monitor's filter was analyzed with a Ge-detector and multichannel analyzer and several fission products were identified such as 88Pb and 138Cs. Next day, when the reactor was in operation, no fission products could be found during five hours' operating time. So during the following days either seme of the channels gave an alarm or not. At last only two hours' time was enough to get the alarms both in the primary water activity monitor and in the aerosolmonitor. To localize the leaking fuel element a new experimental arrangement was developed as shown in the Figure 1. In this method water is sucked just above the core, circulated by means of a pump through a half-open 4 1 container and led back to the reactor tank. In the container the fission gases, as 88Kr and 138Xe, escape frcm the water into the air. The air is further led to the aerosol monitor. The daughter products of the fission gases are then gathered in the aerosolmonitor ' s filter. The recorder connected to the aerosolmonitor was used to follow the changes in the activity of the filter in each experiment. To start one experi ment the sucking tube was set above the core on the place to be inspect ed, a new filter for the aerosolmonitor was changed and the reactor was started. First tests were made at the power level of 10 kW. This level was enough to localize the leaking element roughly. Later tests were made at 100 kW power level, which gave greater signals. The indication of the aerosolmonitor was about ten times higher when water was sucked above the leaking element compared to that situation when water was sucked far away from the leaking element. 1-74 After the leaking element had been localized to a certain area of the core, one suspected element at a time was removed from the core and the reactor was operated to check if fission products appear or not. At this stage water was sucked in each test from thé nearly same place, which produced the maximum indication of fission products. Typical tests needed less than 5 minutes' time of reactor operation. After several tests the leaking element was found. It took totally 8 hours during two days to find the damaged element. The leaking element was an instrumented, stainless steel cladded fuel element with one of the highest burn up values (14 %) in our reactor. The leaking point could not be find visually using an underwater TV- camera. The search tests also confirmed that the leakage was a rather small one, because no radiation protection problems occurred when the tests were carried out at said power levels. The search method described here where fission gases are gathered above the core, separated from the water in a small container and detected with an aerosolmonitor, showed to be very useful because the method is simple and sensitive. It is much more sensitive than e.g. the direct water activity measurement due to the concentration of fission products to the filter. 1-75 1 AIR FLOW Tim AEROSOL WATER MONITOR FLOW i \ r il RECORDER Figure 1. Experimental set-up for the detection of gaseous fission products. 1-76 OPERATING EXPERIENCE AND MAINTENANCE AT THE TRIGA MARK II LENA REACTOR. F. Cingoli, S. Altieri, F. Lana, G- Rosti, L. Alloni, S. Meloni Laboratorio Energia Nucleare Applicata, Universita' di Pavia, Pavia < Italy ). INTRODUCTION The last two years at the Trigs Mark II LENA plant were characterized by the running of the n-n oscillation NADIR experiment. Consequently reactor operation was positively affected and the running hours rised again above 1000 hours per year. The LENA team was also deeply involved in the procedures for the renewal of the reactor operation license. New require ments by the Nuclear Energy Licensing Authority, ENEA for Italy, most of which concerning radiation protection and envi ronmental impact, have been already fulfilled. In some cases the installation of new apparatus is underway. REACTOR OPERATION • The reactor operation period considered in this paper ranges from July 1986 to June 1988. Reactor operation time, MWD and burn-up data are shown in Figures 1 and 2. During this period 626 applications for reactor use were brought in. In Table 1 the applications are grouped according to the research projects and the distribution of reactor time among the groups is reported in Table 2. Table 1 : Reactor utilization in the time period July 1, 1986 - June 30, 1988 Research Reactor time Time percent projects ( hrs ) Activation analysis 4,269 61 .8 Radioisotope production 430 6.2 Forensic 171 2.5 NADIR experiment 1,627 23.6 NeutTonography 16 0.2 Neutron spectroscopy 374 5.4 Teaching and training 20 0.3 Total 6,907 100 1-77 Table 2 : Reactor users in the time period January 1, 1986 - June 30, 1988 Users Reactor time (hrs) Average 1986 1987 1988 Total 85 1982 - 1985 Jan. - June Pavia University 1 ,356 1 ,044 426 547 2,017 20. 7 Other Universities 89 228 210 33 471 4-8 CNR - CRAA 1 ,286 1 , 196 737 433 2,366 24.2 Euratom 987 872 670 590 2,132 21 .8 Court of Justice 103 196 223 15 434 4.4 NADIR 188 1 ,522 834 - 2,356 24- 1 Total 4,009 5,058 3, 100 1 ,618 9,776 100 % 10 uo tc 3 O I z o u a. o YEAR Fig. 1 - Reactor operation time from start-up to 1987. Oo O YEAR Fig. 2 - MwD and burn-up at the LENA plant in the time period 1965 - 1987. REACTOR COOLING SYSTEM Any water, poured by accident in the reactor room, is conveyed by a trench system to a water collection well. The water may be succesively discharged after Health Physics controls. The well is equipped with an overflow signal, in the case that too much water is being collected, and to avoid the possible flooding of the reactor room ; whenever the overflow signal is activated, the inlet of any water in the reactor building is hindered. In December 1987 a spill from one of the pumps of the primary cooling system occurred during reactor operation. The water was conveyed to the collection well, the overflow signal was triggered and the inlet of water in the reactor building stopped, including, of course, water for the reactor cooling. Soon after the reactor operator observed an unusual instabili ty of reactor power, characterized by large power oscillations on the linear recorder, and an increase of temperature of the water in the primary circuit. Reactor was immédiatly shutdown and the cause of the power instability soon identified. However, in order to avoid any break in water feeding to the reactor building during reactor operation, a water level gage was inserted in the collection well at a lower level with respect to the overflow device. The level gage is connected to an acoustic alarm in the control room. By now, at the LENA plant, if a water leakage occurs in the reactor room, the water collected in the well triggers first the water level gage without interrupting the water feed to the building; the reactor operator is alerted and the reactor team may quickly identify the cause and take care of the inconvenience. THE n - n OSCILLATION EXPERIMENT The n - n oscillation NADIR experiment, whose set up had been described in previous Conferences, was completed in July 1987. Data acquisition was stopped after about 2400 hours of neutron beam given to the target. The experiment set up is still mounted as well as the interblocks connecting the experiment and the reactor. No future use of the NADIR neutron beam is at the moment foreseen, whereas the detector's room, with its three-dimen sion array, may be potentially utilized in cosmic rays studies. ELECTRONIC EQUIPMENT Due to age and obsolescence, parts of the reactor electronic equipment must be substituted. In 1987 the "percent power" meter was removed from the reactor control console and replaced with a new one especially designed and manufactured by the LENA electronics group. The block diagram of the new circuitry is shown in Fig. 3. The input signal from the ioni zation chamber after amplification, enters the meter, whose scale is calibrated in " % power", and the trigger circuit. The latter, connected to a scram level selection device. 1-81 SIGNAL INPUT >- • 12V OUTPUT SCRAM LJ CHECK SCRAM LEVEL I SELECTOR 03 ,/pOWER METER A AMPLIFIER T TRIGGER RESET SCRAM HOLD ON MEMORY Fig. 3 - Flow diagram of new % Power Meter triggers very quickly the scram circuit by inducing a polarity change in the scram power circuit. The scram signal after a further amplification is fedback to the scram power circuit, thus keeping activated the scram condition, that in turn can only be resetted by opening the reset circuit. LICENSE RENEWAL The Italian Nuclear Energy Agency < ENEA ), acting also as licensing authority for nuclear plants, before renewing for a five years term the operation license of the Triga Mark II LENA reactor has required to implement the radiation protec tion instrumentation and to carry out radiation dose measure ments inside and outside the plant ( with the reactor shutdown and in operation ), as well as an air particulate monitoring campaign. New apparatus for radiation measurement, such as dose and contamination monitors, have to be stored outside the reactor building, ready to be used in any emergency situation, both internal or external to the plant. As previously cited, a campaign was carried out, in cooperation with the ENEA officers, to measure with very sensitive instrumentation (Reuter-Stokes ionization chambers), inside and outside the plant, the radiation doses, with the reactor shutdown and in operation. Instruments were able to detect radiation level change above natural background, every time that the reactor was operated. Radiation levels were found compatible with radiation limits for occupâtionally exposed and unexposed people. However, for a further safety purpose, it was decided that the general public should not be admitted in a 10 meters belt around the LENA building. A campaign was also carried out to look for radionuclides in the air particulate collected on filter paper in the exhaust chimney. Filters were submitted to gamma-ray spectro metry, but only radionuclides deriving from natural radio active decay chains were detected. The ENEA officers, howe ver, required the installation of a continuous monitoring of the exhaust air. A gamma- ray spectrometry system is being installed in a calibrated by pass derived along the exhaust air piping. The system's detector is set to monitor the 41-Ar gamma ray. An acoustic alarm, activated for radiation levels above preset values, will be available in the reactor control room. The system can be modified to detect also gamma rays from gaseous fission products, thus acting as a monitor for emergency conditions. At the LENA plant there are absolute filters on the exhaust of the radiochemical hoods and activated carbon filters in the emergency air ejection system. The ENEA offi cers required that the installation of these filters needs to be periodically checked as well as their filtration efficien cy. To this aim, the air exhaust piping from hoods and emer gency air ejection system has to be modified to allow the insertion of devices for the required evaluations^ A layout of the new air flow from hoods and emergency system is presented in Fig.4. Collection of some meteorological data at the LENA plant 1-83 Fan Fa - Absolute filters SM5 Nadir Fca— Activated carbon filters —tx) Building VV4 PF - Subsidiary filters Fan Fc - Paper filters tot-ß Hoods 41 Nal- Check of Ar VV3 i GM— ß -y counter m NaJ n e Emergency system y Control Fa Room SMi CA FA ÏSr, -01 tv Fca VV5 I floor iilll FaC PF mVVi ;sr, Plenum SM4 tab. £T~ VV2 I PF #— 00 Fc W - Fans ;sr GM SM .... Sr ... - Shutters PF _,n Lab. Rabbit II floor Beams Euracos :sr« A dm. Fig. 4 - Layout of the new air flow from hoods and emergency system. such as temperature, atmospheric pressure, direction and speed of wind, was so far carried out manually. The readings from a small meteorological station were periodically recorded by the reactor operator on a logbook. The station has now been inter faced with a small personal computer, data reading takes place automatically at prefixed intervals and data are recorded on floppy disks and printed out on a printer located outside the LENA building, so that meteorological parameters are available even if an emergency condition is started inside the reactor hall. Moreover, the availability of meteorological data on floppy disks allows their treatment for statistical purpose and trend analysis. 1-85 The Research Reactor TRIGA Heidelberg II institut für Radiologie und Pathophysiologie Deutsches KrebsforschungsZentrum Heidelberg W. Maier-Borst and 0. Krauss Introduction The research reactor TRIGA Heidelberg II in the Institut für Radiologie and Pathophysiologie of the Deutsches Krebsfor schungszentrum Heidelberg is of the type TRIGA Mark I (without irradiation tubes) with a thermal power of 250 kW. The reactor is in operation since the beginning of 1978. It is the second TRIGA reactor of the Deutsches Krebsforschungszen trum, since in the first stage of completion of the research center, a TRIGA MARK I (1966-1977) has been in operation. On the base of the working experience gathered during that time employing the TRIGA in biomedical research, especially the ir radiation units have been extended or newly developed. On the occasion of the TRIGA Users's Conference in Rome in 1986, several TRIGA users reported about difficulties they had encountered using the rotary irradiation system. It became ob vious to us that the alternatives to the original Lazy Susan are not commonly known. Therefore, in this speech the open ro tary system fed by a hydraulic rabbit system, which has proved successful in this form during the past ten years, shall be presented. Core Configuration The core of the TRIGA Heidelberg II is that of a Mark I to which a G-ring has been added (Fig.l). On the whole, the core has 121 positions, which are, for core-symmetrical reasons, arranged in a hexagon. Presently, the core is loaded with 99 fuel elements (about 3.5 kg 235U) , of which 56 elements are can ned by aluminum (former TRIGA HD I elements), and 43 elements are caned by stainless steel. For the operation of the reactor and for the shut-off, four absorber elements with fuel element followers have been integrated. Meanwhile, nine fuel elements with aluminum cans have been eliminated due to their excessive bending and are presently stored in a rack inside the water pool of the reactor. 1-87 Irradiation Devices The TRIGA HD II is the first TRIGA reactor furnished with a hydraulic rabbit system with an open rotary specimen rack and special in-core irradiation stations. As irradiation facili ties, seven in-core stations and 40 stations in the rotary sy stem are available. Five in-core stations and all of the ro tary stations are fed with a hydraulic rabbit system, and two of the in-core stations are fed with a pneumatic rabbit sy stem. All of the irradiation stations are constructed such that during the irradiation the rabbit is situated in the sym metry plane of the core, i.e., in the region of the highest neutron flux density (Fig.2). For the activation of inert gases, an aluminum in-core irra diation facility with a volume of 0.651 is available. This station has been designed for a pressure of 4 bar. In contrast to the normal TRIGA construction, the rotary sy stem runs in a pervading gap between core and graphite reflec tor, so that the pool water directly cools the samples. During the development of the rotary system special attention was gi ven to easy accessibility and serviceability. The rotary sy stem is supported by five bearing stands, which can easily been positioned or dismounted from the platform. The in-core stations of the hydraulic rabbit system can readily be lifted or displaced inside the core. Hydraulic Rabbit System The hydraulic rabbit system has six loading tubes, one for the rotary system, and the remaining five for the five in-core ir radiation stations. These six tubes lead from the core upward and inside a shield over the edge of the pool, from there (in side the biological shield) into an underground duct to the reactor engine room, where they end in a shielded box in the B laboratory (Fig.3). Along the entire length of the transport tubes, radiation protection is assured. The system has a pres sure and a suction pump for loading and unloading the diffe rent stations of the core and in the rotary system. Above the reactor pool and below the shielded box of the B-laboratory with the loading station, sensors have been installed which signal the passage of a rabbit. 1-88 The distance between the core and the loading station is about 30m; the travel time of the rabbits between the station and the reactor core is about 30 sees. For the hydraulic rabbit system, a special rabbit has been constructed. For reasons of radiation protection, the handling of the irradiated rabbits and the removal of the secondary irradiation rabbits is possi ble only behind shielded walls and with the aid of a manipula tor. Operation of the Reactor Production of isotopes and neutron activation analysis are the major applications of TRIGA. During the period from Feb.28, 1978 to Dec.31, 1987 TRIGA has produced 3.6 MW. During this time there were over 4,500 different experiments with nearly 22,000 individual samples being irradiated. The reactor is in operation on about 180 days per year. 30 days are needed for government mandated inspections. About 15 more days are re quired for maintenance. The daily irradiation time has in creased from 4.7 h/d in 1978 to 10.7 h/d in 1987. Our TRIGA is meantime the only research reactor in southwest Germany for activation analysis and production of shortlived isotopes. Besides using the reactor extensively ourselves, we have many users from universities, Max-Plancfr-Institutes and industry who irradiate samples for physical, biological and material investigations. Radiation Protection and Radioactive Wastes during Routine Operation In compliance with the conditions for the permit of the rese arch reactor TRIGA HD II, and to establish proof of protection against ionizing radiation, a number of radiation protection measurements are taken periodically. The staff is monitored with personal dosemeters. For the control of the research re actor, the activity in the pool water is measured weekly, and both the gamma dose rate just above the pool surface and the activity in the exhaust air are monitored continuously. The activity inside the ion exchanger resin of the purification circuit is measured before it is treated as radioactive waste. In the following, the results of our measurements will be pre sented. 1-89 The personal doses are small. They are below 10 man-mSv per 7 staff members and per year (operation of the reactor: 3 em ployees; engineering workshop: 4 employees). A repair of the ball race of the rotary system which was due in 1984 resulted in an increase of the personal dose to 17 man-mSv. The mean concentration of radioactive nuclides in the pool wa ter at the end of an operating week is shown in Fig.4 (power: 250 kW, operation: maximum of 10 hours per day on 5 days per week). Tritium is produced by the (n,1)-reaction of the deute rium, 24Na by the (n,or) -reaction of the aluminum, and 41Ar by the (n,?) -reaction of the argon solved in the pool water. The remaining radioactive nuclides having significantly smaller activities are produced by activation of the impurities in the aluminum and in the metal components of the stainless steel of the fuel cans, screws, and ball bearings. In the table, the radioactive nuclide 14C does not show up. It is produced by the (n,p)-reaction of 14N and by the (n,a)-reaction of 170 and 14 is found in small concentrations as C02 in the pool water. The activity was measured in the exhaust air from the pool wa ter surface (P. Gesewsky, Bundesgesundheitsamt, Neuherberg, 1985). The results of these measurements will be explained in the chapter treating the air activities. The activity concentrations of 24Na and 41Ar in the pool water mainly produce the dose rate above the pool surface with a ma ximum value of 40 juSv/h after an operating time of about 10 hours at 250 kW. The remaining activities do not contribute significantly. In contrast, the 16N activity in the pool water (160 (n,p) 16N , not listed in the table) and the direct irra diation from the core do contribute to the dose rate; their contribution to the above dose rate is about 5 /xSv/h. The air space above the pool water surface is separated from the reactor room by aluminum gratings with pane of plexiglas and is purified at a rate of flow of 1,000 m3/h. The gas acti vity in this exhaust air mainly contains 41Ar as well as tra ces of Tritium and 14C. The 41Ar concentration is monitored con tinuously with a gas. flow-through counter tube in a by-pass system and is about 23 MBq/h-250 kW. Additional measurements by drawing air samples in a pressure bottle and gamma spectroscopy measurements yielded a value of (10 kBq/m3•1,000 m3/h =) 10 MBq/h-250 kW. A conservative overall balance of the exhaust air activity of 41Ar for 1,500 operating hours per year 1-90 at 250 kW yields an emitted 41Ar activity of (23 MBq/h-1,500 h =) 34.5 GBq per year. The evaporation of a maximum of 10 m3 of the pool water results in an emission of activity of Tritium (see Fig.4) of 40 MBq per year. The UC activity in the ex haust air above the pool water surface has been measured to be 1.82 Bq/m3 and therefore results in an emission of activity of WC of (1.82 Bq/m3 • 1,000 m3/h 1,500 h =) 2.7 MBq per year. Monitoring the aerosols in the air above the pool, only traces of 24Na, ^Co, and 109Cd are found (i.e. some hundred Bq per week and aerosol filter) . Since the exhaust air is being lead over absolute filters, no aerosol activities can escape. These re marks show that only 41Ar has to be taken into account for dose calculations in the reactor environment. Two ion exchangers, each with 240 kg of resin, are integrated in the purification cycle. They are operated alternatingly in order to be able to carry out the regeneration for the washing of carbon dioxide. After decay times of more than a year, two fillings have been treated as radioactive waste up to date. The activities in the barrel are ^Co with a maximum of 15 MBq, Mn with a maximum of about 2 MBq, and 65Zn and 58Co with less than 1 MBq. The additional small Caesium activities in the se cond barrel are traces of the Chernobyl accident. 1-91 TRIGA HD II O Watai^-Tube-Position (A1J)8,D12,E4,E22) O Aü-Tnbe-Poaition (G4, F22) O Control-Rod (C1,C4,C7,C10) Fig. 1 0 Gas Radlaiion Position (F11) 1-92 TRIGA HD II Neutron Fluxes 1E12 n/sqcm/sec E22 E4 Rotary-SysternOutside Radiation Positions Ml Neutron Flux Fig. 2 W7T s;;;;;;;/;;/;;;///;;////////////////////// B-Laboratory Shielded Box Suction Pump Pressure Pump Loading Station •o Transport Tubes with Monitoring System Monitoring System Rotary Drive, Position Transmitter Rotary System 5 In-Core Irradiation Positions Fig.3 1-93 TRIGA HD I! Mean Concentration of Radionuclides in the Pool-Water 3H 56Mn 99mTc 5lCr 54Mn 6OC0 Radionuclides kBq/m3 Fig. 4 1-94 ABSTRACT OPERATION AND MAINTENANCE EXPERIENCES AT THE C.R.E. CASACCIA TRIGA REACTOR. A. Festinesi. Servizio Esercizio Impianfco TRIGA - ENEA, Dipartimento TIB. C.R.E. Casaccia - Strada Anguillarese Km 1+300 00100 ROMA (ITALY). The memoir explains TRIGA RC-1 plant activities from last European TRIGA Users' Conference till today. In particular, measures following reactor exercise license renewing (march 1987) are described. Finally, difficulties and measures about shielding tank's water funguses and spores contamination, are explained. 1-95 OPERATION AND MAINTENANCE EXPERIENCES AT THE C.R.E. CASACCIA TRIGA REACTOR. A. Festinesi. Servizio Esercizio Impianto TRIGA - ENEA, Dipartimento TIS. C.R.E Casaccia - Strada Angui1larese Km 1+300 OOIOO ROMA INTRODUCTION During the last four years the most important activities of TRIGA RC-1 operation group have been the revision and the elaboration of technical documents to renew plant exercise license. On March 87, ENEA has been authorized to reactor exercise, by INDUSTRY MINISTRY (MICA), for five years more. Till now, reactor operation group has realized or begun part of the measures agreed and/or requested with/from Italian Regulatory Body (ENEA/DISP) for the new licensing, as (essentially): -a) renewal of reactor instrumentation referred, in particular, to power control channels, ß - Y and aeriform radiations monitoring (in the reactor room and in the reactor building chimney outlet); -b) installation of an authomatic fire protection system in the reactor control room; 1-96 y -c) physical separation of reactor room inlet-outlet corridors and aß -Y portal monitor installation in the reactor room outlet corridor; -d) documents elaboration for the Quality Assurance reactor management. Originally, as we have referred at the last European TRIGA Users' Conference, instrumentation renewing actions in a), were included in the ENEA design intended to realize a computerized TRIGA control consolle of advanced conception, with the principal aim to develope an ergonomie control room that could be a reference point for the planning of itaiian power reactor control rooms. Consequently to changes in the itaiian energetic politics, after Chernobyl, that results of papular referendum about nuclear plant employ as energy sources reinforced (November 87), the computerized control consolle project has undergone a delay, Contemporaneusly to the technic-administrative actions for plant exercise license renewing, TRIGA RC-1 reactor has been operated <5 days in a week, 6-7 hours/day), essentially, to continue in pile irradiation activities finalized to neutron activation analysis and radioisotopes production for ENEA and extra-ENEA users. 1-97 REACTOR OPERATION. In the January 86-August 88 period, TRIGA plant has been operated at 1 MW for a total time of about 19^0 hours, as table 1 summarizes. Figure 1 shows temporal distribution of ractor operation and, in particular, plant arrests whose reasons ^re following listed: February- : heat exchangers removal and -March 1988 substitution because Silicium incrustation on the secondary site. July-August 1986, August 1987 : summer holidays of reactor operators. September 1986 : dismounting of CNR spectroscopy apparatus because experience cessation. July, September, October and November 1987 : water purification of the bulk shielding tank because funguses contamination (see after); reactivity insertion because core burn-up compensation; apparatus installation because neutrons and ?v beam (of prefixed energy) 1-98 realization (as referred in the last European TRIGA Users' Conference). ENEA, CNR and Rome University searches, essentially, have utilized TRISA reactor as table H shows. Finally, table 3 summarizes the irradiation facilities utilization. ENEA and Rome University searches have used central thimble and Lazy Susan facilities also for \f irradiations, during reactor shut-down , periods (nights, holidays, Saturday and Sunday time) for V rays damage studies on plastic materials and epoxy resins of italian industrial production particulary. To estimate above y irradiation dose, figure 2 is purposely inserted. Reactor room accesses restructuration is responsible of the Rabbit limited utilization. In fact, such restructuration has involved radiochemical laboratory (see fig. 3) in which Rabbit external terminal is installed. Further, also Rabbit external terminal has been modified Co automatize irradiated samples removal and to increase its lead shielding; so, only recently, the pneumatic transfer system has taken again to work. OPERATIVE PROBLEMS AND TROUBLES From the last European TRIGA Users' Conference (October 1986) till now, particular troubles and problems have not 1-99 occurred, but it may be interesting to s.ignalize a funguses contamination of the shielding tank water and the actions we have undertaken to solve the problem. Since autumn 1986 a thin layer of gelatinous and colourless mould was observed on the epoxy covered walls of bulking shielding tank. Initially, a particular attention was not attribuited to such substance but, in few months, its layer became remarkable and the substance fell to pieces in the shielding tank's water which grew of a turbid lightly colour. Water and substance microbiological analysis disclosed the presence of fungus colonies classified as ASPERGILLUS FUMIGATUS. Probably, some wandering spore achieved the shielding tank's water from reactor carriage door during its opening in the reactor shut-down periods. To purify tank's water two products of commercial type were found, by DIVERSEY s.p.a. both products manufactured: MERAK HE DIVOSAN FORTE First product (essentially "chlorine constituted) was discarded immediately because its green colour dyed intensely the water and therefore DIVOSAN FORTE was preferred. Table 4 summarizes the principal characteristics about DIVOSAN FORTE. 1-100 Purification work (in the shut-down reactor conditions) was realized as -following: - DIVOSAN FORTE 15 liters insertion (equivalent to a concentration of 0.07'/. about) inside shielding tank's water ; - shielding tank's pump on after the ion exchange resin container bypass, to blend water and to clean shielding tanks piping. Pump was in operation for about 3 hours/day for 5 days, nights excluded. Consequently this treatment, funguses in the tank's water assumed an intense black colouring, water becoming fully not transparent; - microbiological analysis of shielding tank's water; laboratory results disclosed fungus colonies and spores death, by their black colouring preannunced; - reinsertion of the ion exchange resin container and shielding tank's pump on. Although, for many days, ion exchange resin and additional filtering system were inserted in the pump circuit (till 1 - tank's water removal and its transfer to C.R.E. CASACCIA waste tanks. RC-1 Operators were let below shielding tank for a better clearing of gelatinous remainder on the tank's bottom. In corrispondence of the thermalizinq column door 1-101 a v dose of about 30 mR/h was measured during the cleaning operation and, obviously, it was not a radiological problem. Tank's bottom and its lateral surfaces were cleaned furtherly with DIVOSAN FORTE imbibed clothes (not dilui ted); - reintegration of démineraiized water (~ 22 m ) in the shielding tank; - ion exchange resin substitution; - clearing of conductivity probes at the inlet and outlet of ion exchange resin container for anomalous informations due to gelatinuos material presence inside probes. In consequence of this operation, the internai parts of tank's p.v.c. piping were controlled and they were found strongly dirty for the presence of gelatinous residuals as inactive funguses and spores at last. For further precautions, on January 88, the PVC piping of tank's plant were substituted. Till now, ASPERQISLUS funguses or spores are not more detected in the tank's water. However, for prudential reasons, at the beginning of August, and since rea.ctor was in the shut-down conditions, 10 liters of DIVOSAN FORTE (0.04-V. about) have been newly inserted in the shielding tank's water. Water pump has been maintained in operation during August days (nights excluded) after ion exchange resin container bypass. 1-102 In September, before to start at 1 MW, DIVOSAN FORTE residuals have been removed from water by ion exchange resins reinsertion. PRESENT AND FUTURE ACTIVITIES As we have referred more in detail in the last European TRIGA Users' Conference, RC-1 plant is employed essentially in the following activities: a) neutron irradiation for tracers production and activation analysis, requested by ENEA's laboratories, research institutes and national industries (as we have referred also in the REACTOR OPERATION paragraph); b) neutron spectroscopy and diffraction, in magnetic systems liquid and amorphous substances c) realization of a new radiochemical laboratory for the production of medical diagnostic radioisotopes; d) realization of a Tritium handling laboratory, to support italian programmes on nuclear fusion technologies. Activities as per c) and d) are still in the planning and authorizative phases in consequence of financing delays after Chernobyl. According with Italian Regulatory Body (ENEA/DISP), a technical-typographic revision of RC-1 Safety Report is realizing to keep in account the numerous documents 1-103 elaborated for new licensing and any modification and/or improvement from first criticality (1967) today. In this general revision is inserted a verification and control programme about aeriform waste production and discharge to reactor building chimney. In particular a théorie experimental verification of Argon 41 production is just in act (inside reactor tank's water, in the reactor room and in the external part of reactor building). A. first critical revision about Argon 41 production theory has been completed and it will be published after experimental data analysis, which collection is just in act. 1-104 TABLE 1 DATA ON REACTOR OPERATION FROM 1986 TO AUGUST 1988 \ .EXERCISE. PERIOD OPERATION HOURS AT 1 MW MWD 1986 643 •26,8 1987 625 26 JANUARY 88 AUGUST 88 670 27,9 1-105 TABLE! 2 REACTOR UTILIZATION FROM 1986 TO AGUST 1988 USER EXPERIMENTS 1MW REACTOR OPERATION TIME FACILITY activation analysis for: environmental pollutior studies, forensic purposes, nuclear materials qualification, agricultural improvements, geolo Lazy Susan gical informations, constructive works qualifi 1000 Central thimble cation and restoration. Rabbit ENEA radioisotopes production (tritium essentially) to support italian programs on nuclear fusion technology. radiation damage studies on plastic materials andp.v.c, essentially. (*) spectroscopy and diff rattometry for solid state CNR physics studies 1900 2radial beams ROME activation analysis for geological and 700 Lazy Susan UNIVERSITY environmental studies (*) Central thimble activation analysis for: ambiental pollution 200 Lazy Susan OTHERS studies, materials qualification, geological informations (*) ENEA and ROME UNIVERSITY have used "Central thimble" and "Lazy Susan" also for X irradiations during reactor shut-down periods for / rays damage studies on plastic materials and epoxy resins of italian production particularly. TABLE 3 IRADIATION FACILITIES UTILIZATION FROM OCTOBER 1986 TO AGUST 1988 IRRADIATIONS WITH REACTOR FACILITY ON DOWN (1) NUMBER TIME (hrs) NUMBER | TIME (hrs) CENTRAL THIMBLE 15 780 21 | 780 RABBIT 153 1,5 LAZY SUSAN 1182 19.180 78 j 1.140 ( 1 ) only y irradiation 1-107 TABLE 4 DIVOSAN FORTE ACID DISINFECTANT PHYSICAL CHEMICAL CHARACTERISTICS Composition (%): peracetic acid 15,0 j oxygenated water 23,0 ] acetic acid 16,0 ! stabilizer 1,0 ! demineralized water 45,0 ] Aspect: colourless limpid liquid Properties.: disinfectant for plants treatment in alimentary and bottling industries. DIVOSAN FORTE solutions may be used on stainless steel, tinned stripe, galvanized iron, teflon, polystyrene,polyethylene, syntetic recoverings, enamel and rubber. Density: 1,12 gr/cm at 15°C USE DOSES AND MODALITIES Temperature (C) Contact time Concentration {%) i less than 8 15' 0.1 ~ 0.2 | 8-=- 22 5* 4- 10' 0.1 4- 0.2 j 8 •?• 22 15' 0.07 4-0,015 ! 22 -r- 40 5' 4- 15* 0.07 -f- 0.015 | 70 -5- 90 5' 0.07 { (l)ambiental temperature 6' 4- 12 hours 0.05 j (1) This possibility must be preferred in the case the plant may be kept full of disinfectant solution. 1-108 TRIGARC-1 Reactor operation 120. f U n 100. n <%?& i 80. n g 60 40. annual total : 643.05 hours X annual total : 625.38 hours total : 670 hours 20. gïé y r /m?/M?/m'/s*?7t ± ^rm -LL_1L ï£ mm G F M A M J J A S 0 H D G F M A M J J S 0 N D 6 F M A M J aJ A S 0 N D year 1986 year 1987 year 1988 FIS. \ GAMMA DECAY AFTER BEACTÖR SCRAM 1 10 TIME AFTER SCRAM (lu) - CBNTPAl rwmtB \ - - I • ' 1 • ili —, i_^ , 10 TIME AFTER SCRAM (hr) F16.Z unn REACTOR ROOM ADMITTANCE LAYOUT Peactor room Reacto r room At inlet corridor Gt outlet corridor Ci hot dressing room Di decont«mi nat ion shower-bath Ei safety door (normally closed) Gt cold dressing room Ht radiochemical laboratory I : emergency wardrobe Lt health.physics site N: security office Ot electronic laboratory P?Ot portal monitor Ri rabbit terminal Ii telvcamera Si health phytic« monitoring instrument 3t wardrobe *• t contamina tea dresses conta Iner St health physics alarm bell 6t clothes peg 7t overa 1 Is hanger Boforo modification After modification fJGJ OPERATION EXPERIENCE WITH THE TRIGA REACTOR VIENNA H.Bock, J.Hammer, G.Zugarek Atominstitut Vienna, Austria ABSTRACT The TRIGA reactor Vienna operated during the last two years without any major unscheduled outage. The average power production per year is 230. MWh. Presently the reactor operates with 72 fuel elements (mixed core), 54 of them are Al-clad and from the first core loading in 1962. Totally 10 fuel elements had to be removed due to fission product release or excessive elongation. In view of future fuel element failures a comibined wet/dry sipping test cell was constructed and tested which allows the detailed investigation of suspected failed fuel elements. In 1986 a new purification loop has been installed being completely independent of the main coolant loop and operating round the clock. The reactor was submitted to a new licensing procedure resulting in an increase of periodic inspections of all major reactor components. Visual tank inspections showed that frequent cleaning of the pool is necessary to remove various particles from the tank bottom and tank installation. This particles originating from work above the pool may lead to serious corrosion spots if not removed periodically by a strong underwater suction pump. A near term plan for the reactor in the renewal of the reactor instrumentation which is 20 years old. Although the overall experience was excellent problems with spare parts and requirements for additional safety features are pending. Quotations were already obtained from 3 companies but no decision has been made. 1. INTRODUCTION The TRIGA reactor Vienna operated during the last two years without any major undesired shut down. As the reactor is now more than 26 years of age and still 54 of the original Al-clad fuel elements are used in core, in the near future new fuel elements will have to be purchased. Further the present reactor instrumentation has been installed in 1968 and will have to be replaced soon. 2. FUEL ELEMENT SITUATION The TRIGA reactor became originally critical at 100 kw with 64 Al-clad fuel elements. During the past 26 years the total number of fuel elements available at the institute increased to 90, 26 of them with SST cladding, 9 out of them are FLIP elements. From these 90 fuel elements 6 fuel elements show either excessive elongation or bowing and other 6 fuel elements had to be removed due to fission product release. These fuel elements are stored underwater in the fuel storage pits and are checked regularly. Until December 31, 1987, the total thermal power production was 6830 MWh. Using the relation of 1.25 g U-235 for 1 MWd this means that 1-113 356 g of U-235 has been consumed during the life-time of this reactor which is equivalent to the uranium content of about 9 standard TRIGA fuel elements. In view of exspected fuel element failures a combined wet/dry sipping cell for TRIGA fuel elements was constructed [1]. The sipping cell is designed to allow either wet sipping with a closed water loop or dry sipping once through with Nitrogen gas. As an additional feature the equipment allows to heat the suspected fuel element with two electrical heating coils up to 300C°. Using two separate coils it is possible to heat either the upper part, the lower part or the total fuel element. The temperature distribution is controlled by four Ni-CrNi thermocouples. A general view of the sipping cell is shown in Fig.l. It consists of two stainless steel tubes, the inner tube contains the fuel element and supports on its outside the heating coils and thermocouples. Through the lower end plug a pipe is connected which contains the purging medium (water or gas). The outer cylinder acts as a protective shield and support. The volume between inner and outer cylinder can be evacuated to improve temperature isolation. For shielding purposes two lead plugs are provided one fixed installed around the top of the inner cylinder for lateral radiation shielding and one movable lead plug which is loaded on top of the fuel element with the fuel handling tool for vertical radiation shielding. The fission product detection system consists of a high purity germanium detector with a preamplifier and multichannel analyzer. In wet sipping tests the water circuit is wrapped around the active part of the detector, while in dry sipping tests Nitrogen gas is passed through the cell and a glass fiber filter, used in front of the detector which measures the gamma spectrum. The count rate versus energy is recorded and fission products are identified from standard tables. The wet sipping test can be performed while the cell is suspended inside the reactor pool on the wall (Fig.2). To load the capsule, the cell is lowered to the bottom of the reactor pool and the fuel element is loaded into the cell under water. Then the cell is lifted and suspended on the pool edge. As the cell is now flooded with pool water, this water has to be removed by a small pump and replaced by fresh destilled water before starting the measurements. Dry sipping test is more complicated as the water has to be removed completely from the cell. It can principally be performed in the reactor pool, but it takes an extended period to remove all water from the capsule, which is introduced during fuel element loading. Another possiblilty is to position the cell somewhere else in the reactor hall in a dry shielded place (i.e. spent fuel storage pits) and transfer the respective fuel element to the cell by the fuel transfer cask. In this case the sipping cell is kept dry all the times and measurement can be started immediately after loading (Fig.3). As purging gas Nitrogen is recommended in once-through-then-out loop. The gas together with any particulates passes through a glass fiber filter in front of the high purity germanium detector, which measures the gamma spectrum. By stepwise increase of the temperature, the temperature dependent release of any fission products can be monitored. 1-114 Care has to be taken not to contaminate the detector area, therefore frequent cleaning with alcohol of the filter housing is recommended. 3. REACTOR TANK As reported at the last TRIGA Users Conference in Rome in 1986 the tank has been checked optically in 1985. These inspections together with a general cleaning process has been repeated in 1988. unfortunately several corrosion spots with a maximum size of 8 mm and a depth up to 2 mm were found on top of the thermalizing column. The spots have been checked optically and the dimensions of these spots were taken using a replica method. As the material thickness at these areas is 12.5 mm there is no immediate danger, but the spots will have to be observed closely. Out from those experience with cleaning several systems are now used which are: - a strong underwater pump with an integrated fine filter - a water jet nozzle with a pressure up to 160 bars to stir up deposits - rotating brushes to clean surfaces - small but strong underwater spot slights for observations in tank - replica material which allows to take exact prints of corrosion spots within a period of 5 minutes. The experience with these system is described in more detail in [2]. 4. REACTOR INSTRUMENTATION The present instrumentation was installed in 1968 and is a hard-wired relay instrumentation. Although its performance was excellent there is now a lack of spare parts and ageing effects can be observed. Three quotations have been received, but no decision has been made as the funds must be supplied by the competent ministry [3]. 5. PROBABILISTIC SAFETY ANALYSIS During the last two years a detailed probabilistic safety analysis for the TRIGA reactor Vienna was carried out which will be presented in detail in another paper at this conference [4]. To obtain input data for the event trees the data collection of component failures was continued and information exchange with other TRIGA facilities took place. 6. REINSPECTION PLANS A detailed reinspection plan was developed [5] which list all components and prescribes the procedure of inspections. The inspection intervals range from weekly to monthly, semi annual and annual inspection. Certain inspections are carried out by specialized companies while other inspections are carried out during the presence of an appointed expert. Those reinspections consume quite a bit of time (i.e. a few days per month), but the benefit for the facility justifies this effort. 1-115 7. SUMMARY The TRIGA reactor Vienna is operating successfully with only minor problems due to ageing of certain components. Frequenct reinspection and careful maintenance is very important to assure safe and efficient operation in the coming years. REFERENCES Hl H.Böck, H.Gallhammer, J.Hammer, M.Israr: "A combined wet/dry sipping cell for investigating failed TRIGA fuel elements", AIAU 87310, August 1987 [23 H.Böck, J.Hammer, K.Varga: "Optical inspections of research reactor tanks and tank components", AIAU 88306, September 1988 C3] J.Hammer, G.Zugarek, H.Böck: "Concepts for the renewal of the instrumentation of the TRIGA reactor Vienna", 10th European TRIGA Users Conference, Vienna, September 14-16, 1988 [4] H.Böck, C.Kirchsteiger: "Probabilistic safety analysis for the TRIGA reactor Vienna», AIAU 88305, July 1988 [5] H.Böck, J.Hammer, G.Zugarek: "Reinspection plan for the TRIGA Mark II reactor Vienna, AIAU 88303, May 1988 1-116 expan heating sion end plug -ring lower coils bellow recircu flange gasket tube lation tube upperflange lead for media shield M Nam* fl*o*w o».»t &.TOMINSTITUT G*pmn »9"i VIENNA/AUSTRIA Norma Maftst COMBINED WET/DRY SIPPING 1.2 Fig. 1 CELL FOR TRIGA FUEL ELEMENT Ersatr für Erf tri dLreh sipping unit suspended on flow direction of the crane the sipping medium reactor ai . A platform "s'il pool / water / level / ', / / / / / / / / / / 8:: lead plug / / / / ', / / fuel / / element / / / / o / / / / / / / / jj 1 y_ / / • B i a / / / / / / / / / / / ', / / / / / / / / / / 140 / ' / / A ul reactor top reactor top \\\v^vqk ? grid plate grid plate thermal thermal column column liner liner CELL LOADING Fig.2: Schematic arrangement for CELL IN , t~n~é wet sipping test OPERATION crane glassfiber fuel handling filter tool high purity gas flow Germanium detector fuel , direction transferer cask TRI G A fuel dry instruments for element nitrogen temperature measurement temperature regulation reactor building" floor a lead plug TRIGA fuel element ¥sA sipping cell fuel storage m pit ( 3 m x o.24 m ) Fig.3: Schematic arrangement of the sipping cell in dry operation mode ! \ GHANA'S NUCLEAR PROGRAMME ALBERT K. AHAFIA NATIONAL NUCLEAR RESEARCH INSTITUTE GHANA ATOMIC ENERGY COMMISSION P.O. BOX 80 LEGON GHANA ABSTRACT The Paper gives the purpose of Ghana's Nuclear Programme and describes some specific research activities and peaceful applications of atomic energy in agriculture, medicine and industry. A discussion of some of the problem facing the programme concludes the Paper. 1. INTRODUCTION; The Ghana Atomic Energy Commission was established in 1963 by an Act of Parliament [Act 204 of 1963]. The Act vested in the Commission the responsibility in Ghana for all matters relating the peaceful uses of atomic energy. The PNDC Law 37 of 1982 gave substance to the original Act 204 of 1963. NATIONAL NUCLEAR RESEARCH INSTITUTE (NNRI) In order to discharge its responsibility to the nation as far as peaceful uses of atomic energy is concerned, the GAEC established the National Nuclear Research Institute in 1964. The Institute is situated at a sight 24 kilometers to the North of Accra and about 10 kilometers from the University of Ghana. The staff strength of the Institute is about 400, made up of highly qualified scientists, engineers, technicians, administrators and supporting staff. RESEARCH ACTIVITIES At the moment, the Institute has three academic departments, namely Departments of Physics, Chemistry and Biology, Food and Agriculture. All the three departments collectively and individually are conducting research into peaceful application of nuclear science and technology in Agriculture, Medicine and Industry. The results achieved so far are encouraging. It is not possible nor desirable to enumerate this morning all the research findings the Institute has achieved. A few results in each area of research should be sufficient for to-day. 1-121 In agriculture, the Institute has completed the research phase of food preservation by gamma irradiation. Pilot stage trials will begin as soon as the Gamma Irradiation Facility (GIF) currently under construction is commissioned probably by March 1989. Thereafter the nation would see the dramatic reduction in post-harvest losses that gamma irradiation can achieve in crops such as cocoa and maize. Additionally, the Institute has mastered the technique of mass rearing of tsetseflies, a first step to the eradication of tsetse flies in the cattle rearing areas selected in Northern Ghana. Besides improving animal production in Northern Ghana, the reverine tsetse fly project would also reduce sleeping sickness in the area. Another important research area in which the Institute has obtained good results is Tissue Culture. The tissue culture technique has the advantage of not only selecting high yielding species but also disease-free ones. Our scientists are now in a position to produce in the test-tube 650,000 yam plantlets from a single carefully selected tuber. The Green House now under construction would, when completed, enable the Institute to transfer the plantlets into the farm for assessment and eventual supply of such plantlets to farmers. Other crops being investigated include pineapple, beans, kernel palm. The Institute is confident that the Tissue Culture Project would make the desired impact on food and cash crop production in Ghana in the near future. The National Nuclear Research Institute is not oblivious of the importance of cocoa to Ghana's overall development. In collaboration with the Cocoa Research Institute at Tafo, our Institute is conducting genetic mutation breeding research into cocoa. The road to success may be long but the NNRI shall never tire until the distination is reached. HEALTH The NNRI is contributing its fair quota to the health needs of the country. It provides diagnostic services to the Korle Bu Teaching Hospital in the area of scanning the heart, liver, lungs and kindneys, using radioactive techniques to ascertain their state of proper functioning or malfunctioning. Every month, the Institute, using TLD badges, monitors 109 X-ray operators in Accra and Kumasi to ensure that they are not exposed to unacceptable levels of radiation. Commercial organisations using nuclear radiations in their operation receive regular advice and inspection from the Institute. The Institute will soon rehabilitate its monitoring stations in all the regions so as to resume the monitoring of radioactivity in food, air, rivers and grass in the regions. 1-122 Nuclear techniques are finding increasing demand in industry. The Institute has mastered these techniques [NAA, XRF] and is applying them in mineral exploration and uranium prospection. A programme for non-destructing testing to be introduced into relevant industries is under way. MANPOWER TRAINING One of the most important and regular features of the Institute's activities is manpower training. Over the years, the Institute has organised several training courses in nuclear techniques, such as NAA, XRF, BNF, computing and electronics not only for its own staff but also for participants drawn from a wide spectrum of the country's scientific and commercial organisations. These are what we call national training courses. The Ghana Atomic Energy Commission often hosts regional training courses for participants from Africa and the Middle East. The FAO/IAEA Regional Training Course on the Use of Isotopes and Radiation Techniques in Studies of Biological Nitrogen Fixation and Soil/Plant Nutrition, Accra, Ghana, 22 August - 16 September 1988 is one of such courses. There is good collaboration between the Institute and the Universities in the training of undergraduate and postgraduate science students. Some final year students from the Universities come to our laboratories to conduct research for their project dissertations. Postgraduate students researching Into nuclear-oriented fields also conduct their experiments in our Institute. Both sides gain from the collaboration. Our scientists register for higher degrees, M.Sc and Ph.D, in' our Universities. INTERNATIONAL ACTIVITIES Ghana became a member of the IAEA in September 1960, the third country from Black Africa to be admitted to the membership of the Agency. Since then, Ghana has been elected four (4) times onto the Agency's Board of Governors and is confidently seeking re-election this year. Ghana regularly attends the Annual Regular Sessions of the General Conference of the IAEA and contributes her fair share to the debates. The Commission is aware of the existing deficiency in bilateral co-operation and has therefore initiated moves for approaches to be made to some friendly Member States of the IAEA. Before long, we hope to see such co-operations flurishing. PROBLEMS The GAEC has had more problems than most scientific institutions in the country. The one that stands out is the perpetual absence of the nuclear reactor 1-123 expected to have been commissioned in 1966. The absence of the reactor has continued to dampen the spirit and lower the morale of the scientists trained specifically to operate and utilize the facility. As a result of the lack of job satisfaction, over thirty nuclear scientists trained abroad have either failed to return or resigned. The very patriotic ones now at post need only one incentive, namely the provision of the nuclear reactor. The other mundane problems include transport difficulties, lack of foreign exchange allocations to buy scientific equipment, consumables and scientific journals. For more than a decade, the NNRI at Kwabenya has had no telephone link with the rest of the country. Every conceivable message meant for Accra has had to be delivered by hand. We must concede that Government has solved some of our problems during the last two years. This will not, however, prevent us from persistently reminding Government of our problems. It is only natural. There is, however, a ray of hope in the tunnel. From 1986 up to date, the Institute has employed twenty (20) scientists as research officers. Training both at home and abroad is on the increase. Local inputs are also forthcoming. We now have telephone and telex facilities. Thanks to the IAEA, the provision of equipment and consumables is reasonable. This Report has decidedly been made brief because other documents such as the Brochure and the Status Report contain detailed accounts of the activities of the GAEC. Copies of these documents can be made available for your information. APPEAL On behalf of the Ghana Atomic Energy Commission, I appeal to all the organisations represented here and those who may read this Report to help the NNRI in any way they could: Fellowships, research grants, exchange visits and opportunities for Ghana's young nuclear scientists to work in your laboratories. Donations of nuclear equipments, journals and books shall be appreciated. Thank you. 1-124 SESSION II New Developments & Improvements of TRIGA Components & Systems, Including Instrumentation CONCEPTS FOR THE RENEWAL OF THE TRIGA REACTOR INSTRUMENTATION VIENNA J.Hammer, G.Zugarek, H.Böck Atominstitut, Vienna, Austria 1. INTRODUCTION The present reactor instrumentation was installed in 1968 and was at that time a modern, fully transitorized system with internal failure announcement. The system worked successfully during the last 20 years without any major problems. However in the last years due to ageing of components an increased failure rate was observed. In addition the reactor was submitted to a new licensing procedure which resulted in an expert's requirement of a new instrumentation (see also /l/). This new instrumentation should make a maximum use of existing equipment (such as boron ionization chambers, fission chambers and control rod drives) and of existing cable pathways. Further it should include today's state of the art of research reactor instrumentation taking the specific TRIGA characteristics into account. An additional requirement is the reactor use for students training- and education courses where various signals from the reactor are processed, but this must not have any feedback on the reactor itself. All signals utilized for students purposes must be extracted through isolation amplifiers. Finally all important reactor data should be documented and processed by a.computer for storage and evaluation. A replacement of the reactor instrumentation should also include a new design of the reactor control room. The control desk should include all important displays and necessary keypads (actuators) for reactor operation. For auxilary systems such as cooling and ventilation systems or signal processing a 5 m wall panel may be used which should be accessible from behind. 2. REACTOR PROTECTION SYSTEM Due to the inherent safety features of the U-Zr-H fuel a redundant and diverse reactor protection system (2 out of 3 systems) as used in high power research reactors or nuclear power plants is not necessary. A conservative relay type system to interrupt the magnet current of the control rods proved to be very reliable during the last decades. One additional suggestion for improvement in case of relay failure would be to trigger the reactor scram additionally by a programmable control unit as a back-up protection system. This program unit could be also very useful to document the detection and consecutive evaluation of reactor scrams. For the magnet current surveillance an AC control is proposed, which is able to detect the release or contact of the magnets in a diverse way (Fig. 1).- 2-1 24V control rod ° coupled rod magnet W Fig. 1: Proposal for AC control of rod magnet. The events which should lead to an immediate reactor scram are listed in table 1 including several spare inputs for later scram requirements. The individual steps during a reactor scram are listed in table 2. Only if all scram conditions have been cleared, a start-up is again possible. The start-up interlock is not only activated in case of scram conditions but also if the console key is in OFF position or in case of low source signal. All events leading to an alarm signal but not to a scram are listed in table 3. Alarm signals should be accompanied by an optical (blinking) and an acoustical signal. Again spare alarm inputs (about 8) should be available. The accustical signal may be reset manually while the optical signal may only be reset if the alarm condition has been eliminated. Table 1: Scram events * removing key from key-switch * manual scram * power level lin. channel 1 too high * power level lin. channel 2 too high * period too short * any fuel temperature too high * surface water temperature too high * primary water temperature too high * 220 VAC power failure •"• 110 VAC power failure 24 VDC power failure any low voltage power supply failure any high voltage power supply failure limit of transient energy reached * loss of pressure in pneumatic pulse rod system * waterlevel 1 meter too deep (2 channels) 2-2 Table 2: Shut down * current interrupt for control rod and shim rod magnets * opening of magnet valve of pulse rod pneumatic system * run down control rod and shim rod motors * activation of scram indicator * display of scram origin * activation of start-up interlock system Table 3: Alarm events * push buttom - control rod up - pressed and control rod not magnetically coupled * push buttom - shim rod up - pressed and shim rod not magnetically coupled * waterlevel too deep * door of the thermal column open * primary water temperature too high * any fuel temperature signal missed * no ventilation and power level greater than 10 kW * primary water cooling system not operating and power level grater than 100 kW * secondary water cooling system not operating and power level grater than 100 kW * purification circuit out of action * sumppump - primary cooling system - is running longer than 1 minute * sumppump - secondary cooling system - is running longer than 1 minute * conductivity of purification circuit too high * aerosolmonitor alarm * waterlevel in neutron radiography collimator too deep * loss of underpressure of reactor hall longer than 2 minutes and reactor in operation * fire alarm 3. NUCLEAR CHANNELS Each nuclear channel should be equipped with its own power supply with control display and with a variable high voltage (50 VDC - 1000 VDC/5 mA) with display. A failure should lead to a reactor scram and a test of each channel should be possible. It should be possible to vary the input sensitivity of the amplifier to ±20% to facilitate an electronic power calibration procedure. As mentioned before from each channel it should be possible to extract its signal through an isolation amplifier to use the signal for other purposes without feedback. The TRIGA reactor Vienna uses four nuclear channels being the - start-up channel - log- and period channel - linear channel 1 - linear channel 2 (or percent power channel) The requirements to these channels today do not differ compared to two decades before and are standard to most of the low power TRIGA reactors. One requirement which may not be standard is a reactor scram at 20% over power in any range of linear channel 1. At low power operation this requirements limits a power excursion to 20% over power of the set range and not to 20% overpower of nominal power. 2-3 Some improvement could be made with the chamber housing of the nuclear channels. It has been observed that chamber failures originate in the désintégration of the plastic isolation of the chamber cables due to radiation and/or humidity in the housing. The chambers itslef usually do not fail for an extended period, but the cables near the chamber connection look rather obsolete after about 3 years. One improvement could be a one meter piece of mineral insulated cable at the chamber connection to get the normal cable out of the high radiation area. In addition the chamber housing should be a vacuum tight system filled with an inert gas such as He. Further the chamber should be installed electrically isolated in the housing. As the compensated ionization chambers (RC6EB) are equipped with a high voltage surveillance system this should also be implemented in the reactor instrumentation. For the signal cable a triaxial cable should be used to avoid noise currents or pick-up especially at low neutron signals. The cable connectors should be different for positive-, negative- and signal connection to avoid misconnection and should be tight to avoid moisture influence. Finally the complete chamber housing should be mounted on a simple drive system to allow vertical movement {about ± 5 cm) for easy chamber adjustment during thermal power calibration. A proposal for an improved chamber housing is shown in Fig. 2. mechanism to move chamber housing vertically (± 5 cm) nobel gas filling plug electrical connections sealing suspension rods isolation plate isolated chamber Fig. 2: Proposal for an improved chamber housing configuration 2-4 4. REACTOR CONTROL SYSTEM The reactor control system in use has proven successfully. It uses presently the signal of linear channel 1 and of the log.- and period channel. A very useful feature would be an automatic power increase with a preset period between 5 s to 40 s. The rod position display should be both in digital- and in bargraph form. 5. TEMPERATURE CHANNELS Besides fuel temperature measurement by two instrumented fuel elements (totally. six thermocouples) several temperature sensors in the primary water system are required. The fuel temperature system does not require changes as it worked reliable during the past except with some thermocouple failures. It should be possible to select any of these six temperatures and to put the signal on a recorder. For the surveillance of the primary water system temperature measurement is required at primary water input and exit directly at the pool and at the pool water surface. This temperature measurement will be taken by PT-100 sensors in 4-wire technique and should be displayed digitally at the control desk. Each temperature instrument should have one lower and two upper variable alarm settings to actuate alarm or scram signals. Several additional temperature displays should be available as temperatures or also measured in other areas such as reactor building and outside air temperature. 6. OPERATION MODES At the TRIGA reactor Vienna the standard operation modes - off - normal operation - pulse operation are presently required with a possible extension for a reserve mode as special operation (such as square wave operation). The above mentioned three operation modes require some special conditions and interlocks which are well established and do not need to be repeated here. During pulse operation it is important that the transient and the fuel temperature is recorded over a preset time by a digital storage oscilloscope. Also the peak pulse power must be recorded. The stored values should be later on recorded on a strip chart recorded for documentation. After a preset time or after any transient limits have been reached the reactor is shut down still in the pulse mode. 7. SUMMARY After 26 years of operation and two reactor instrumentations (lifetime of instrumentation no. 1 about six years, and of instrumentation no. 2 about 20 years) the TRIGA reactor Vienna needs a new state-of-the-art instrumentation which should incorporate all international accepted 2-5 necessary instrumentation features as well as features specific to the facility due to its special application for education and training. Flexible up to date data processing and documentation is one requirement as well as a hard wired reactor protection system. The Atominstitut has obtained three quotations from three different companies each of them incorporate practically all the requirements. No decision has been made until now as no financial guarantee has been obtained from the actual reactor owner which is the Federal Ministry of Science and Research. /!/ H.Weiss, Some aspects of licensing, periodic inspections and backfitting of TSIGA reactors (this conference). 2-6 / I / REPLACEMENT OF THE COOLING SYSTEM OP THE TRIGA MAINZ REACTOR H. Menke Institut für Kernchemie, Universität Mainz D-6500 Mainz INTRODUCTION The present cooling system of the TRIGA MAINZ reactor consists of a primary cooling circuit, from which a part of the water is taken into the purification system (fig 1), and an open secondary cooling circuit with an evaporation cooling tower. The heat transfer from the primary to the secondary cooling circuit is performed by a tubular type of heat exchanger, which contains U-shaped stainless steel tubes, through which the demineralized primary cooling water flows. Normal water from the secondary cooling circuit, replenished from the public water system, flows on the shell side. Supporting and deflecting plates as well as the shell, which is inside clad with neopren rubber, are made of normal steel. In 1978 we learned from the TRIGA HEILDELBERG I reactor, where exactly the same heat exchanger was used for the same period of time that the supporting and deflecting plates were severely corroded. We therefore did not dare to clean the secondary side of the heat exchanger, because we were afraid to destroy these plates and to damage the tubes. As a result of the returning inspections of the reactor facility by the competent authority, it was advised already in 1979 to install a new heat exchanger.- At the same time it was recommended, to separate the primary cooling circuit and the water purification system, to prevent a waste of energy by heating up the primary cooling circuit and the pool water by the primary water system pump; the latter must be in operation even with the reactor shut down, because of the water purification system in this circuit. Probably due to the deposition of lime and organic matter on the shell side of the tubes, the heat transfer rate decreased in the following years, as can be seen from table 1 . At the same time the advise to replace the heat exchanger became more urgent. Meanwhile, in 1984, a new KTA-rule became effective. According to this rule, the pressure in the secondary cooling circuit must be permanently higher than in the primary cooling 2-7 circuit, tö prevent radioactive water from passing from the primary into the secondary cooling circuit in the case of a leak in the heat exchanger. Otherwise the water in the secondary cooling circuit must be permanently monitored, which reqires the investment of expensive apparatus or manpower. Table 1 HEAT TRANSFER FROM PRIMARY TO SECONDARY COOLING CIRCUIT The pool-water was heated to 32 °C with the reactor in operation at 100 kW, primary cooling circuit on, secondary cooling circuit off. Outdoor temperature 5 'C. Test 1 : reactor shut down, secondary cooling circuit on Test 2: reactor in operation at 100 kW, secondary cooling circuit on Year Test 1 Test 2 degrees/hour degrees/hour 1979 6,0 2,8 1984 3,7 1,6 1988 2,4 0,4 DESIGN So we started to plan new cooling circuits, the design of which is shown in fig. 2. The essential differences between the old and the new water system are: 1 The primary cooling circuit and the water purification sytem are in separated circuits; 2 the secondary cooling circuit is provided as a closed system, allowing to keep a higher pressure than in the primary cooling circuit; 3 the heat will be delivered to the environment by means of a cooling tower, using ventilation, which may be enforced by sprinkling water on the tubes; 4 a plate-type heat exchanger will be used instead of a tubular type; 5 all parts of the water system which are essential or due to exchange are provided twice; 6 all piping and valving will be of stainless steel instead of aluminum, also the heat exchanger. 2-8 PRIMARY COOLING CIRCUIT The layout of the primary cooling circuit is shown in fig. 3. The system consists of stainless steel piping and valving, a stainless steel-type heat exchanger and two pumps, which are used alternatively, thus making possible an exchange of a pump during reactor operation. The tank-inlet stream of the water is directed towards the core to delay the upcome of N-16, formed in the core during reactor operation. Temperatures and water pressures are measured before and behind the heat exchanger. WATER PURIFICATION SYSTEM The water purification system, also shown in fig. 3, forms an independent circuit. The intake of this circuit is a surface skimmer, which collects at the same time foreign particles, floating on the water surface. Piping, valves and vessels are made of stainless steel. Again two pumps are provided, one of which is in opereation, pumping the water through the purification system. The water passes through a water monitor box, where conductivity and activity are measured, then through a filter, a mixed-bed ion-exchanger and again through a filter. Filters and ion-exchangers are provided redundant, to allow an exchange during reactor operation. At the outlet of this circuit is again a conductivity probe and a flow meter. Pressure differences are measured parallel to filters and ion- exchangers . All piping, immersing into the pool-water of the tank, has 13 mm vent-holes below the water surface to prevent a drainage of the pool-water in the case of a broken pipe. SECONDARY COOLING CIRCUIT In fig. 4 the secondary cooling circuit is shown. Stainless steel piping makes the connection between heat exchanger - cooling tower - heat exchanger. Two parallel pumps are again provided, one of which is used to transport the water in this circuit. Two motor-valves may be used to close the piping against the heat exchanger. As this circuit is closed, an expansion-vessel with safety-valve is provided. The cooling tower operates as air- or evaporation-cooler, 2-9 whatever is demanded by the heat to be removed from the circuit.- Water from the public system will be fed into the circuit automatically to maintain the demanded higher pressure. At the same place the circuit may be emptied into the sewer system. Temperature and pressure are measured before and behind the heat-exchanger, also the flow-rate of the water. The differential pressure between primary and secondary cooling circuit is measaured and used to establish the prescribed higher pressure of the secondary circuit. SURVEILLANCE AND SAFETY The position of all valves in fig. 3 and 4 will be indicated on an illuminated diagram in the controll room and in the machine room. For this purpose the valves are equipped with end- position switches, indicating on the switching diagram the position "open" or "closed". The pumps will be indicated as operating or not. A redundant temperature probe in the pool water will at preselected temperatures switch on and off the secondary cooling circuit and the different steps of the cooling tower.- A redundant water-level meter in the reactor tank will indicate the water level and will give an alarm on reaching the lower or upper limit. All temperature and pressure measurements of both cooling circuits will be indicated and the temperatures will be registered in the controll room. The differential pressure measurements will be indicated in the controll room and will give alarms on reaching lower or upper limits. If a leak occurs in the heat exchanger, water will pass from the secondary into the primary cooling circuit because of the higher pressure in the former. The following will occur.- 1 Water will be fed into the sedcondary cooling circuit, because the differential pressure between secondary and primary cooling circuit is falling; this will be done only for a preselected limited time; 2 increasing conductivity will be indicated in the primary cooling circuit; 3 on reaching the upper limit, the water-level meter in the reactor tank will a) stop the feeding of water into the secondary circuit, b) switch off the pumps of the primary and 2-10 secondary cooling circuit, c) close the motor valves of the secondary cooling circuit before and behind the heat exchanger; 4 on reaching the lower limit the differential pressure indication between secondary and primary cooling circuit a) the pumps of the secondary and primary cooling circuit will be switched off and b) the motor valve s of the secondary cooling circuit before and behind the heat exchanger will be closed. CONCLUSION The detail planning of this project was completed in December 1987. In response to regulatory requiremnets a motion for replacement of the cooling system was proposed to the authorities, to obtain the necessary licenses. The money for the performance has been granted meanwhile. About one year after obtaining the licenses, for which we are currently waiting, we can start with the performance, which makes necessary a shut-down of the reactor for about eight month.- It should be mentioned finally that in planning all these changes a later upgrading of the steady state power of the reactor to 300 kW was taken into account. 2-11 FILL PUMP -t>><3- /—D£J—>—D§}—0" GEIGER TUBE, TEMPERATURE PROBE AND CONDUCTIVITY CELL -© r-C^h- FILTER | .WATER-SURFACE SKIMMER ^Bh^ -© OEMINERALIZERJ 1/2 IN.VENT ^ HOLES CONDUCTIVITY CELL | FLOW METER REACTOR TANK -t&h •Ä Fig. 1 Present Reactor Water System 2-12 EXPANSION VESSEL Qmnx = 300KW * / 0 A A A A T = 22.2°C F 31A/B Ï=22.2°C F 32 A/B P = 100KW SEC0N0ARY COOLING CIRCUIT W02 Ni T = 25 °C F 30 A/B SEWER SYSTEM : 3 Q=4m /h P50 P40 -« WATER FEED AND PRESSURE KEEPING T=25°C —o- 3 PURIFICATION CIRCUIT Q=30m /h OP30 —oP10 — PRIMARY COOLING CIRCUIT Fig.2 Design of Cooling Circuits OP20 Fig. 3 Primary Cooling Circuit and Purification Circuit i 2-U A A A A A 7V" S68t S 52 SS3 Zi. Ä- J^ S66 •a© B50 V 0*0—rg COOLING TOWER S 63 s fi; IP WITH CLOSED CIRCUIT SEWER SYSIEM I SECONDARY COOLING CIRCUII AUTOMATIC WATER FEED AND PRESSURE KEEPING SEWER SYSTEM Fig. 4 Secondary Cooling Circuit Use of a Digital Reactivity Meter on TRIGA Reactor for Instrumentation and Training Purposes Bogdan Gktmac University of Ljubljana "J. Stefan" Institute, Yugoslavia for presentation at 10** European TRIGA Users Conferrence Atominstitut Wien - Austria, September 14* -16**, 1988 univerza, e. kardeija. ££ institut "joief stefan" ljubljana, Jugoslavia w w Odsek za reaktorsko fiziko ™ 2-17 Abstract Development of a realtime reactivity calculation algorithm, based on recursive so lution of the inverse point kinetic equation, and adapted for digital computer calcula tion, is described. Application of this algorithm to an "off-the-shelf microcomputer is presented in order to produce an extremely low cost piece of equipment that can both alleviate the work of operators as well as serve as an extremely versatile tool for training purposes. Various applications of this device in a day to day work of a research reactor operators are given. A method of control rod worth measurement after core recon figuration using this digital reactivity meter is described. Demonstration of various effects, such as control rod insertion or extraction, void coefficient measurement, temperature feedback effects, etc., is described. 2-18 1 Introduction A "Core Design Report" is prepared for a reactor for which the refuelling has been scheduled. The parameters that can be experimentally verified and that are essen tial for a safe performance of a refuelled core are measured on a low power level immediately following first criticality after refuelling. Those are, for instance, the differential (££) and integral (Jo* £f • <&) reactivity worths of the control rods, the temperature coefficient ($j») and, if the reactor has chemical reactivity control system, the boron coefficients and boron endpoints. When performing any of the above mentioned reactor experiments, we must limit ourselves to the use of ionisation cells that constitute permanent reactor instrumenta tion meant to monitor the neutron fluxi n the core. These cells are usually positioned in such a way adjacent to the reactor core that they perform effective space integra tion of the angular neutron flux. All measurements should be performed at such power level that no temperature feedback on reactivity is observed. The power level is only slightly rised above the zero level so that any external neutron sources that may be present in the core can be neglected. All the above mentioned core parameters can be measured using extremely sim ple experimental means, but it must be pointed out that such measurements and subsequent data evaluation take a lot of time. To cut down on this time a digital measuring device has been designed that can sample the ionisation cell current and in real time calculate reactivity from the sampled values. Only the ionisation cells that are already present as a part of permanent reactor instrumentation are being used. 2 Working Equations of the Digital Reactivity Meter Based on the assumption thai the ionisation cells adjacent to the reactor core perform effective space and energy integration of the angular neutron flux we try to develop the digital reactivity meter algorithm from the well known equation: and precursor concentration equations for six groups of delayed neutrons : &/(*) _"(*)• ßi x. r(i\ ., , * (9s 2-19 where the relevant quantities are : • n(t) - current over ionisation cells , • p(t) • reactivity , • ß - delayed neutron share, • Î* - prompt neutron generation time , • Xj - j* delayed group decay constant, • */(*) " effective neutron concentration of the j* group , • a(t) - effective external source , • ßj - j* delayed group share . The space and energy averaged kinetic equations (1) and (2) can be obtained by the perturbation theory approach (ref. 1,2) which at the same time gives expressions for/? = Ej=iAandf*. Set of equations (2) can be analytically integrated to yield the result: cy(i) = <-V<. â. jf' »(*). «M. d8 j = 1, •.., 6. (3) Now we perform "per partes* integration of the integral in (3) and obtain: By introducing (4) into (3) and after some rearrangement we obtain for precursor concentration equations: *&••&;•{•&-£*$•*+*+•*> y-V»A (5) By putting (5) into (1), where we neglect the external source term, we obtain the following expression for reactivity: ^-^•KO+j(M-fjf^-rW«..«, (e) 2-20 and, by knowing that ß = E/=i Ä» we end up with the final expression: r • *$• = M• -M-pi- /.' *$*• • rv«• *. m We now expand n(t) into Taylor series on a certain time interval (*, t + Ai) and, by catting the series after the first derivative, obtain at some time t + r from the interval: n(* + r)*n(t) + ~&-T. (8) By putting (8) into the integral from (7): •M-J£TP-«'MM'-*. *-».•••.• (») we easily obtain from (9) a set of recursion formulas for the time point t +1 • Ai: *(t + iA*} = î,.(*-iAiH-V» + Ai^U^-U* (10) where the weight functions: t>,- = Cb(A,--At)-Ci(A,--A<) «>,- = Ci(ArA*) (11) are given by degenerate hypergeometric function: Q(x)= /V«~-{1"f) dy / = 0,1,... (12) The procedure is now simple: if the reactor is stationary at times t < 0 the integrals (9) are equal to zero since ^ is zero. If we collect 2 • m -f-1 equidistant samples over the time interval t and calculate smoothed n(t) and *jjü. values at the mtt sampling point ( at time t + jAi), we first calculate integrals from equation (10) and then the reactivity at time t+^At from equation (7) and then amply recursively continue this process. The algorithm is stable. The calculation of the smoothed values n(t) and &jfi- is done by the use of the convoluting integers method (ref. 3) in fivepoints . 2-21 3 Digital Reactivity Meter Hardware and Software On Figure 1 the schematics of the digital reactivity meter, assembled from the "off - the - shelf* components is given : Ion Current Amplifier Celb and W*» ADC Personal DAC Computer y-t recorder CUT W),»(*» Mass Storage Figure 1 : Digital Reactivity Meter Schematics Standard IBM Personal Computer hardware with 256 kByte RAM and at least one floppy disk drive is necessary. To digitalise the n(t) signal and to plot out the reactivity one of readily available 12 bit AD/DA expansion cards was used, namely Microway 6910. For this AD/DA unit it is very straightforward to assemble all the necessary support routines. Signal n(t) was amplified by Keithley 614 electrometer, signals n(t) and p(t) can be recorded on any strip - chart recorder. Main program, based on the equations from the Section 2, was written in standard IBM PC Fortran. All the hardware support routines were written in Assembler and added to Fortran code at link time. 2-22 4 Results Various versions of the reactivity meter, based on the equations from the Section 2, were assembled, tested and were regularly In use on "J. Stefan" Institute TRIG A research reactor and in two nuclear power plants to perform regular post-refuelling startup physics tests. 4.1 TRIGA Reactor On Figure 2 a segment from the "rod swap* measurement of the JSI TRIGA reg ulating rod is given. Intercomparison of the values for the integral regulating rod worth shows excellent agreement ( 651 pern for the "rod swap", 640 pern for the "rod drop", 650 pem for the "rod insert" Method ): JSI TRIGA, 1986, "Rod Swap" Snap ASS/COMP 0 2» «» Timkmxaii) Figure 2 : »Rod Swap" Measurement of the JSI TRIGA Regulating Rod, 1985 Digital reactivity meter is being regularly used by the JSI staff during training courses for the NPP KrSko operators to demonstrate such effects as temperature i"rod insert" method is explained in the subsection 4.3 2-23 feedback, response to step power change, void coefficient determination, "swap* of the fuel elements, etc. 4*2 Krfto Nuclear Power Plant One of the versions of the digital reactivity meter is being regularly used to perform startup physics tests in the Rrâko Nuclear Power Plant3. On Figure 3 a segment of the dilution measurement of the Rod Control Cluster Assembly (RCCA) "A* during 1987 startup is given: NPP Krsko, 1987, RCCA "A* alone Caution nwHUraiHnt » a 2» 15 » (\ A 5 T A A A r / I 0 J 1 / -5 / \ J \J\t ( 1/ \ -10 V 1/ -IS t 2» 4Q0 a» Tin*/(saooncaj) Figure 3 : RCCA "A* Dilution Measurement, KrSko NPP, 1987 startup »Kriko NPP is Westinghouse two-loop PWR with « 630MWe output that went into operation in 1981 2-24 4.3 Rod Insert Measurement at Kernkraftwerk Obrigheim »Rod Insert" is a new method that is under development at the "J. Stefan" Institute and uses slightly different approach than the dilution/boration or rod swap mea surement, where the reactor is kept approximately critical during the measurement. When measuring by rod insert the reactor is made subcritical by continuous insertion of the control rod. Signal n(t) is scanned and stored online (appx. 3-4 minutes for full rod travel). Irom this stored signal, after appropriate smoothing and filtration, the integral reactivity curve of the control rod is calculated offline, using basically the same set of equations as presented in Section 2. This method was tested with success on the TRIGA reactor, in Kräko NPP and in Siemens made nuclear power plant Obrigheim in the Federal Republic of Germany. On Figure 4 the differential reactivity curve'of eight simmetrically distributed control rod clusters of the NPP Obrigheim is given: KWÛ 1988, ref. cluster, differential CufPpwww flMJdfe range,no d tovt -»«»«i -^— 43 - t ^ \ 35 - \ / a i \ 3 - \ a a \ su 2 - a• C 15 - - y \M as - / o --U \ 0 40 80 120 1» 200 240 280 Stop îwvtod Figure 4 : Control rod clusters YR22,YR25,YR31,YR28,YRlYR2,YR3,YR4, differential reactivity curve, Kernkraftwerk Obrigheim, 1988 »calculated integral worth for this cluster is 884 pem, integral worth measured by rod insert is 871 pan 2-25 5 Conclusions Digital reactivity meter, developed at "J. Stefan8 Institute, and its application on a research reactor and in a nuclear power plant is described. Results are given for varions methods of measurement showing its versatility and applicability both for research and power reactor operators. Great advantage is also its extremely reasonable price - this reactivity meter is assembled only from the low cost "off the shelf components. 6 References 1. G. R. Eeepin, Physics of Nuclear Kinetics, Addison - Wesley, 1965 2. B. Glumac, Digital Reactivity Meter, US DP 4976,1987, IAEA Research Con tract No. YUG - 4088/RB 3. A. Savitaky, M. Golay, Smoothing and Differentiation of Data by Simplified Least Squares Procedures, Anal. Chem., Vol 36, No. 8, Page 1627,1964 2-26 Digital Automatic Control System for the TRIGA Reactor Susumu Harasawa Institute for Atomic Energy, Rikkyo University 2-5-1, Nagasaka, Yokosuka, 240-01, Japan Mohamad Amin Sharifuldin, Syed Nahar Syed Hussin Shahabudin Unit Tenaga Nuklear, Kompleks PUSPATI, 43000 Kajang , Selangor, Malaysia and Kunishiro Mori, Sadayuki Uchiyama Clear Pulse Co. Ltd . 6 - 25 - 17, Chuuoo, Oota-ku, Tokyo, 143 , Japan 1. INTRODUCTION One of the authors has designed a digital automatic control system with two sets of comparators for the TRIGA reactor of the Rikkyo University " . The control system has been designed in resemblence with manual operation of the reactor. The fundamental idea was that the time interval of the digital control action depends on the deviation between the power output and the demand power. In the case of the Rikkyo University Reactor, when the deviation is smaller than ± 0.2 %, no action occurs, when the deviation exceeds ± 0.2 % but below ± 3. 5 %, a regurating rod responds for 0.1 second with stepwise movement at an interval of 1 second, and when the deviation exceeds ± 3.5 % the regulating rod moves continuously. These actions has been materialized by using two sets of comparators and a pulse generator. Another control system for the PUSPATI TRIGA Reactor applying similar idea was designed and tested in 1987. The control system has one comparator which detects the difference of the reactor output and demand power as conventional control systems have, and the time interval of the control action of the newly designed system is determined by a voltage-to-frequency-convertor. The idea and results of the automatic control systems for the Rikkyo University Reactor and PUSPATI TRIGA Reactor are discussed below. 2-27 2. LOGIC CIRCUIT 2.1 The logic circuit for the Rikkyo University Reactor Input-output relation of the logic circuit in Rikkyo University Reactor -3. 5X- ..0. 2M -0. 2 1) üiSliiiyü 2) ilSiüaüiili lsec 3) -3.5X INPUT OUTPUT Fig. 1 Illustration of input and output signal of the logic circuit of the Rikkyo University Reactor automatic control system 2-28 While the input to the logic circuit are the signals from the ionization chamber which are continuous, the output from the logic circuit are pulse-like and are divided into three categories according to the deviation of the output power from the demand power : 1) Nan-sensitive region ( dead band ) 2) Region of pulse-like movement of a control rod 3) Region of continuous movement of a control rod These relations are illustrated in Fig. 1 1) When the deviation of the power-output from a demand power is between ±0.2 1 there is no output from the logic circuit. 2) When the deviation is between 0.2 % and 3. 5 %, or - 0. 2 % and - 3.5 1 the output of the logic circuit are stepwise with 0.1 second width and 1 second interval, and 3) when the deviation exceeds 3.5 %, the output of the logic circuit is constant. The diagram of the circuit is shown in the appendix [ Fig. A-l ] . Some power output under the automatic control operation in the Rikkyo University Reactor Some results of the power fluctuation of the Rikkyo University Reactor under automatic control are shown in the appendix [ Fig. A - 2 ( 100KW ), Fig. A - 3 ( 100W ), Fig. A - 4 ( 10W ) and Fig. A - 5 ( 1W ) ] . The fluctuations due to the reactor noise are smaller than ± 0.2 % of the operating power higher than about 10 Watt. On the other hand, the reactor noise are larger than ± 0. 2 % of the operating power lower than 10 Watt. As can be seen from the figures, the average power level is satisfactorily controlled even when the noise fluctuation is larger than ± 0. 2 % . At demand power level of higher than 1Q Watt, the noise to singnal ratio is smaller than ± 0. 2 % and the output power is controlled whithin ±0.2 %. 2.2 Modified logic circuit for the PÜSPATI TRIGA Reactor We have designed a new control system with variable interval stepwise action for the PUSPATI TRIGA Reactor in order to reduce the power fluctuation at high power region. In the new logic circuit the concept of resemblence with manual control 2-29 a voltage to frequency converter. The frequency of the pulse is proportional to the power deviation, that is, the pulse interval is inversely proportional to the deviation. The input-output relation of the logic circuit in the system is illustrated in Fig. 2. /\ 0. 1 i a c -41- Logic. £ \ - s Circuit V INPUT OUTPUT Fig. 2 Illustration of input and output signal of the logic circuit for the PUSPATI TRIGA Reactor automatic control systea Pulse interval r = K / 1 g I , where e is the deviation of the power output from the demand power. K is a constant. 3. CßmSßL SYSTEM in the PUSPATI REACTOR 3.1 Back-up automatic control system for the PUSPATI TRIGA Reactor As the PUSPATI TRIGA Reactor has an existing automatic control system, we designed the digital automatic control system as a back-up system. Digital automatic systems usually consist of following three parts : Sensor Logic Circuit , and 2-30 Actuator . As the sensor can be shared with the existing control system, we only prepared the logic circuit and the actuator for the back-up system. The back-up control system is connected as shown in Fig. 3.. control rod » reactor automatic controller «—*• neutron sensor e* iilliiïipiliiili lijpijjifsijjj Siiiiiiiii Back-up Automatic Control System Fig. 3 Automatic Control System of the PUSPATI TRIGA Reactor 3.2 Logic circuit The simplified block diagram of the logic circuit is shown in Fig. 4. The gain of the amplifier for the deviation from the demand power, A, and the pulse interval, v, are designed to be adjustable . The detailed logic circuit diagram is shown in the appendix [ Fig. A - 6 ] . 2-31 4. RESULTS AND DISCUSIONS Power fluctuations of PUSPATI TRIGA Reactor under (a) manual, (b) the original automatic control ( GA AUTO ) and ( c ) the digital automatic control ( DIGITAL AUTO ) are shown in Fig. 5 ( 50 W ), Fig. 6 ( 500ÏÏ ), Fig. 7 ( 5kW ), Fig. 8 ( 50kW ) and Fig. 9 ( 750kW ). It can be seen from these figures that the digital automatic control system has reduced the frequency of the control rod movement compared to the operation under the original automatic control system, and power fluctuation has been improved for demand power of 5 kff or lower. However, there is no improvement of average deviation from demand power at 50 kW or higher. Output power under both automatic control systems of PUSPATI TRIGA Reactor have larger fluctuation than that of manual mode. This is because both automatic control systems tend to enhance the deviation by responding constantly to statistical fluctuations. The digital control system was also designed to solve this enhancement by varying the interval between control action through the use of a voltage- frequency converter. However, the result showed that this is not sufficient to solve the problem. From the experience with the automatic control system of the Rikkyo University Reactor, inclusion of a dead band into PUSPATI TRIGA Reactor digital automatic control system will overcome this problem and reduce the deviation. ACKNOWLEDGMENT The authors would like to thank Mr. Adnan Bokhari for checking the circuit and connecting the back-up system to the existing control system, and Mr. Gui Ah Auu for fruitful discussions and critical reading . Fund for the back-up system is financed by the Japan International Cooperation Agency. REFERENCES *) S. Harasawa and K. Kawaguchi : Automatic Control System of Rikkyo Research Reactor, Proceedings of First Asian Symposium on Research Reactors, 247- 252, 1986 2-32 50ld HRNUflL S4 • • • • Mr»1 53 52 51 50 49 48 47 46 n RH inn 150 ?nn ?5n 3 50ÜJ GR-fiUTO 50 100 150 200 250- 300 SOU DIGITAL flUTG 200 250 300 TIME (Sec) Fig." 5 Example of the power output of 2-33 PUSPATI TRIGA Reactor at s5n0 mIV 500ÜJ MANUAL 50 .100 150 200 250 . 300 500U GRflUTO 50 100 150 200 250 300 500UJ DIGITRL RUTÖ o 100 150 200 250 300. TIDE! (Sec) Fig. 6 Example of the power output of 2-34 PUSPÄTI TRIGA Reactor at 500 « 5'KbJ HRMUfiL 0 50 100 150 200 250 300 5Küü Gfl RJTG 0 50 100 150 200 250 300 5KUJ DIGITAL RUTO 50 100 150 200 250 300 TiriE (Sec) Fig. 7 Example of the power output of 2-35 PUSPATI TRIGA Reactor at 5 KW 50KUJ MANUAL o 50 100 150 200 . 250 ... 300 50KUJ GA AUTO 50 100 150 200 250 300 5QKU DIGITAL AUTO 100 150 200 250 300 ?-% ^ (Sec) Fig. 8 Example of the power output of 2 3b PUSPATI TRIGA Reactor at 50kff 750K MflNURL X LU 3 Q_ 0 50 100 150 200 250 300 750K Gfl flUTG x OL !^ CD Q_ 50 100 150 200 250 300 750K- DIGITAL AUTO X LU 3 CD Q_ 100 .150 200 250 300 ... TirF (Sec) -37 fig- 9 Example of the power output of PUSPATÏ TRTGA Reactor at 750! Fig. A - 1 , Diagram of logic circuit of Rikkyo University Reactor automatic control system. Example of the controlled power output ,„ of Rikkyo University reactor at 100 KW . Fig. A - 3 , Example of the controlled power output of Rikkyo University reactor at 100 W . Fig. A - 4 , Example of the controlled power output of Rikkyo University reactor at 10 W . Fig. A - 5 , Example of the controlled power output of Rikkyo University reactor at 1 W . Fig. A - 6 , Diagram of logic circuit of PUSPATI TRIGA Reactor automatic control system. Fig. A - 7 , Diagram of actuator of PUSPATI TRIGA Reactor automatic control system. 2-38 § § s § © Fig- A " ! ' DÄ?cf ÄColCisry?t?iL0f Rikky° UniVersity Reactor 2-39 T " ~" L^ — — LJJI^_: S=^ j__ y—i . •—.—*= i £1* —• • ï— - -•i •!••-! 1 H 1 r- •: i —. , . J . . —• ; : 1 . —; j : j 1 • '—r^p^rri. i r ._. _ , ' ' | p :zrz. -4 f- •• i i i = •: 1 i-l.-:::- • .- ; ..•;•: |—. : :. : : : 1 • • i. - __ _, _.._.,,.,,.,,. ._, ..... i- X \ -1 • é== — - _; 1 1 :—:—:—J 1 j 7* •:••)::: ! ••• • i:!1: ; • ::r:-i':! 1 1 ••—1 .—i ••:..'! :—i . «^•-•••l : . •••:; i ; i' —: r—! ;—: , , . , , , . i :—K^— - ___ i 1 1 ..." ' • ' • •: ••!• - ; " i: ,; ' i: : - : : : - • 1 • :^|7H ' - • Ï : 1 •. i • - . • • | • • . 1 t: 1 J i 1 . ' • • . • • 1 -_ .: . I • i i ; . .• • • 1 • 1 : - r- < i J . . 1 • . . . i • • i ••. 1i . .,i i -. -0..jq '. .- • ; ,•'•'• i... ^ 1 j 1 • •-:••.•;•; UL_ —i-^i-i , , , _; —• i •- 1—*—-H i. ' : ' \ l ' •. '~ j—: , ; 1 ; "—1 i . • t .-,...• | • r1) • • i ••!••.. r • i J •i . " -.]• • <£U » ' • : • • -4U .- : , . i • • -. -jj : • 1 . . 1U IÙ . . j . . • l j . : : 1 • • - . ..,-*: • j i i . i • • ' t | t i : . i • i i • j • • ( | i • • 1 i • • . 1 . , i . 1 • . . . | i , . . • • i • .... 1 • \ Vi*<_| ("• '; . i , , , i ....,.,.. | . ... , , . " •• • . . • i i • . , 1 1 .. 1 1 1 i I . : 1 , . -/; !.*_* I 1 . 1 . 1 1 1 . 1 1 -1 I 1rf— • . j ,,.,,.,, 1 . • ,(,..., . i . , M 1 I > 1 < . < - 1 1 i 1 1 i 1. 1 , 1 .. 1 ..... | • ! | | : ..• i .... | •• ' • •• V ' ' .' ?•] .","" 1 , | , j • , . j j , -, . , , 1 , !. L- ' 1 : . •• 1 .... | ..i- | i -.. —;, I ' Lr*"1-' l i . i ; . . 1 1 i ' i . . . . ; 1 • ' i ' • • i i 1 . 1 , ." 1 1 | • , . • 1 . . 7\ . il •i l .'. ' ' j . •• ...... 1 • • : | • . . . 1 . . 1 • i 1 1 1 . : • • 1 . . l: 1 II] 1 1 t i i • i i | i • : : : i i • _ • 1 i "'I • • ! • ' ) • . - j • 1 : 1 , ... i i j ! i i , J : * t i i . , • J .t , . : 1 1 t ( f • 1 1 ] 1 1 1 1 J . r 1 I . • i . 1 1 . ' 1 j -« " '• 1' ' ' 1 1 T > 1 :' ' i i ; t t i | J i i . | i \ i i t ' ? 1 J -H • ' i • • : i L • i • > i • • • • i • i • < 1 ' • i . • i i . • ; . i i ' i i i i i t j • ' i i | ; i » i . J- t i • . . . 1 .'il- | • . . i . • 1 • 1 • i • • j • . . i . . • • | i • i . . . i ... •' , ' "f i" I ' ' 1 1 r • 1 r \ t , i -. i i i i i ! ! ! : : . 1 1 : : : : [,i-> 1 1 1 i . . ! . i . ii(l 1 1 1 l"l 1 1 1 i 1 [ 1 1 . . • 1 1 • ! 1 • ' • 1 . . . t ... j . . | i I 1 i ' i ' ' ' . S; • I • i i i i i i ; j i i > i i t 1 i . 1. 1 1 1 1 I . . I • :. » = ' 1 • ' . : • : i i • • • i --;••• i • : - ( • J - • • 1 • 1 . J I jç..,.. i - - • • • • 1 • i . .- . . 1 . 1 . . 1 . . . . *""'• "1 .' Fig. A - -2 , Example of the controlled power output of Rikkyo university reactor at 100 KW . 2-40 Fig. A - 3 , Example of the controlled power output of Rikkyo University reactor at 100 W 2-41 Fig. A - -4 , Example of the controlled power output s of Rikkyo University reactor at 10 W 2-42 Fig. A -'5 , Example of the controlled power output of fiikkyo University reactor at 1 W . 2-43 I î * % 3 ^ «î 1 1 5 S VF3 ' îl-l « I . | « I j TMm RSffF^w. Fig. A - 6 , Diagram of logic circuit of PUSPATI TRIGA Reactor automatic control system.2-44 1 »0 SS ad -*—t ! r o i ____ *5t 1 Ç* ? Ci a at Fig. Diagram.of actuator of PUSPATI TRIGA Reactor automatic control system. 2-45 I I OPERATING DATA DOCUMENTATION SYSTEM FOR A RESEARCH REACTOR F.Kasparec*, J.Hammer* Atominstitut Vienna, Austria 1. History and Discussion of Requirements The documentation of reactor operating data is an important task at research reactor stations. Though not directly neccessary for the safety of the reactor, it is an essential tool for maintenance and safeguarding purposes and therefore made obligatory by the licensing authorities. Until 1985, the documentation system of the TRIGA reactor in Vienna com prised an automatic data logger with a separate hardcopy terminal. A textual report on all channel inputs was printed out each hour at normal conditions. When given thresholds were exceeded, warnings or alarms were issued together with an out-of-order log message, which conveyed information about the offending parameters. In 1985 the development of the documentation system described in this paper was started. It was based on former experiences with the data logger which had led to a selection of representative data and useful data formats, but also made clear that the paper reports produced by the hardcopy terminal were not useful for further evaluation. The documentation system was expected to be an extension of the logger- based system. Data had to be directed to the in-house computer center rather than to a terminal, though the terminal would be kept for reasons of diversity. In particular, the documentation system had to meet the following require ments : a) The central computer used for data archivation (a VAX-11/750) had not to be obstructed beyond neccessity. It was essential to avoid time-consuming high priority jobs. b) Data security was neccessary, in the sense that VAX system crashes would not cause losses of acquired information. c) A high degree of automatization, especially in the system crash and recovery phases of the VAX, had to be implemented in order to minimize human failures. d) Upon request, data had to be available at any terminal of the VAX within seconds or minutes. (This was assumed to be particularly important in the case of warnings or alarms.) e) Data management tasks (sorting into daily reports etc.) had to be performed *) At present at Alcatel-Elin Research Center Vienna, Austria (F.Kasparec) and PSI Wiirenlingen, Switzerland (J.Hammer) 2-47 automatically, so the operator would only have to copy data to tape and clear up the VAX disk directory every week or month. f) The transmission protocol used to transmit collected data to the VAX had to be portable, i.e. not relying on any specific hardware. 2. System Description i • N< Data 2.1. Microcomputer as Prebuffer P • Ù A U< Logger Hardeopy T To meet all these requirements, Tarminal 3- special prebuffer hardware was provided to connect the data logger to the VAX. Because of its proven reliability an Epson-HX20 microcomputer was selected for this purpose. Microeomputar 3=2. Today a personal computer would also Tap* be a good choice, especially because of Epson HX20 Cartridge its large memory and built-in mass storage which is far more powerful than the microcassette drive of the HX20. However, the fact that the operating system of a PC is resident in dynamic RAM may cause reliability problems, which would have to be solved by special watchdog hardware (e.g. a MAX-691 chip on a prototype adapter card in a PC extension slot). Fig.l depicts the structure of the documentation system hardware obtained by introducing the microcomputer. Fis.l: SYSTEM STRUCTURE The microcomputer software performs five tasks: a) Input Processor: This task processes the data logger output in order to cut the data stream into Meaauramen« records. Oat* t Tim« The logger output stream is a sequence of ASCII- Measurement coded text lines of 23 characters each, called Comment frames, which contain information about the result Measurement of a physical measurement, a comment, or Measurement -*± information about date and time. The latter case occurs whenever a new measurement cycle starts and is used by the input processor task to identify # the begin of a new record (Fig. 2). The number of Comment frames per record is variable. Measurement _^=L The input processor task also assumes the end of Data > Time the current record whenever the stream pauses for more than 10 seconds. The frame following a pause always is a date/time frame. Fi6.2: LOGGER OUTPUT 2-48 b) Buffer Management: This task handles Leading End accesses to a 24kByte ring buffer, (Writ*) using a special algorithm which will be discussed in detail in chapter 3. All newly received data from the data logger are written to the leading end of the occupied memory area as if they were pushed on a stack, whilst data for transmission to the VAX are Trailing End taken from the trailing end. (R#ad> Since any buffer access beyond top or below bottom is wrapped around, repeated accesses make the occupied Fis.3: RING BUFFER ARCHITECTURE memory area wander through the ring (Fig.3). c) VAX Interface: This task serves transmission requests issued by the VAX by sending entire data records to a VAX terminal port, which are then cleared out of the ring buffer. As only whole records are handled, the time-out mechanism mentioned at the end of (a) makes sure that the most recent information becomes available as soon as the data logger stops sending. Without time-out, the input processor task would have to wait for the anticipated date/time frame (leader frame) of the next record to identify the end of the current record. All communication runs through a serial 1.2kBaud link using 20mA current loop standard. A special protocol, which will be discussed in detail in chapter 3, traps line breaks and VAX system crashes, so no data are lost. d) Buffer Supervision and Tape Access Control: This task periodically checks the ring buffer occupation. If the amount of occupied memory area exceeds an adjustable threshold higher than 16kByte, a 16kByte section of stored data is scanned for end-of-record marks, proceeding from the trailing buffer end in forward direction. The last end-of-record mark found identifies a memory block containing the largest number of entire records that fit into the 16kByte space (see Fig.4). This block is then saved out to tape in reverse order. After successful completion of every transmission cycle to the VAX, which is assumed to have emptied SAVE the buffer except for partial records (when input OUT from the data logger is in progress), this task checks for the existence of data blocks on tape. J TPMTH In case that some is found the block is read back and appended to the trailing buffer end, pro ceeding backward character by character. In case Fi6. 2-49 VAX requests received during an already active save-out or read-back operation are rejected. Typically, the reason for the impending buffer overflow is a line break or VAX system crash, so requests need not be expected during a save-out operation. e) Tape Management / Operator Interface: This task controls the HX20 micro- cassette drive as well as operator accesses to this drive, which are the only manual intervention possibly required. Together with six multicolor signal lamps the HX20 liquid crystal display informs the operator about the system state. Tape management is stack based, i.e. the blocks exchanged with task (d) are stored first-in / last-out. Each of the MC90 tape cartridges used can contain four 16kByte blocks in distinct sections. The cartridges need no formatting; the sections are assumed at fixed offsets and accessed by first rewinding the tape to the begin and then winding forward to the desired location. This task needs the most extensive functional modification to run on a PC. Fig. 5 illustrates the data flow among the five tasks. Serial communication via the main data path (input from data logger and output to VAX) is fully interrupt-driven. Such a design was possible by using an asynchronous communication adaptor (ACIA) chip in an extension box attached to the HX20, whereas input from the VAX is polled. There is also a possibility to output messages to the data logger for the purpose of dynamic reprogramming. If, for instance, task (d) anticipates an overflow condition and finds the tape cartridge to be full, the data logger could be automatically reprogrammed to drop useful, but redundant information and restrict its output stream to absolutely neccessary data. As this facility has not been implemented in the actual software version, the corresponding line in Fig.5 has been drawn dashed. Since the five tasks partly run in parallel and there is no system program ming language available for the HX20, the HX20-software has been written in assembler. Dat. Log. Ramota Control Input Data logear Procaaaor RS 232 » Opatator, Suffar Tapa Suparviaian Managamant VAX Sava Out Fis.5: DATA FLOW AMONG PREBUFFER-RESIDENT TASKS 2-50 2.2. VAX Software The VAX, on the other hand, also executes software, matching to the VAX Interface task on the HX20 (or PC). This software comprises three modules: a) the Controller Module, which holds the code of an always resident controller task. This task is responsible for calling one or more acquisition tasks at fixed time intervals or upon request from the operating system. b) the Acquisition Module, holding the code for all acquisition tasks, of which each serves an individual buffer subprocessor (HX20 or PC) . As the present configuration contains only one HX20, only one acquisition task has been implemented. c) the Sorter Module, which provides the code for all sorter tasks. Acquisition tasks and sorter tasks are organized in pairs, in a way that each acqui sition task activates an individual sorter task on exit. As there is only one acquisition task, only one sorter task can be active in the present system configuration. The activation does not take place unless the acquisition task has been informed by the assigned buffer subprocessor that there are no outstanding data on tape (also see buffer subprocessor task (d)). The sorter task reads a workfile provided by the acquisition task, sorts the records contained there into a time-ordered sequence using their leader frames (date/time frames) as key, puts them into daily report files, and finally deletes the workfile. If there were no accesses to tape, the workfile is time-ordered. On the other hand, Fig.6 gives an example of 3 First tr«n«m. cycle completed: five data stream sections in the Suiter empty, block2 read-oack started workfile of which two have been saved 2 out to the tape stack. Since the section limits depend on the time when save-out or transmission to the 4 VAX are started, the section defini ——— ^ Second transm. cycle completed! tion is entirely random. Suffer empty, blockt read-beck started The workfile has to be preserved till the end of the last read-back and 1 then sorted section by section or- as section lengths are undefined- 5 record by record. Third transm. cycle completed: The Sorter Module uses the subroutine Suffer empty, start sorting calls of the VAX/VMS Sort-Merge Utility, which exhibits a good Fi6.6: EXAMPLE OF TAPE- behavior at this particular sorting READ-BACK ORDER problem. Fig.7 gives a concise formal description of the VAX task interaction, using the real-time language Occam /!/. Although this language has been developed for concurrent programming on transputer systems, it is a good tool to describe software that extensively relies on VAX/VMS low level system calls. 2-51 An alternative description language could be seen in PEARL ('Process and Experiment Automation Real-Time Language'), as it is defined in the papers by the 'Kernforschungszentrum Karlsruhe' /2/. Among the VAX/VMS system calls used, there are process hibernation, scheduled wakeup, wakeup enforced by a parallel process, as well as event flag services which are used for locking the acquisition task, in order to grant the operator exclusive access to the workfile for testing purposes. Except for the Sorter Module, which has been written in Fortran, all VAX software has been programmed in VAX Macro-Assembler. ^^ PAR WEILE TRUE SEQ PAR SEQ ... Acquisition Task 1 IF no_outstandingjiata ... Sorter Task 1 TRUE SKIP ... Other Acquisition / Sorter Pairs (SEQs within PAR) ... Compute next_time ( = schedule wakeup) ALT (TIMER ? AFTER next_time) â notJLocked SKIP (forcedjpakeup ? any) S notJLocked SKIP — End of infinite-loop body (outermost sequencial construct) ... Other jobs under VAX/VMS (write to channel forced_wakeup) \ / Fis.7: VAX TASK INTERACTION (FUNCTION OF THE CONTROLLER TASK) In Fig.7 SEQ denotes a process comprising any number of subprocesses (next indentation level in the list) to be executed sequentially. In a similar way, PAR indicates a process consisting of two or more parallel (concurrent) sub- processes of which the slowest is responsible for the termination of the PAR. An ALT construct specifies two or more alternative processes which can be triggered by individual timer, interprocess-communication or real-time events masked by static conditions ('&' operator). Only the alternative triggered first of all is executed; its termination is that of the entire ALT. SKIP is a dummy process used in constructs where only communication or synchronisation is of interest. 2-52 3. Fault Tolerance Precautions In the context of this paper the term 'Fault Tolerance' expresses that VAX system crashes and transmission errors are trapped. No means has been provided to meet hardware faults in the data logger or prebuffer system, because this could not be accomplished except by extensive hardware redundancy. In transmitting data from the microcomputer (prebuffer) to the VAX, a record is the smallest unit that can be handled by the protocol. Whenever a line break occurs during transmission, both sides respond with a time-out and the last, only partially received record is discarded in the VAX. Hence the buffer management task in the microcomputer must use a double- pointer algorithm. This algorithm distinguishes character-level pointers with names ending in 'PNTR' and reeord-level pointers with names ending in 'REC. A transmission pointer (TPNTR) always holds the address of the next byte to be sent to the line. This definition implies that TPNTR is incremented after every byte transmission, as is usual in system program ming. A transmission record Wait for and pointer (TREC) is over racatv.acho written by the current (with timaout TPNTR value whenever TPNTR and arror chack) points to the first byte of a record and the preceding end-of-record mark - which has been used to identify the record begin - has been succesfully transmitted-. Hence TREC never can point at other locations than record begins and always is updated record by record. When aborting the trans mission, TPNTR is set back to the current TREC value, so the next communication cycle will start with the begin of the record which Fis.8: PREPARING TRANSMISSION OF ONE CHARACTER was affected by the error. There are two possible reasons for a communication abort, closely related to the error-checking mechanism used. This mechanism is based on the echo returned by the VAX terminal driver whenever a character is fetched by the VAX CPU. Such an echo-checking protocol requires a full-duplex line, but has the advantage that no software overhead in the VAX is caused as is by protocols like XModem or Kermit. A communication abort by the microcomputer may be caused by an echo time out or by an accumulation of echo errors. In case that an erroneous echo 2-53 arrives a limited number of retries is done, making the VAX discard the record torso by sending it a backspace character and then restarting at TREC position. If the retry limit is exceeded or the backspace is not echoed properly, communication is shut down. As some time is required for a character to travel to the VAX and back, at least three unechoed characters must be on the way if the line shall not be slowed down. (See /3/ for a detailed description.) For this reason the VAX interface task in the microcomputer allows a credit of five characters, except if an end-of-record mark or the first character after an end-of-record or backspace (i.e. the leading character of a record) is to be sent. The first exception ensures that the record has arrived properly before the VAX is informed to accept it definitely. The second exception does not allow to continue with a new record as long as it is not sure that the old one is on disk. Taking this into account and regarding the TREC update condition explained above, TREC obviously must be set equal to TPNTR when an end-of-record echo is received (see flowchart in Fig. 8). The double pointer philosophy is also CXaractar "\ used at the leading buffer end, which is C controlled by the input processor task. Here the pointer pair is named RPNTR and RREC. RREC is overwritten by RPNTR when Stor« to ever the last character written to the bulf« buffer is an end-of-record mark (Fig.9). Transmission to the VAX always stops at RREC, so records just being received by the data logger remain unaffected. In the microcomputer-resident Buffer Supervision / Tape Access Control task (task (d) in chapter 2.1.) a pointer named SREC is used to cut off the save- out block, whereas SPNTR runs through the buffer during tape operations. How ever, save-out is done on a predecrement Fi6.9: READING ONE base whereas read-back operations keep CHARACTER to the post-increment convention. It should also be mentioned that every transmission burst (from the pre- buffer to the VAX) is preceded by a password exchange, so it is difficult to interrupt the line and extract data from the prebuffer by simulating a VAX request. The VAX always begins the exchange by sending a password to the micro computer. The microcomputer then answers with another password, which is not echoed by the VAX, and starts the transmission protocol. However, since it is easy to monitor a serial line, this is a facility for avoiding accidental data losses rather than a protection against 'hackers'. In endangered areas glass fiber cables should be used, as they are hard to tap. Assemblies with optical transmitters and receivers integrated in RS232 connector cases, which are available from various manufactures, allow data rates up to lOOkBaud over distances of 40m and more with excellent reliability at a cheap price. 2-54 4. Experiences and Conclusion By now, the Documentation System has been operational for about ten months, collecting a daily average of 30kByte of data, and overcome a number of VAX system crashes, maintenance days and VAX/VMS updates. During this time various software extensions have been developed for pro cessing the daily reports- Among these extensions there are programs for convenient tape backup, procedures for formatted printer output, and other. It is especially helpful to use commercial personal computer software for further data processing, in particular when PCs are connected in a network with the VAX or, on the other hand, emulate VAX terminals. There is a vast number of spread-sheet or database programs with excellent computational and graphic support available for PCs. Most of them allow data import, though format conversion programs will be neccessary in most cases. Fig.10 gives an example where nuclear reactor power and radiation level at the reactor pool surface (the latter measured by an aerosol detector) are set against each other graphically, using the program MS-Chart by Microsoft. The measurement time extended over ten days, including a week-end. The picture is based on data collected by the Documentation System. Aerosol Monitor 14- 12-- \ \ \ \ -1 ' 3 Aerosol Monitor mR/h, 100kW • Reactor Power ite VIIIH! WED THU FR\ SAT SUN MON TUE WED THU FRI • Day Fi6.10: AEROSOL ACTIVITY AT POOL SURFACE (WED; JUN 15, 1988 - FRI; JUN 24, 1988) References: /!/ 'OCCAM Reference Manual', INMOS Ltd. /2/ 'Full PEARL Language Description1, Ges. für Kernforschung m.b.H. Karlsruhe, PDV-Report KFK-PDV 130 /3/ F.Kasparec, 'Betriebsdaten-Dokumentationssystem für For schungsreaktoren' (Diploma Thesis, Tech. Univ. Vienna) /4/ J.Hammer, 'Computer Controlled Area Surveillance System for the TRIGA Mark II Reactor Vienna', Acta Physica Hungarica 59 (1985) 2-55 SESSION III Fuel & Fuel Management REACTIVITY EFFECTS IN MIXED TRIGA CORES M. Ravnik, I.Mele University of Ljubljana "J. Stefan" Institute Yugoslavia September 14, 1988 Abstract Temperature effect on reactivity is treated for standard, FLIP and LEU fuels. Large differences are observed, making predictions of mixed core reactivity effect very difficult and not reliable without explicit reactor calculations. Their accuracy depends primarily on the effective few group cross section calculations. WTMSD-4 programme with TRIGA extended library is tested for these purposes. It is con cluded that it can be used for calculation of all important reactivity effects of TRIGA fuel. 1 Introduction Experience shows /l/ that it is relatively difficult to predict reactivity effects of heterogeneous cores. This is esspecially true for mixed cores due to large differences in physical properties of fuel elements. Reactivity effects of mixed core do not depend only on properties of different types of fuel elements used in core, but also on their number and loading pattern. Reliable predictions can be obtained only by performing detailed reactor calculations. One of the most important reactivity effects for practical operation is treated in this paper, i.e., fuel temperature coefficient. It is presented separately for standard, FLIP and LEU fuel elements and as it appears in typical mixed cores for normal operating conditions of TRIGA Mark II. Power interval 0-400 kW is considered. It is calculated using WTMSD-4 with extended library /2/ and TRIGAP /3/ codes. Comparison to references and measurements is preformed in order to verify the applicability of both codes for reactivity calculations. WTMSD-4 code is used here for calculating reactivity effects in unit-cell approx imation. Detailed description of the code is given in ref /2/. We use it for TRIGA calculations without modifications in 18 group unit-cell approximation with critical buckling. Original 69 group library was extended. Isotopies specific for TRIGA fuel 3-1 were added (Er-166, Er-167, Sm, H in Zr-H). Among them scattering cross-section of hydrogen in zyrconium hydride is esspeciaUy important for fuel temperature coef ficient. It was generated from ENDFB-3 (scattering matrices) and ENDFB-5 (total scattering cross-section) data files, using programme FLANGE, as reported in /4/. WIMS was used also to generate effective two-group cross-section library for diffusion calculations included in THIGAP program. Temperature dependence of cross sections is included in simple linear form (but dependent also on burn-up), so that main temperature reactivity effects can be calculated for the whole reactor. Intercomparison with WIMS results shows that in spite of simple approximation temperature reactivity effects are accurately modelled for conditions of normal op eration (fuel temperature range 20°C - 300°C). Detailed description of formalism used in TEIGAP for reactivity by effects has been presented at previous TRIGA conference or can be found in the program manual /3/. 2 Fuel temperature coefficient Fuel temperature effect is main contribution to power deffect in TRIG A reactor. Experiment and calculations show, that it contributes ~ 95 % , the rest are effects of water temperature or density changes and reactor thermal expansion. The same is valid for fuel temperature coefficient at normal conditions. Physical mechanism of temperature reactivity effect in ZrH is well known and explained, e.g. in ref. /5/. It consist of Doppler and thermal spectrum hardening effects. Since Doppler effect is similar as in other reactors, only brief discussion on spectrum hardening effect is presented, relevant for reactor calculations. Spectrum hardening is induced by increased temperature of moderating and thermalising medium. In addition to normal shift of thermal (almost Maxwellian) spectrum, spectrum hardering is significantly affected by specific resonance structure of thermal scattering cross-section of hydrogen in Zr lattice. Up-and down scattering cross-sections are strongly dependent on temperature. For typical operating temper atures, increasing of fuel temperature means increasing of up-scattering probability in upper part of thermal spectrum, causing its deformation and hardening. Hardening of spectrum is stronger in fuel element than in water due to small diffusion length of thermal neutrons in comparison to fuel element radius. If tem perature is increased, thermal utilisation factor (f) is decreased because relatively more reactions take place in water. Relative absorbtion in water is increased and reactivity is decreased. From aspect of reactor calculations, fuel temperature effect requires detailed unit cell modelling. Flux distribution in fuel and water must be calculated in multigroup approximation with several thermal groups. Microscopic thermal scattering cross- section of H in Zr-H must be tabulated as a function of temperature. Influence of leakagae on spectrum must be considered. Described effects occur in thermal and epithermal part of spectrum and con tribute ~ 70 % to fuel temperature coefficient in standard fuel. The rest contributes Doppler effect. In FLIP and LEU, temperature induced spectrum hardening has even stronger influence on reactivity, because of the spectrum shift towards low en ergy resonances of Er-167. This can be observed from Fig. 1, where fuel temperature 3-2 coefficient is presented for standard, FLIP and LEU unit cell calailated at ~ 70°C. o FLIP A STANDARD • LEU o o £ u a. Of 80 100 e IV.J Figure 1: Fuel temperature coefficient of standard, FLIP and LEU fuel at 73° C as a function of enrichement. It is calculated as a function of enrichement. Enrichement is varied to explain differences between temperature coefficients of different fuel types. As standard and FLIP fuel elements are identical in all respects except enrichement and erbium con tent (standard: no erbium, FLIP: 1,5 w % erbium), temperature coefficient of TO % enriched standard fuel would be the same as of FLIP, if erbium would not be important. However, it is evident that burnable poison significantly increases fuel temperature coefficient in FLIP in comparison to standard fuel. For LEU this com parison can not be directly performed since it contains more uranium then standard and much less (~ .5 w %) erbium than FLIP. 3-3 Calculated values presented on Fig 1 well agree with references for fresh fuel elements in this temperature range: 8.5 pcm/°C for standard, 3.0 pcm/°C for FLIP and 5.5 pcm/°C for LEU, /5/, /6/. The same is true for depleted elements, as can be seen from Fig. 2. Ol ^3 o FLIP A STANDARD • LEU 10 20 30 40 50 BU [ °/. 1 Figure 2: Fuel temperature coefficient of standard, FLIP and LEU fuel as a function of burn-up For standard fuel the absolute value of temperature coefficient is monotonously increasing with burn-up. This is in agreement with results on Fig. 1, since burn ing of fuel is approximately equivalent to reducing enrichment. In case of poisoned fuel, the absolute value of temperature coefficient is first decreasing, because de pletion of erbium is stronger than depletion of uranium 235, and then increasing with approximately the same rate as in standard fuel, after the burnable poison has vanished. Increasing of fuel temperature coefficient with increasing burn-up and shape of the curves on Fig. 1 (approximately 1/e) suggest, that temperature coefficient otf strongly depends on effective absorption cross section of fuel, £„. It can be observed that aj is inversely proportional to enrichment or uranium content, which are proportional to £a. Simple derivation «/*3P* if8Y,a 3-4 really shows, that a/ is in first approximation inversely proportional to ab sorption, if it is assumed, that temperature induced change in absorption ££aj is proportional to temperature change 6T. This explains smaller value of coefficient for LEU fuel than expected from its enrichment (20 %, same as for standard). £a of LEU is normally very high due to its high uranium content. Fuel temperature coefficient is also temperature dependent, as can be seen from Table 1. Table 1: Fuel temperature reactivity coefficient in dependence of temperature for fresh standard, FLIP and LEU fuel, mean value over temperature interval is pre sented 0 Temp.interval [°K] -af[pcm/ K] standard FLIP LEU 296 - 400 8.7 3.9 5.7 400 - 600 10.9 7.3 7.7 600 - 1000 10.5 14.1 11.6 Values are calculated with WIMSD-4 code with extended library. Poten tial user should be aware that only average values over the whole temperature interval of hydrogen cross-section tabulation have physical meaning. Discon tinuities of coefficient on interval boundaries are observed, suggesting that only linear temperature interpolation of scattering matrices is considered in the code. However, average values for broad intervals agree well with references /5/, /6/. Maximal discrepances do not exceed ± 1 pcm j°C, which gives confidence in results of calculations with WIMS. 3 Mixed core fuel temperature coefficient It was shown that fuel temperature coefficient depends on fuel element type, temperature, bum-up and enrichement. Implicitely it depends also on neu tron leakage from reactor, which effectively appears as increased absorption in the system. Fuel temperature coefficient of a realistic mixed core depends therefore on size of the core and its loading pattern . It can not be calcu lated with pin-cell code (like WIMS) only. Global reactor calculations must be performed. It is also very difficult to compare calculated reactor fuel temperature coefficient to unit cell calculations or measurements, because average fuel temperture of the core can not be calculated or measured precisely . If discrepancy is observed, it can be attributed either to calculational error of 3-5 temperature or reactivity. Usually it is preferred to compare temperature or power defects and not coefficients. Due to uncertainty of temperature determination, it is also reasonable to use such empirical correlations between power and fuel temperature, that give appropriate values for power reactivity defect. Fuel temperature reactivity deffects of two hyphothetic mixed cores are presented in Table 2. together with values for relevant homogeneous cores. They are calculated with TRIGAP program in two-group diffusion approxi mation. Two group cross-sections are generated by WIMS. The cores consist of 66 fuel elements and operate at the same power (400 kW) in all cases. Homogeneous cores consist of standard or FLIP fuel. Mixed cores consist of 33 FLIP and 33 standard elements. In case A FLIP-s are loaded in inner rings and in case B in outer. Results for mixed cores show, that temperature effect coefficient is not proportional to number of fuel elements of each type. It depends also on power and temperature distribution. If power and temperature would be uniformly distibuted over the core, mixed core power defect would be the same in both cases and approximately equal to mean value of homogenous cores (1264 pcm). However, in case of FLIP fuel in the centre, power defect is ~ 200 pcm closer to FLIP core value. Table 2: Power reactivity defect of homogeneous and mixed cores operating at 400 kW Core k, P = 0 k, P = 400 kW Ap [pcm] 66 STANDARD 1.01617 .99876 1716 66 FLIP 1.02410 1.01566 812 33 FL. + 33 ST. (A) 1.01237 1.00169 1039 33 ST. + 33 FL. (B) 1.02860 1.01541 1263 4 Conclusions Calculation of temperature reactivity coefficient of mixed core requires re liable unit cell and diffusion calculations. WIMSD-4 with extended library and TRIGAP programs are used and tested for these purposes. Their results proved to be consistent and in agreement with references. It is shown that accuracy depends strongly on physical and calculational model, esspecially for mixed cores. Testing and upgrading of both codes will be continued to make reliable calculations of reactivity within reach also to research reactor centres, equiped with small computers. 3-6 References [1] M.Ravnik, I.Mele, S.Slavic Calculational Analysis of TRIGA Mark II Reactor Core Composed of Two Types of Fuel Elements IAEA Progress Report, RB 3512/Rb, IJS-DP-3479, April 1984 [2] K.Kowalska The S-WIMS Code for CYBER-72, NEA CPL, Saclay, France, 1976 and the WIMSD-4 Program Manual Ibid., 1983 [3] I.Mele, M.Ravnik TRIGAP - A Computer Programme Package for Research Reactor Cal culations IJS-DP-4238, Dec. 1985, NEA CPL, Saclay , Jan. 1986 [4] A.Trkov, M.Ravnik Cirkonijev hidrid v WIMS knjiznici grupnih konstant XXVII konferenca ETAN, Struga 1983, V. zvezek, 473-479 [5] Research reactor core conversion from the use of highly eriched uranium to the use of low enriched uranium fuels guidebook, IAEA-TECDOC-233 [6] Safety Analysis Report for the NSCR Texas A-M University June 1979 •3-7 THE EFFECTS OF SS-304 CLADDING- ON CORE CALCULATION OF ISTANBUL TRIGA MARK-II REACTOR Süleyman Güngör Çukurova Üniversitesi Fizik Bölümü Balcali-Adana TURKEY 10th European Triga Users Conference Atominstitut Vienna/Austria 14.9.-16.9.1988 3-9 ABSTRACT According to the Safety Analysis Report of TIGA MARK II reactor, the initial core loading of about 63 fuel elements, a cold, clean excess reactivity of approximately 2.1 $. The first criticality experiment at 11 March 1979 showed that the reactor was just critical with 63 fuel elements, and had no excess reactivity which was predicted in safety report. The criticality with 2.1 $ excess reactivity was reached about 69 fuel elements. The calculations are performmed with one-dimensional in 40-energy groups and two-dimensional in 5-energy groups. It is found that the excess reactivity of about 3 $ is calculated with 69 fuel elements, by using SS-304 cladding in the fuel elements. 3-10 1.GENERAL DESCRIPTION The TRIGA MARK II reactor core assembly is located near the bottom of a cylindirical aluminum tank surrounded by a reinforced concrete shield structure. A typical MARK II installation is shown in Fig. 1. The standard reactor tank is welded aluminum vessel with 1/4 in-thick walls, a diameter of approximately 6-1/2 ft, and a depth of approximately 21 ft. The tank is all-welded for water tightness. An aluminum angle used for mounting the ion chambers, fuel storoge racks, and other equipment,is located around the top of the tank. The core is shielded radially by a minimum of 7 ft 9 in. of ordinary concrete, 1-1/2 ft. of water and 12 in. of graphite reflector. The overall core diagram of TUR-TRIGA- MARK II is shown in Fig.2. It contains 69 fuel, 16 dummy graphite and one instru mented fuel elements. The reactor is equipped with a central thimble for conducting experiments or irradiating small samples in the core. It consists of a 1-1/2 in. o.d. it is normally filled with water. 2. FUEL - MODERATOR AND DUMMY ELEMENTS The active part of each fuel-moderator element, shown in fig.3 is approximately 1.43 in. in diameter and 15 in. long. The fuel is a solid, homogeneous mixture of Uranium-Zircouium hydride alloy containing about 8-1/2 % by weight or uranium enriched to 20 % U-235. The hydrogen -to- Zirconium atom ratio 3-11 is approximately 1.6. Each element is clad with a 0.02 in. thick stainless steel can. In order to serve as top and bottom reflector, graphites are inserted to the top and bottom of fuel, the total length of the fuel element is 28.8 in. The total weight of a fully loa ded fuel element is about 7 lb. Table.1 gives a summary of the fuel element specifications. Graphite dummy elements may be used to fill grid positions not filled by the fuel-moderator elemnts or other core compo nents. They are of the same general dimensions and construc tion as the fuel elements, but are filled entirely with grap hite and are clad with aluminum. 3.CALCULATI0NAL METHODS It can be seen from the fig.2 that the reactor is nearly axial symmetric. A one-dimensional or 2-dimensional R-Z confi guration will be sufficient to calculate the whole core. To generate group constants which are used to solve multigroup diffusion theory codes, two steps are performed. At the first step GGC-4(1) which is a 200 group (out of which 99 are fast and 101 are thermal) spectrum calculation code is used. A fuel element is thus assumed to be removed from fig.2, with its water neighborhood of 1 cm of thickness. The material present in the cylindirical geometry is homogenized and flux energy dependency calculation is performed. Fourty group cross sections are thus found by collapsing the 200 group cross section of GGC-4 via the use of the estab lished flux energy dependency of the reactor. 3-12 At the second step ANISN(2) which is a one-dimensional multigroup S-N transport code is used to compute 5-group, the space wise averaged equivalent cell, and the related isotopes cross sections based on the GGC-4 output, within the actual heterogeneous structure of the cell fig.4. 4.ONE AND TWO DIMENSIONAL CORE CALCULATIONS One-dimensional core calculation is undertaken (for a core of 69 fuel elements) with the one-dimensional ANISN in 40 group using the configuration as shown in fig.5. According to the configuration,the first ring is water which is the central thimble. The second region is made of an homogeneous material of fuel, clad, water containing 57 fuel elements and 3 control rod guide tubes which are full of water. The third region is made of 12 fuel elements and 16 graphite dummy elements. The fourth region is only water and the last fifth region is made of graphite reflector. The eigenvalue k _f came out of one-dimensional calculation to be : kgff(l-D)= 1.0222221 The same calculation is repeated using two-dimensional code GE'REBUS(3) by adding Z-dimension, without changing the configuration in R-dimension. In Z-dimension to the top of reactor water an graphite reflectors are added, fig.6. The calculation is performmed in 5-groups which the 5-group cross sections are collapsed by ANISN from 40-groups.From a corres ponding two-dimensional calculation the k ._ result was: erf keff(2-D)= 1.0192362 The 5-group cross sections for some isotopes which are used in the 2-D calculation are given in table.2. 3-13 CONCLUSION The k ,.,. values which are obtained from the 1-D and 2-D ef f calculations have nearly no differences. If we compare these values with the value which were predicted in the Safety Analysis Report of ITÜ-TRIGA Mark II reactor(4) for the initial core loading, there is nearly no differences. The only difference is in between the number of fuel elements. The calculations are performed with 69 fuel'elements but in the safety report this number is 63 fuel elements. The first criticality experiment at 11 march 1979 showed that the reactor was just critical with 63 fuel elements, there was no excess reactivity, and 2.1$ excess reactivity was reached about 69 fuel elements. The differences between the present calculations and res"ul£ of safety report is probably coming from the cladding which are used. Up to now there is no information why these difference are occured. It is clear that if aluminum is used as clad the ^eff ^or ^3 fuel elements will have this excess reactivity. 3-14 References 1. J.Adir and D.Lathrop; GGC-4 Multigroup Cross Section Code GA-9021 2. W.W.Engle, Jr. ANISN, A One-Dimensional Discrite Ordinates Transport Code With Anisotropic Scattering. K-1693 1976 3. GEREBUS, GKSS version of the Multigroup Diffusion Depla- tion programme in 2-Dimension. FN-E-88 (FIAT 1967) 4. Safety Analysis Report for the TRIGA MARK-II Reactor, August 1978. 3-15 Figure.1. Typical Mark II Reactor Pulsing Detector Safety ° O' 0 channel ' Jßrs /SOURCE\ VE Î5 J/*~~\ w so o-o •CO Wide o Range Oetector SatetY Channel EU ELEMENT - U 2 iri) GRAPHITE ELEMEHT CONTROL R00 SOURCE INSTRUMENTEE! FUEL ELEMENT Figure. 2. TUR TRIGA MARK-II Core Diagram 3-17 STAINLESS STEEL TOP END FITTING GRAPHITE 3.U5 IN. STAINLESS STEEL TUBE CLAOOING THICKNESS 0.02 IN. URAN IUM ZIRCONIUM HYORIOE- WITH AXIAL Z I RCONIUM ROO. 28.8 IN. 15 IN. TOTAL 1.1*3 IN. •1 .It7 IN. 9 GRAPHITE 3.<*5 IN. STAINLESS STEEL BOTTOM END FITTING Figure. 3. Triga stainless steel clad fuel element with triflut end fittings 3-18 Figure.4. Fuel cell structure for an ANISN calculation 3-19 r^ 2.208 cm r2=l785 cm r3= 21793 cm r^a 24.689 cm rr=54.51 cm Figure.5. One-dimensional core configuration 3-20 z Water 7.62 cm Graphite 16.38 cm Homogenized fuel + mod. £ £ C9 a a o 35-43 cm £ £ £ £ £ »- r o u u o CO LD (NI en 00 en CO O-' c-^ Figure.6. Two-dimensional core configuration 3-21 Fuel-Moderator Material H/Zr ratio 1.6 Uranium content 8.5 wt % Enrichment (U-235) 20 % Diameter 1.43 in. Length 15 in. Graphite End Reflector Porosity 20 % Diameter 1.43 in. Length 3.44 in. Cladding Material Type 304 SS Wall thickness 0.020 in. Length 22.10 in. Table.1. Summary of Fuel Element Specifications 3-22 1 .group 2.group 3 .group 4.group 5.group CT :L.321E00 2.288E00 3.787E00 5.710E01 4.595E02 a 4.760E00 9.390E00 4.954E01 7.199E01 4.706E02 tr 1.065E00 1.152E-2 1.324E-2 0.0 Ü-235 4.077E-1 2.791E-1 7.963E00 5.188E-1 1.920E00 a 4.951E00 9.810E00 3.616E01 1.122E01 1.260E01 tr 1.630E00 1.407E-2 8.668E-3 0.0 U-238 6.113E-3 1.144E-2 1.302E-1 2.877E-2 1.307E-1 a Zr cr 0.0 8.348E-5 8.865E-3 5.163E-2 2.346E-1 3 °"tr 1.007E00 3.687E00 6.736E00 8.362E00 2.862E01 Table.2. cross sections for some isotopes 3-23 GAMMA SPECTROMETRICAL EXAMINATION OF IRRADIATED FUEL Edvard KriStof and Gvido Pregl* Institut "Jo2ef Stefan", Jamova 39, Ljubljana *VTS - Univerza v Mariboru, Smetanova 17, Ljubljana Gamma scanning is the only nondestructive technique for quantitative measuring of fission or activation products in spent fuel. Data, obtained from such measurements serve for studying the migration of fission products, for characterization the performance of the fuel elements during irradiation, etc. On account of the attenuation of gamma rays in matter, merely reliable gamma spectrometric determination of the desired amount is obtained from a measurement of the relevant isotopic density in the fuel. With the methods resembling those used in the medical tomography, these problems have been worked on in the field of nuclear technology. However, the negligence of local variation of the linear attenuation coefficient of gamma rays in the irradiated fuel remains the main source of systematic error. To eliminate it we combine the (single) emission gamma ray scanning technique with a transmission measurement. Mathematical procedure joined with the experiment is particularly convenient for fuel elements of circular cross-section. In such a manner good results are obtainabale even for relatively small number of measuring data. Accomplished routines enable to esteem the finite width of the collimation slit. Owing to the great number of gamma ray spectra measurements, the experiment has been partially automated. Trial measurements were carried out, and the measured data were successfully processed. This research work is auspieced by the International Atomic Energy Agency. It is liable to the Agency Research Contract No. 2997/RB. 3-25 2. The experimental set-up For measurement of gamma-ray spectra a high resolution Ge Fig.l Horizontal section through the experimental set-up. Shielding is of lead and heavy concrete. The effective width of the gap on the detector side of the collimator C for the 662-keV line is 0.07 cm. Rotation of the lead shutter S by 90° screens the additional.fuel element A. The iron slab IS between precollimator PC and detector serves for suppression of the continuous background and for attenuation of the direct beam. E denotes the examined fuel element. 3-26 experiment. The examined fuel element may be rotated and moved normally to the axis connecting the detector and additional fixed source (Fig. 2). It hangs on a torsional wire fastened to the vertical axis of the rotator, which is a part of the moving device. The lower end of the examined fuel element dips into a vessel filled with water. In such a way small oscillations appearing when the measured object is moving are damped out. TJPB ^<<,V7X\V/AvyA\y/x>x/,<\y/x\y/AV/, Fig. 2 Vertical section through the experimenhtal set-up. Here M denotes the moving device with rotator R, W the torsional wire, G the collimation gap, S the scraper, E the examined fuel element, PB lead bricks, and V a vessel filled with water. Owing to the great number of gamma ray spectra measurements the experiment has been partially automated. We worked out a group of congruent routines, which starts and stops the partial measurements, transfers the measured spectra into files, and governes the moving device5,6. The spectral data processing is based on ORTEC's analysis software package. 3-27 3. Efficiency of experimental set-up Due to highly active spent fuel, the efficiency of an apparatus such as ower must be very low (^ 40 ^.Therefore standard gamma sources are not suitable for calibration purposes. For that reason we activated a rectangular silver plate with dimensions 1,12 cm x 2,18 cm in a homogenous beam of thermal neutrons. The activity of thus obtained source was determined by gamma-spectrometry. Because the activity is too high to be measured directly, are weakened the radiation during the data acquisition. Theree measurements using iron slabs a, b, and a + b with thicknesses 6,0 cm, 5,2 cm and 11,2 cm respectivelly, were performed. Each time the virtual activity Ax was determined. The result is A - A^/A - 147,12 f*.C/cm2 (18th June 1986 at 00:00). In the final step the flat silver source was placed in the rear of the collimator slit. In absence of attenuation plates acquisition of 1122600 integrated live timeseconds of spectral data was -performed over an extended period of time. With results from these measurements we obtain X (gHUEn A^lr 1,^ = W* (E) "I RlH where is the counting rate in the photopeak belonging to the energy E, if the superficial activity of this line equals unity (1 gamma decay cm-2 s'1) is thickness of iron plate used during the measurement with the fuel element 2 log^a - - 48,2692 + 10,3003 y - 0,708539 y - - 36,6369 + 7,11506 y - 0,490530 y2 - - 67,7625 + 14,6454 y - 0,935882 y2 logvD a + b = - 0,158815 - 3,60863 y + 0,216113 y2 y » log E, E in (keV 1 3-28 The root mean square of the uncertainty in the energy range from 600 keV to 1600 keV is 0,4 %. If we use the iron plate b (x • 5,2 cm), the efficiency for 662 - keV line belonging to the isotope :37Cs is 2,652.10"8. 4. Field of examination Counting rate and activity on the observed cross-section It is known that an object can be uniguelly reconstucted from an infinite set of projections. In practice, the number of data measured is limited and their value is of a finite accuracy. Consequently there does not exist a single way which would be able to satisfy all geometrical and precision requirements. During the measurement the fuel element is moved in the horizontal plane. As the broadening of the slice in the vertical direction is negligible, we have in essence a two-dimensional image reconstruction. Pig. 3 Field of examination 3-29 We treat the field of examination as a circle of unit radius in which the points are given with polar coordinates P and \^ belonging to the body system of the object of examination (Fig. 3). The origins of the coordinate systems x, y and p , *P coincide. The abscissa x represents the distance between the centre and the fixed axis y' connecting the detector with the additional source, and the angle represents the rotation of the system 0 ,\0 with regard to the system x,y. ' An array of pixels is formed by dividing the unit circle into K concentric rings with the same width 1/K (Fig. 4). These rings are composed from particular area elements (pixels). Areas of all pixels are equal. Inside the single area element the linear attenuation coefficient of gamma rays and the activity are treated as constants. Fig. 4 Array of pixels 3-30 In the case of a sufficiently fine division of the field of the examination, the simplification of a thin slice breaks down. For this reason we have developed the model which takes into account a real transpassing response u( ç parameters). This function is proportional ^to the counting rate when a point source is positioned at © in the body system while parameters are kept constant. Because the broadening of the slice in our field of examination is negligible in comparison with its width, we treat the transpassing response as a function of only one variable x and parameter p (Fig. 5). 1 u(x-p) j u(x-p)dx = ! 0 i 0 A X Fig.5 Transpassing respose u(x-p). Thy symbol p represents the distance between the origins of the laboratory system x',Y' and of the body system During the measurement counting rates are being determined for chosen distances p as a function of the rotation o< . The number of measured convolutions is determined in such a way that to each position pk belongs a certain number Nk spectra measured with the additional source both screened and unscreened The t h measurement is done at «C (i-1)' 2 11/N. . We avoid here a derivation and only give the result of discretization: W nt(MA-a(V)= |^M- a ) and 3 Jv;C«0 «?«)£;,• £10 &V (40 Fl« 3-31 where R (k) is the length of the prime (I,II)m at x = xm within (i + l)th pixel in kth ring (Fig.6) is the element of the selfabsorption th matrix which corresponds to Rlm I i •! ) pixel in k ring ndices m Fig. 6 Illustration of denotations u. and R. (k) ATI 3-32 •'i**-* A KM) is the activity of the observed gamma line at the P°int * /^(*f—0 / *? S(<<°<,£y|f) is the attenuation of gamma ray on the straight line between points . ,^ and — — — ,Ç cos (3 - <<) cos (* -«< ) A/V, is the linear attenuation coefficient K(X,-É ,3 ,ç ) - s(x,-c ,5 ,^r ) Dy (x»3 -•<) / ^5 gt(k) - f (k) h (k) p Here, the measuring data are the counting rates f in the absence of the additional source, g in the presence of the additional source, and Po for the unscreened additional source while the examined one is being shifted away. Matrix elements Rx. are wholly geometrical quantities, and values u. are obtainable by a separate measurement. From the nonlinear system of equations (**) we estimate the linear attenuation coefficient of the observed line in the area 3-33 elements of the ring k. The next step is the determination of the matrices q, 5, Error of the measurement An average value T of the activity of the observed photoline in the field of examination can be considered as the final result of the measurement. Short derivation showes that the variance S of the average activity is given with variances and G; =I /*»* m » k, The number of numerical operations corresponding to this formula increases faster as fourth power of indices number. For this reason we make an approximation in such a way that we compute for each k only one pair of partial derivatives : 01 x 5" ! 25_ _A_ )%(l. + l«ki) + k V&fVv2'^] In this equation, t means a live time of the particular measurement, u and v are counting rates of backgrounds for both screened and unsereened measurements, respectively. 3-34 6. Measurement In recent years, whereas the method has been developed, some test measurements were performed. We describe the latest of them. The fuel element used was E 6195. It was cooled more than four years prior to the measurement. The diameter of the field of examination is 4,5 cm. It is divided into 300 pixels. The observed cross section of the fuel element lies almost concentrically within this field. The eighth ring (R - 1,80 cm) overcovers slightly the circular fuel boarder (R » 1,76 cm). The number of measured spectra for the additional source screened and unscreened was determined so that it equals in both cases to the number of pixels. Prom the evaluated gamma-ray spectra the dominant lines were separated. In Fig. 7 we present activity distribution of the 662 - keV line belonging to the isotope 137Cs. The average value of the specific activity in the observed cross section of the fuel is 4,743 (1 + (0,0040 + 0,0047)) •• 10+s s"1 cm"3. This corresponds to 7,800 (1 + 0,009) • 1013 13 7 Cs atoms per cm3. Here, the error of the calibration measurement is 0,40 %, and the statistical error of all the measurements on the fuel is 0,47 %. The «ccuracy of the calibration point sources was not taken into account. 3-35 Fig. 7 The distribution of the isotope 13 7 Cs in arbitrary- units. In the lower picture the body system is rotated for «ft-. REFERENCES 1. Jonh R.Phillips, Nuclear Technology 28 (1976) 282; 2. S.T.Hsue, Atomic Energy Review 16 (1978) 1; 3. E.Kriätof and G.Pregl, Fizika 11, Supplement 1 (1979) 117; 4. E.Kriätof and G.Pregl, Preiskava gorivnega elementa s presevanjem, IJS-DP-2191, November 1980; 5. J.Krajnik, Sotware package AVTOZ, Podgorica, August 1981; 6. B.Glumac and I.Jencic, Software package GOR, Podgorica, July 1982 7. I.Jenöic and B.zefran, Software package GORIVO, Podgorica, April 1987 8. E.Kriätof and G.Pregl, Nondestuctive Method for Assessment of Nuclear Fuel, Fizika 17 (1985) 2, 179-189 3-37 FUEL ELEMENT REPLACEMENT AND COOLING WATER RADIO ACTIVITY AT THE MUSASHI REACTOR T. Nozaki, T. Honda, N. Horiuehi, 0. Aizawa and T. Sato Atomic Energy Research Laboratory, Musashi Institute of Technology Abstract The Musashi reactor (TRIGA-II, lOOkW) has been operated without any serious troubles since 1963. In 1985 the old Al-cladded fuel elements were replaced with new stainless cladded ones in order to insure a long and safe operation. By using a semi-automatic equipment the old fuel elements have been transferred into the bulk-shielding experimental pool, which was remodelled for the spent- fuel storage. In order to reduce the exposure during the transfer work, the old fuel elements were cooled in the core tank for 3 months. After the replacement, the radioactivities in the cooling water have been drastically changed. The activity of Na-24 decreased about one decade, and the activities of Cr-51, Mn-54, Mn-56, Co-58 and Co-60 increased about two decades. At this conference we. will report on the following points : (1) semi-automatic equipment for the transportation of the Al-cladded spent fuel, (2) structure of spent-fuel storage pool, and (3) radioactivity change in the cooling water. 3-39 FUEL ELEMENT REPLACEMENT AND COOLING WATER RADIOACTIVITY AT THE MUSASHI REACTOR T.Nozaki, T.Honda, N.Horiuchi, O.Aizawa and T.Sato Atomic Energy Research Laboratory, Musashi Institute of Technology (Japan) I Introduction The Musashi Institute of Technology Research Reactor (TRIGA 11 lOOkW, the Musashi Reactor) has been operated without serious problems since 1963. The occasion arose, however, when we needed to choose one of the following three options, because there were no more spare fuel elements; (1) to obtain some new aluminium cladded fuel elements and operate with both these and the old aluminium cladded ones, (2) to obtain some stainless steel cladded standard fuel elements and operate with both those and the old aluminium cladded ones, and (3) to obtain many stainless steel cladded fuel elements and operate with only those fuel elements. Option (1) is the easiest way of obtaining operating permission from the regulating agency (the Science and Technology Agency) of the Japanese Government. Option (2) have often been adopted in the case of other old TRIGA reactors around the world, but some difficulty in negotiating permission from the regulating agency was predicted by reason of the complex fuel arrangement in the core. Option (3) is the best from the technical point of view, for long-term, trouble-free operation, though it did require some negotiation with the regulating agency and it is the most expensive way. In the end, we decided to adopt option (3) under the auspices of the Ministry of Education, Science and Culture of the Japanese Government. The bulk shielding experimental pool was remodeled as storage for the spent fuel elements, where the neutrons from the thermalizing column were shielded by cadmium and boron- polyethylene plates. Equipment for transferring the spent fuel elements was built and temporarily set up between the core tank and new storage. This work started in fall 1983 with the motion of the change of permission with the regulating agency of the Government, and finished in summer 1985 when the agency's permission to recommence operations was received. 3-40 After the. reactor was restarted, the count-rate of the conventional cooling water monitor, which was set in the cooling system using a GM counter, drastically decreased as shown in Fig.l. Therefore the radioactivity in the cooling water was precisely surveyed again, as had been done in the time of the old core [1]. In this paper, we describe the spent fuel storage, the work of transferring the spent fuel elements and the cooling water activity. II Fuel Element Replacement 2-1 Spent fuel storage Figure 2 shows a vertical sectional view of the spent fuel storage and transfer equipment which was temporarily set up. The storage is composed of a storage pool and a storage tank which is made from stainless steel and sunk at the bottom of the storage pool, supported by a floating prevention supporter. The storage tank is composed of an inner tank and an outer lid as shown in Fig.3. The spent fuel elements are lined up cylindrically in three lines in a fuel stand in the inner tank (see Photo.1). Ninety spent fuel elements can be stored in this storage, and remain in a subcritical state under all conditions. The storage also has a purification system of water and a nitrogen gas system. After the spent fuel elements were inserted in the holes in the inner tank and the lid was put on, nitrogen gas was supplied to the tank and water was excluded from the tank. Thus the spent • fuel elements are stored in a nitrogen atmosphere in the bottom of the storage pool. The purification system is shown in Fig.4. The system is composed of three subsystems: They are; (1) system for supplying purified water to the storage pool, (2) purification system for the pool water and (3) exhaust system for the pool water. If water has to be supplied to the pool, purified water is supplied from the supply system. The water is continuously circulated between the pool and the purifier and the conductivity is kept below 2uS/cm. Waste water is transferred to a waste water storage pit which has already been built for other radioactive waste water, and it is disposed of the other radioactive waste. The nitrogen gas system supplies the nitrogen gas from a gas cylinder to the storage tank and the nitrogen gas is exhausted through the stack of the reactor building when the tank is opened. 3-41 T r 14 Sep.'84 CO ( AT clad fuel ) - Q. Ü 1.0 J l 1 I. I L 0 12 3 4 5 Operating Time ( h ) Fig.l Change in count-rate of water monitor. 3-42 Fuel transfer equipment Stora£ pool (unit : cm) 2 Vertical sectional view of reactor core and spent fuel storage. —Fuel transfer equipment is assembled between core and storage— 3-43 Outer Lid / Inner Tank itr • , ¥ i|l ill '1' 1 Fuel Stand Spent fuel storage- Fig.3 Spent fuel storage and fuel stand in inner tank. Photo.1 Opened inner tank storing the fuel elements 3-44 P: Pump F: Flow Meter C: Conductivity Meter Purification G: Pressure Gauge system of pool water Purifier Water supply system Wast Water Storage Pit Storage Tank Tap Water Fig.4 Conceptional design of purification system 3-45 2-2 Equipment and work for fuel transfer The spent fuel elements were transferred after a cooling period of about three months. The radioactivity was about 370 GBq for each fuel element. This value shows that the fuel elements can be transferred easily using a relatively simple and thin shielding capsule with low personnel exposure. We, however, preferred another more reliable and safe way with which even lower personnel exposure was to be expected. A vertical view of the transfer equipment is shown in Fig.5. The spent fuels are semi-automatically transferred from the reactor tank through the reactor top to the new spent fuel storage. In Fig.5 the one movable part of the equipment is described as part of both sides, the reactor core tank and the storage, as if it were in the two parts. The transfer work was performed as follows: (1) The fuel element was manually pulled out of the core using a conventional fuel-handling tool for the TRIGA reactor, and inserted into a fuel case in the transfer equipment ((1) in Fig.5), (2) The case was put into the lower edge of expansion and contraction rod ((2) in Fig.5) by using the swing handle which is fixed at the reactor top ((3) in Fig.5), (3) After step (2) was completed, the workers left the reactor hall in order to operate the equipment remotely and avoid unexpected radiation exposure. (4) After all persons had left the reactor hall, the following operations were automatically performed by the pushing of a button; 1) the expansion and contraction rod was shortened and pulled up, then the case with the spent fuel also went up to the reactor top, 2) when the case in the expansion and contraction rod had arrived at the reactor top, the drift table which hang the rod with the case ((4) in Fig.5), was drifted to the storage side on the top, 3) when the expansion and contraction rod arrived at the storage side and the drift table stopped there, the expansion and contraction rod expanded farther and the case went down with the lower edge of the expansion and contraction rod, 4) when the case in the expansion and contraction rod arrived under the water at the center of the inner tank, the motion of the expansion and contraction rod stopped. (5) After the signals on the operation panel and TV monitor confirmed that it had stopped, the workers went into the reactor hall again. (6) The case was removed from the expansion and contraction rod by use of the swing handle ((5) in Fig.5). (7) The fuel element was manually pulled out of the case using the fuel-handling tool and inserted in a hole in the inner tank. (8) The case was again put in the expansion and contraction rod and it was returned to the reactor tank by remote control. 3-46 Fig.5 Fuel transfer equipment. 3-47 The transfer of one element took about 20 minutes. These operations were repeated 65 times in 3 days. During the work, TLD at the reactor top recorded an exposure of 3.5 R but the workers' individual monitors did not show any exposure. After the work, the equipment was removed. Ill Cooling Water Radioactivity As shown in Fig.l, the count-rate of the cooling water monitor much decreased. We, therefore, surveyed the radioactivity in the cooling water by the method used before the replacement, and the results were compared as the cases before and after the replacement. 3-1 Method of radioactivity measurement Anion or cation ionexchange resin of about 80 cm^ was used for the collection of radioactive nuclides in the cooling water. For the sampling, the cooling water was pumped up from about 2m depth below the top surface of the reactor tank at a flow rate of 13cm3/s and it was passed through an ion exchange resin column of 80 cm^ for about one hour. The collecting efficiency for each of Na-24, Te-99m, 1-132, La-140 etc. in the sample was about 80#. Hence the value of 80% was used for each nuclide. The sampling was performed after the reactor shut-down on Friday, when the radioactive concentrations would be almost at their maximum. The measurement of radioactivity was carried out using a Ge(Li) gamma-ray spectrometer. 3-2 Cooling water radioactivity with the old and new cores The change in radioactive concentration of various kinds of radioactive nuclides in the cooling water before and after the replacement is shown in Fig.6. The change from the old core to the new one is shown by the direction of the arrows. The values for the old core were measured on September 2 and 16, 1982 by use of cation and anion exchange resin, respectively, and the mean values of three cases measured between April and November, 1986 are used for the new core. There was almost no change in the values for the new core in 1987. The radioactive concentration of Na-24 decreased by about one order of magnitiude after the replacement. This change can easily be understood as being caused by the replacement, which resulted in a decrease in the amount of aluminium in the core, because the nuclide Na-24 was mainly produced by the 24 reaction 27A]_ (n ( a) Na . 3-48 Otherwise, the radioactive concentrations of such nuclides as Cr-51 Mn-54, Mn-56, Co-58 and Co-60 increased by about two order of magnitude. The new cladding material SUS 304 includes chromium (18.3%), nickel (8.7%), manganese (1.8%) and maybe also a small amount of cobalt. It is considered that the nuclides Cr-51, Mn-54, etc. mentioned above may have been produced by the following nuclear reactions: Mn-54 54Fe(n,p)54Mn, 55 56 56 56 Mn-56 Mn(n,Y) Mn, Fe(n,p) Mn1 50 51 Cr-51 Cr(n,Y) Cr, Co-58 58Ni(n,p)58Co Co-60 59Co(n.Y)S0Co 60Ni(n,p)60Co From these results, it can be concluded that the cladding material is the main source of many radioactive nuclides in the cooling water of the Musashi reactor. Fission products such as 1-132, 1-133, Te-133 etc. were also detected. The change in concentration between the old and new cores is not large. It decreased to about one-half as shown in Fig.6. A small amount of fission products could be produced from trace amounts of uranium in or on the core components, coming into contact with the cooling water in some kinds of research reactor [1,2,4]. In order to have some data, the uranium content in the aluminum cladding of the graphite dummy element, which was made at the same time as the fuel elements, was measured and determined as 0.2 ppm by the neut ron activation method [5]. Moreover, we could find little uranium content in the three kinds of SUS 304 using the fissio n track method. It is also reported that the values were 0.3 2-0.17ppm for the JMTR fuel cladding and the fuel side plates made from aluminium, and below 40 ppb for the SUS-304 of t he core materials [4]. If the uranium content in the alum inium cladding of the fuel element is 0.2ppm, the same as th at in the dummy element and that of the stainless steel cladd ing is below 40ppb, and if the contribution of the cladding to the fission products is large, a greater change in the r adioactivity should occur. The results shown in Fig.6 may suggest that the fission products were released from the o ther unchanged core material including more uranium. VI Conclusion The replacement of the fuel elements was successfully performed without any problems. The replacement has meant that there are hopes for the future of the Musashi reactor. 3-49 24 Na 10' 56 Mn 10' Old Core 9 9mT( c 187. 1 10 ! 51 New Core Cr 72 Ga 2 (> 58 1Ö Co 134. 135. I J 122 Sb 54 Mn 6 0„ 10-3L 133T ? r * < i # Co 132 « J | x 65 59 Zn I Fe r4 132Te 10 131 I1 124 5 Sb • 10' ND 8 104 105 106 107 10 Half Life ( s ) g.6 Change in the radioactive concentrations in the cooling water with replacement of the fuel elements Radionuclides (except for fission products) Fission products ND: Not detected 3-50 Now we are looking forward to upgrading to 250kW. As well, the activity in the cooling water was measured and we found a clear difference in radioactive concentration between the cores with the aluminium cladded fuel elements and the stainless steel ones. References [1] NOZAKI.T., OKAMOTO.M., : Radioactivity in Cooling Water of Musashi Reactor "TRIGA II", J.At.Energy Soc.Japan Vol.25, No.10, 816-821 (1983). [2] MATSUSHITA.R., KOYAMA,M. :The origin of Fission Products in Primary Cooling Water, Annu.Rep.Res.Reactor Ins.Kyoto Univ., 11,57-66 (1978). [3] HONDA,T., et al. : Uranium Determination in Aluminium Clad for Dummy Elements at the Musashi Reactor, Bull. Atom. Ener. Res. Lab. Musashi Ins. Tech. Vol.12, 87-91 (1987). [4] YONEZAWA.C.,et al : Determination of Uranium in Reactor Materials by Neutron Activation Analysis, J.At.Energy Vol.29, No.l, 58-63 (1987). 3-51 SESSION IV Safety Aspects, Licensing & Radiation Protection SOME ASPECTS OF LICENSING, PERIODIC INSPECTIONS AND BACKFITTING OF TRIGA REACTORS H.Weiss Technische Universität Graz, Austria ABSTRACT After a brief historical retrospective view on the early days of TRIGA reactor operation in the light of both technical and legal aspects the considerable changes over the years in reactor safety philosophy and licensing procedure will be discussed. Although an extremely successful operating record has been demonstrated by TRIGA reactors throughout almost three decades, several reactor facilities have been now in operation for more than twenty years. The question, whether to renew obsolete control and instrumentation systems or reactor components will be seen in the light of present days knowledge and technical standards. Upgrading or license renewals require or at least make new credible accident analysis highly desirable. The reason for this and the subjects to be dealt .with in this respect will be shown. Caused by changes of research Programms or other reasons the safety report and other documents may be updated occassionally. To what extent does this fact result in a renewal of the licensing procedure? Further points of discussion are practical aspects of the reinspection procedure as well as possible interactions between operation of mixed cores and licensing. Finally some aspects for future work in the above mentioned area will be discussed. 1. RETROSPECTIVE VIEW 1.1 Situation of TRIGA reactor operation at the end of fiftieth and early sixtieth: In those early days in many countries no comprehensive legal basis for research reactor operation existed. Due to the absence of national codes and safety criteria the regulatory authorities used in many cases the international accepted standards for nuclear power reactors and modified them for application to research reactors. 1.2 Situation in Austria (1958 - 1969): - no Health Physics Act issued - only normal hearing for civil engineering permission - drinking water quality required for the reactor and chemical laboratories waste water rejected into the Danube canal. - Geological and Hydrological requirements as for any normal industrial 4-1 plant - short expert reports as basis for a safety report were compiled from * Vienna Veterinary University; result "no objection concerning fauna" * Vienna university of Agriculture; result "no danger to vegetation expected" * university of Vienna, Physics Department; concerning: heat production, radioactivity, earthquake, sabotage 1.3 Situation in Austria after 1969 After the Austrian Health Physics and Radiation Protection Act of 1969 has been issued (which is the legal basis for licensing of nuclear installations in Austria) a different way of thinking took place. Based on this law a new application for the operating license for the TRIGA, which has been in operation since 1962 was necessary. However the special features of the TRIGA concept considerably facilitated the licensing procedure taking the following TRIGA characteristics into account - the inherent safety characteristic in respect to a reactivity accident. - no forced emergency cooling system for the core is necessary in case of a loss of coolant accident. Therefore no emergency power supply is required. - relatively small activity inventary of the core even after extended periods of operation at full power. - reduced requirements for the specification for control and instrumentation systems (no consequent redundancy or diversity). Therefore the main changes in the safety philosophy since start of the Vienna TRIGA in 1962 can be summarized: - No significant changes concerning technical aspects or the TRIGA concept because of its unique features mentioned above. Main restrictions: * Measurement of deformation of fuel elements. * Austrian TRIGA: Maximum permissible temperature difference between pool temperature and temperature in the reactor building (hall) was limited to 18°C. * increased sensitivity in monitoring fuel cladding failures by detecting radioactive aerosols directly above reactor pool. - Licensing requirements * continuous measurement, registration and documentation of all radioactive effluents from the reactor facility under any circumstances not only for emergency and post accidential conditions but also for reactor shut-down periods. As a consequence of this requirement: emergency instrumentation became necessary with * an uninterrupted power supply * redundancy of registration equipment, alarm indication of failure and following administrative procedure to ensure continuous registration, even for example, manual if no spare recorders are available. - increased safety requirements for the transport of fuel elements into and from the core (e.g. frequent checking of the transport equipment, ropes, chains, container etc.). 4-2 2. LICENSING Upgrading or license renewals require or at least make new and extended »credible accident analysis" highly desirable. There are mainly two reasons for réévaluation of accident analysis: First: New knowledge, data and experience have been gained over the past two or three decades of operating history of the reactors. Second: Over the years public opinion grew more and more sensitive to operating reactors especially if they are located close to cities - like the Austrian TRIGA (3 km from the center of Vienna) - although no objective reasons may be found for this situation. Subjects of credible accident analysis: - Questions regarding the effects of numerous pulses on the fuel. Are predictions still valid? - Earthquakes, reconstruction necessary? (not for Austrian TRIGA). - Fuel handling accidents should be more seriously considered. - Mixed core operation. Caused by changes of research programs or other reasons the safety report and other documents may be up-dated occassionally. Because of the operation of a TRIGA reactor as a research tool, licensing should be flexible enough not to prevent or restrict research work on one hand but must guarantee the basis of a safe operation of the reactor on the other hand. Minor changes of experimental set-ups need not be described in the safety report and thus should not result in a renewal of the licensing procedure. However, any changes or reconstructions of reactor components or experimental facilities which affect reactivity either in positive or negative direction, statically or dynamically to some extent must be documented in the safety report. In this case normally an amendment of the operating license will be necessary. In this connection another point may be mentioned: In research reactors sometimes part of the reactor instrumentation may be used in experimental set-ups for monitoring or initiating of different actions or, on the other hand, signals from experiments may cause certain actions as e.g. triggering the pulse operation of the TRIGA reactor. In these cases usually no licensing procedure is required as long as any feedback interference on the reactor instrumentation can be definitely avoided by proper measures as e.g. the application of isolation amplifiers in all outgoing connections from the instrumentation. Otherwise the whole system including the experimental set-up will be subjected to licensing. Although highly enriched uranium will not be available for TRIGA fuel elements in the future, still several types of fuel elements with different enrichment are used presently even in the same core. The operation of such "mixed cores" requires some attention by the operator. In highly enriched fuel elements the neutron flux density may increase if they are surrounded by low enriched fuel. This effect of local high power density may be negligible as long as steady state operation is concerned 4-3 due to the very low average power density and the low surface temperature of TRIGA fuel elements. However, pulse operation may result in a considerable increase of local power density and thus of stress in the fuel cladding. Due to this fact precaution should be applied in the layout of the fuel elements and control rods in the core lattice. In certain cases - depending on the number of fuel elements with highly different enrichment in the core - the operation of such mixed core may require special licensing procedures. Both, recommendations and/or regulations may result for the configuration of fuel elements in the core. The requalification program for reactor operating staff and other personal should also be mentioned which is required by the regulatory authority. Date and results of this program have to be documented. 3. PERIODIC INSPECTION Periodical reinspection is a routine procedure carried out by the reactor operating staff. However, the amount of reinspection tends to increase and for certain parts of the facilities the presence of govermental inspection authorities is often required, thus the reinspection procedure may grow to a burden for the owner or operator of the reactor resulting in a considerable amount of time and money to be spent. The owner or operator of research reactors - especially if they are operated by universities - are usually not really prepared for periodic inspections due to the shortage of personnel and lack of suitable personnel. Reinspection is an unwanted and burdensome job and not seldom the necessary understanding for this activity is missing. It turned out that establishing of a reinspection plan and schedule along with preprepared forms for the different tests to be carried out on the various systems are very helpful. 4. BACKFITTING OF REACTOR COMPONENTS AND INSTRUMENTATION Although an extremely successful operating record has been demonstrated by TRIGA reactors throughout almost three decades, several reactor facilities have been in operation now for more than twenty years. The question, whether to renew obsolete reactor components or control and instrumentation systems must be seen in the light of present days knowledge and technical standards. - Today no longer tolerable: Increased risks of release of radioactive material or loss of radiation shielding and/or coolant of the core due to the failure in obsolete coolant circuits with components which do not meet the technical standards of today. This also holds for the reactor tank. - Obsolete instrumentation and control systems cause operational problems as well as difficulties in obtaining proper replacement parts. - Systems approach their end of life (Bath-tube curve). - Aging does not only concern measuring channels but also connecting elements, wireing, plugs, coaxial connectors, soldering etc. 4-4 - Increased heat production of obsolete electronic equipment may cause accelerated aging of electronic components. Therefore the requirement for the instrumentation system to fulfil present day technical standards must be met to ensure a safe and trouble-free operation. Due to the unique features of the TRIGA fuel element concept eventually more attention may be paid to the emergency instrumentation than to the control and safety systems. A typical example is the requirement for the operator to continuously monitor, register and document the amount of all radioactive effluents from the reactor facility under any circumstances not only in case of emergency and post accidental conditions but also during reactor shut-down periods. 5. FUTURE WORK Future work which should be planned with respect to the above discussed fields. - TRIGA age and fuel elements accumulate more and more pulses. What is the affect on the integrity of fuel cladding? Are there reliable predictions for the future behaviour of the cladding possible? - Quantification of cladding behaviour under extreme conditions (e.g. metal-water reactions). - More realistic and not pessimistic source terms and modelling of consequences of fission product release for the various hypothetic accidents. 6. SUMMARY The TRIGA reactor Vienna was constructed and put into operation during a period where no special regulations for nuclear installations existed in Austria and its license was based on laws applicable to normal industrial facilities. Only after 1969 when the Health Physics Act was issued the TRIGA reactor Vienna had to undergo a new licensing process. In this process experience collected during the past decades where included. During this licensing procedure it was found that even a facility which was designed 25 years ago could be adapted to the most stringent requirements and to the state of the art with comparatively small problems and minor costs. 4-5 REACTOR PHYSICS ANALYSIS OF THE KUSASHI REACTOR FOR LICENSING PURPOSES H. Kadotani, M. Takami, 0. Aizawa* and T. Sato* Century Research Center Corporation Nihonbashi-honcho, Chuo-ku, Tokyo, Japan *Atomic Energy Research Laboratory, Musashi Institute of Technology Ozenji 971, Asao-ku, Kawasaki, Japan Abstract On the occasion of fuel replacement from aluminum to stainless-steel cladded fuel at the Musashi reactor (TRIGA-II, lOOkW)s a first attempt of reactor physics analysis has been conducted for licensing purposes. Hydrogen scattering cross section bound in zirconium-hydride is not avail able in ordinary cross section libraries. The THRUSH-code has been used to calculate this scattering cross sections and kernels. A unit cell calculation was performed with the WTMS-code system after adding this cross section to its library. The whole core analysis was performed in the diffusion approximation using the CITATION-code. The calculated results for k-eff's and control rod worth etc. were compared with the experimental data. Agreements between experiment and analysis are fairly good. 1. Introduction Objectives of the present analysis were to predict the change of reactor physics characteristics, when the fuel elements were replaced from aluminum to stainless-steel cladded fuel. At the same time, parameters like Doppler coefficients and the effective delayed neutron fractions etc. were evaluated from kinetic analysis of the core. The present analysis followed the established reactor physics method,. The first step was to obtain macroscopic cross sections for homogenized fuel and moderator mixtures. This process of the analysis was performed with WTHS-D code which was developed at Winfrith, UK. 4-7 Hydrogen atoms are bounded in zirconium metal forming ZrH„ molecules. The property of thermal neutron scattering by these hydrogens are well known, but the cross section is not provided in WIMS library. This cross 2 section was newly calculated with the THRUSH-code and was added to the original WIMS library. Whole core analysis was performed with the diffusion 3 code, CITATION to obtain the parameters like k-eff's and neutron flux distribution. The coupled WIMS-CITATION code system was benchmarked against the first critical experiments by using aluminum cladded fuels of the Musashi reactor. 2. Thermal Neutron Cross Section for Hydrogen Atoms in Zirconium-Hydride Scattering cross section for thermal neutrons is dependent on the chemical binding and temperature of scattering atoms. The scattering cross section of hydrogen bounded in zirconium atoms shows different behavior from the scattering cross section of hydrogen atoms in other molecular states, for instance, hydrogen atoms in light water. x 60 N *- 50 o r 40 iz 20 o (X10) CO 10 - 0.05 0.1 0.15 Frequency (cu) (aV) Fig. 1 Spectral density of zirconium-hydride 4-8 It is well known that the scattering cross section for thermal neutrons can be calculated through a function called the spectral density or the frequency distribution. The spectral density for hydrogen atoms in ZrH„ 4 l was calculated by Slaggie . However, in the present analysis we used the spectral density which was evaluated by General Atomic which was a simpli fied and widely used density for this type of analysis (see Fig. 1). The scattering kernel was calculated with the THRUSH-code with this spectral density. Molecules of zirconium-hydride are reported as ZrH for aluminum cladded and ZrH. . for stainless-steel cladded fuel. We used, however, the spectral i. b density evaluated for ZrH-, neglecting these differences. The calculated total scattering cross section is shown in Fig. 2. = 80h- W 70 60 N 50 O 40 >*— UJ 30 » 20 10 10 3 10 2 10 1 Neutron Energy (eV) Fig. 2 Scattering cross section of zirconium-hydride 4-9 3. Group Constants Geometry and components of aluminum and stainless-steel cladded fuel are shown in Table 1. To perform homogenization calculation with the WTMS-D code, it was necessary to cylindricalize the outer boundary of coolant. As the distance between fuel rods in cylindricalized unit cell is different for each fuel ring. Therefore, these equivalent radii are also shown in Table 1. In the first ring which is the center of the core, no fuel element can be loaded. Geometry and components of the control rod are shown in Table 2. Table 1. Geometry and Components of Fuel Elements ^^^__^ Cora Stainless-steel Aluminum Cladded Cladded Items —- _^__^^ Fuel Element Fuel Element Content of Uranium (w/o) 8.5 8.0 Atom Number Ratio (H/Zr) 1.6 1.0 Enrichment (%) 19.8 20 Weight of Uranium (g) 197 175 Weight of U-235 (g) 39.006 35 Diameter of U-ZrH Pellet (cm) 3.6449 3.597 Cladding Material SUS-304 AI 6061 Thickness of Cladding (cm) 0.0508 0.076 Outer Diameter of Cladding (cm) 3.75412 3.749 1st Ring (A-ring) 4.066 2nd Ring (B-ring) 4.624 Equivalent Radii 3rd Ring (C-rlng) 4.584 of Coolant (cm) 4th Ring (D-rlng) 4.590 5th Ring (E-rlng) 4.590 6th Ring (F-rlng) 5.1 10 Average 4.784 4-10 A unit cell model was used for the fuel lattice, however, the multi- cell option of the WIMS-D code was selected to obtain control rod group constants. The effect of neighbouring cell can be treated approximately in the multi-cell option. Effective multiplication factors (k-eff's) calculated with WIMS-D code are shown in Table 3. In this calculation, the bucklings were assumed to be 0.012 cm and 0.00316 cm" for axial and radial directions, respectively. In Table 4, the kinetic parameters of delayed neutron fractions and prompt neutron life times are shown as the cell values, which are different from the adjoint weighted average values for a whole core. Group-constants for structural materials other than control rods were also obtained with multi-cell option of WIMS-D code. These regions were water channels, graphite elements and the samarium burnable poison placed at the end of fuel elements. Group-constants, which were necessary for 2-dimensional analysis of the core, were obtained by axial and radial 2- dimensional diffusion calculations of 36 energy groups. The regions treated by this method were: water as the radial reflector, mixture of water and graphite (water fraction: 38 volume %) as the upper and lower reflectors, mixture of water and aluminum (water fraction: 0.103 volume %) as the upper grid plate and pure aluminum as the lower grid plate. Two-dimensional calculation was performed with 6 energy group-constants in the diffusion approximation. The energy boundaries were 10 MeV, 821 keV, 3.5 keV, 4 eV, 0.621 eV and 0.1 eV. Table 2. Geometry and Components of Control Rods " ' -~^^_Control Rods Items •——-—_^___^ Regurating Rod Shim Rod Radius of B4C Pellet (cm) 1.0 1.5 Outer Radius of 1.1 1.6 Aluminum Cladding(cm) Atom Number Density''' 0.0544 0.0S44 of Boron (X1024 n/om^) * 50% of Theoretical Density 4-11 Table 3. Effective Multiplication Factors Calculated with WIMS-D \„^Paramôters Fuel Temperature (°K) Ring Core ^-v. 293 323 373 8 1.0040 1.0023 0.9991 Aluminum c 1.0053 1.0037 1.0005 Cladded Fuel Element D 1.0051 1.0035 1.0003 E&F 0.9738 0.9720 0.9683 B 1.0350 1.0328 1.0287 Stainlesa-stee! Q 1.0387 1.0386 1.0325 Cladded Fuel Element D 1.0381 1.0360 1.0319 E&F 0.9845 0.9822 0.9775 -2 _ Input Buckling B2r=o.012 cm B22==0.00316 cm 2 Table 4. Kinetic Parameters ^"•\^^ Core Aluminum Stainless-steel ^""^^^^ Cladded Cladded Parameters ^s'\^ Fuel Element Fuel Element Delayed Neutron 0.00653 0.00653 Fraction Prompt Neutron 5 4.1 1 x 10~ 3.58 x 10 "5 Life Time (sec) 4-12 4. Criticality Analysis (1) Initial Core By analyzing the critical experiment of the initial core, we established the accuracy of the present analytical method. The criticality calculation was performed in R-Z geometry shown in Fig. 3. The calculated k-eff was 1.05828. The k-eff should be 1.0 as obtained by the experiment, though we corrected the following effects which were neglected in the calculation. - Effect of burnable poison Change of k-eff was evaluated due to samarium at the ends of fuel elements. For this purpose, the 1-dimensional core was modeled as shown in Fig. 4. Difference of k-eff's between two cores with and without samarium region of burnable poison was calculated to be 0.0207. o Water Upper CM Upper Grid Plate 3rid Pia rP3 Reflector Upper Burnable Poison Reflector z& -22.72- Radial Reflector CO Core CO «9 Core a id CM -30.818- d 1_. Burnable Poison Under y Reflector Reflector Under , Under Qrld Plate Grid Plate o Water d~ unit: cm (unit: cm) Fig. 3 R-Z geometry for Fig. 4 1-dimensional criticality calculation core calculation for burnable poison effect 4-13 - Correction of control rod At the initial criticality experiment, the shim rod had been partially inserted into the core. The worth of the rod was estimated to be 0.0015 experimentally. After correcting the above two factors, the calculated k-eff becomes, k-eff(cal.)=1.05828-(0.0207+0.0015)=1.0361. Even after the correction the calculated k-eff is still larger than the experiment by 3.6%. This difference may be larger than the expected value which should be less than 1%. However, we are not successful in explaining this error so far. (2) Criticality Analysis for Stainless-steel Cladded Fuel The R-Z geometrical model, which is utilized in the previous section, cannot be applied for the core analysis to obtain the number of fuel rods to reach critical or control rod worth. One should model the core directly in 3-dimensional geometry. However, since this needs rather long computer time, the direct analysis of 3-dimensional cores is not applicable practi cally. As the alternative to this, the core was modeled in 2-dimensional R-8 geometry. Vertical neutron leakage was evaluated using the buckling in z-direction. In Fig. 5 the 9-division of the core is shown. 15 }f 13 & u Fuel Element 0 Graphite Element % Control Hod Q Neutron Source Fig. 5 9-division of R-8 geometry for criticality analysis 4-14 The axial buckling was obtained as follows. The initial core was. re calculated in the R-8 model and the axial buckling was searched to give the k-eff»1.0538 which was obtained in the R-Z calculation. The searched axial -3 -2 buckling was found to be 3.54x10 cm The fuel loading pattern was obtained at k-eff=1.0361 for stainless- steel cladded fuel. The bias of k-eff (3.6%) was evaluated in (1). It was found that fuel rods should be added a.t the position F26 and F27 to the initial core pattern. The k-eff for this core was 1.0383. The total number of fuel elements was experimentally found to be 66, when we finished the criticality search. The number of difference between the calculation and the experiment was 4. The agreement of experiment and analysis can be said good. The number of fuel which corresponds to 2$ excess reactivity was obtained for stainless-steel cladded fuels. The k-eff of the 2$ excess core was assumed to be 1.0543 (here we used 0.8% for ß-eff) . To add 2$ excess to the original core, the fuels are to be added at the positions at 1, 2, 3, 28. 29 of F-ring. The calculated k-eff was 1.0555. 5. Rod Worth and Doppler Coefficient (1) Rod Worth The rod worthes of shim and regulating rods were obtained at full withdraw condition. The results are shown in Table 5. The measured rod worthes were -3.0 and -0.72%Ak/k for the core of aluminum cladded fuel . Agreement of experiment and analysis can be said good. Table 5. Control Rod Worth (%Ak/k) """"—•~^^^ Control Rod Shim Rod Reguratlng Rod Core ^"""~--~-^^^ Aluminum Cladded -0.72 Fuel Element -2.90 Stainles3-steel Cladded Fuel Element -2.45 -0.55 4-15 (2) Doppler Coefficient The Doppler coefficient was obtained by subtracting the k-eff's at 323 K and 373°K. The cross sections which are dependent on temperature are resonance cross section of U-238 and hydrogen scattering cross section in the thermal energy region. The expansion effect due to elevated tempe rature are neglected in the present analysis. The calculated Doppler effects are shown in Table 6. The Doppler coefficient for the stainless-steel cladded fuel core is larger than that of the aluminum cladded fuel core, which means that the core of stainless-steel cladded fuel is advantageous than that of the core of aluminum cladded fuel. Table 6. Doppler Coefficient (x 1 0-5Ak/k/°C) ^^-•^^^ Temperature ^~"---~^^ Range 293-323°K 293-373°K Core --^^^ Aluminum Cladded -6.3 -7.4 Fuel Element Stainless-steel Cladded -8.0 -9.4 Fuel Element 6. Conclusion A first attempt of reactor physics analysis has been conducted for licensing purpose at the Musashi reactor. The calculated results for k-eff's and control rod worth, etc. have been compared with the experimental data. Agreements between experiment and analysis are fairly good. References 1. Askew, T. R. et al.: J. Brit. Nucl. Energy Soc., 4 [4], 564 (1966) 2. Kadotani, H.: JAERI-M 8727 (1980) 3. Fowler, T. B. et al.: ORNL-TM-2496, Rev. 2 (1971) 4. Slaggie, E. L.: GA-8132 (1967) 5. Kopper, J. U. et al.: GA-8774 (1968) 6. Nozaki, T.: Private Communication (1988) 4-16 Milan Osredkar, V. Dimic, H. Dusic, H. Kozuh, D. Vojnovic Institut "Jozef Stefan" Jamova 39, Ljubljana PROBABILISTIC SAFETY ASSESSMENT OF THE RESEARCH REACTOR - TRIGA HARK II ABSTRACT The probabilistic safety analysis method was used to assess a risk level - associated with the research reactor TRIGA MARK II The model of the reactor is presented by the system event trees for four initiating events. The system fault trees and the critical sequence event trees are reduced by means of the FTAP 2 program, and the quantitative evaluation was done by means of the IMPORT programme. The , four initiating events selected were: LOCA, Loss of Electric Supply, Loss of Flow and Loss of flow-fuel Channel Blockage. For all four initiating events, event trees were constructed and the probability for core damage was calculated as well as the probability for the radioactive release. Most critical sequences were identified for each event trees and further analysed giving Importance measures (Im), Risk Reduction Rations (RRW) and Risk Achievement Rations (RAW). At the end , beside giving estimate about the risk level, some suggestions are given based on our results. 1.0 INTRODUCTIOH The research reactor TRIGA MARK II from the Reactor Centre in Podgorica has be^n working since 1952. tt has three purposes: i. Operators training 2. Research in neutronics 3. Isotope production 4-17 Operators training includes the possibilities of different errors at controlling the reactor e.g. injection of the fuel element when the reactor is already critical, or position height of control rod over the permitted limit. To fulfil the basic safety criteria (the safety criteria 1) the reactor itself must be safe, that is it must, have a large prompt negative coefficient.lt means that the reactivity is automatically reduced at the increased fuel temperature. Besides the basic safety criteria, which was mentioned above, the second criteria must be fulfilled as well. The reactor is provided with systems which prolong the basic safety system functioning after the appearance of the anticipated undesirable events. The biggest hazard (the risk source) is the release of radioactivity to the environment. There are such accident sequences, though the probability of occurrence of such events is extremely small. 2.0 PSA ANALYSIS The analysis is expected to be carried out to the second PSA level. It means that the results, which assess the estimated probability of the radioactivity release outside the reactor building, are expected. The analysis contains" the modeling of the incident event development for selected initiating events using the reference IAEA material. The analysis is limited by the number of the initiating events. 2.1 CHOICE OF IHITIATIHG EVEHTS The choice of the. initiating events is made using the reference IAEA document(IAEA-TEC DOC-400). The model contains four initiating events: 1. Loss of the power supply. 2. Loss of the coolant circulation (two cases). 3. Loss of the coolant (LOCA). Since the grid is the only power supply source, the initiating event -the loss of the power supply- contains the loss of the power supply from the grid or electrical component failure and disconnection of the cables for the power supply of all reactor systems. The initiating event- the loss of coolant circulation comprises the events liKe stopping of the heat removal from the core as the result of limited or completely impossible natural coolant circulation around the fuel rods. The examples are different plugs of the fuel channels or of the lattice, etc. The failure of the primary coolant circuit is not expected here. 4-18 The second possibility regarding the loss of the coolant circulation is unavailability of the emergency circulation which, results in unavailability . of heat removal from the reactor pool. The reasons can be the following: the pump failure in the primary cooling circuit, the reduction or the loss of the flow because of the plugged heat exchanger, valve, etc. On the basis of the above mentioned results, the initiating event is separated into two initiating events: 1. loss of the emergency circulation in the primary circuit a. loss of the natural circulation through the part of the core(the partial loss of the natural circulation through the core). The initiating event-the loss of the coolant- supposes only the reactor pool inhalt to flow through different penetrations. The flow through the eventual tube breaks of the primary coolant circuit is not considered. Because of the high negative temperature coefficient, the initiating event- the excess reactivity insertion- is not considered. £.2 EVEHT • TREES Four event trees are constructed. For each sequence the core condition and the category of the radioactivity release to the environment is established. It is supposed that the state of any safety system failure defined in the heading lasts at least till the challenge of the negative consequences, i.e. the permanent system shut down. Loss of power supply Assumptions: - the power supply loss is not a short one - the water tower inhalt is sufficient for the emergency core cooling Description: There are five sequences identified. The sequence No. i results from the effective trip of the reactor. The immersed core enables the safe reactor condition. In the case of the unsuccessful reactor trip the temperature of the fuel and pool water raises, because the heat removal from the pool does not exist. .According to the emergency procedures, the emergency cooling from the water tower is used only when there is a danger of the core discovery. At the successful use of the emergency cooling, (the sequence Ko. 2.) the core is not damaged. 4-19 At the failure of the emergency core cooling tlie fuel temperature increases, and. because of the high negative temperature coefficient of the reactivity, the reaction drops on the lower value. In this case the core is damaged later and slowly. If the building is not isolated, the radioactivity will be released to the environment (the sequence No.5). Success criteria: The reactor protection system: -generation of at least one signal for the fast trip; -dropping of one of the three reactor rods into the core. The emergency cooling system: -supplying of the cooling water from the water tower to the reactor pool. The reactor building isolation: -closure of the isolating valves of the reactor building; -recirculation of the building ventilation system. Loss of coolant circulation Assumptions: - one out of the two water sources for the emergency cooling supplies the sufficient quantity. Description: The final states of this initiating event are similar to those in the previous initiating event, i.e. the core is not damaged, the sequence Ho.5 results in release of the radioactive gas products to the atmosphere. In the case of the successful reactor trip, the performance of any other safety system is not neccesary(the . sequence No.i). In the opposite case, the emergency cooling must be ensured. It prevents the core damage. The failure of the emergency cooling is followed by the increase of the fuel element temperature and of the pool water. At this stage the core is slowly damaged and radioactive gas will be released to the reactor building atmosphere ; if it is not isolated, the gases will withdraw out of the building. Success criteria: The emergency core cooling system: cooling water supply either from the deraineralized water tanK or from the water tower; the rest is the same. 4-20 Loss of coolant accident Assumptions: - one of the two water sources for the emergency cooling is sufficient; the cooling water in the pool is not radioactive at the normal reactor performance Description: Seven sequences are identified; two of them result in possibly smaller core damage, and three in the release of certain radioactivity to the environment. The reactor cooling system is not incorporated in the tree, because it is not required after a certain decrease of the water level in the pool (below the suction opening of the cooling system). The result of the sequences Ho.l to Mo.5 is not critical. In the case of the unsuccessful reactor trip the reactivity is reduced because of the negative temperature reactivity coefficient. In the case of successful emergency cooling the core will not be damaged. In the opposite case, because of the fluid loss the core damage can occur, even if the reactor itself shutdown. A certain radioactive contamination of the air in the building occurs , and if it is not isolated also the release to the environment (the sequence Ho.5 and 7). Success criteria: the same as in the previous cases. Loss of normal circulation through part of core Assumptions: no Description: Three sequences are identified. The reactor protection system prevents the undesired consequences (the sequence No.i). Otherwise the partial core damage occurs and the release of the radioactivity, if the reactor building isolation fails. The emergency core cooling is not appropriate. 4-21 INITIATING REACTOR EMERGENCY REACTOR BUILDING CORE EVENT-LOSS PROTECTION COOLING COOLING RELEASE OF ELECTR. SYSTEM SYSTEM SYSTEM ISOLATION STATE POW. SUPPLY SEQ . 1 NO CD NO R •HO CO NO R 3 NO CO NO R 4 CD NO R I.38È-8 5 CD R U8E-I0 FIG. 1 EVENT TREE- LOSS OF ELECTRICAL POWER SUPPLY T. 1.38E-8 1.38E-10 INITIATING REACTOR REACTOR EMERGENCY REACTOR BUILDING CORE EVENT-LOFA PROTECTION COOLING COOLING RELEASE FUEL CHANN SYS TE Ml SYSTEM SYSTEM ISOLATION STATE EL BLOCKAGE SEQ . 1 MO CD NO R 2 P. CD NO R 1.04E-6 3 P.CD R 1.91E-8 FIG. 2 EVENT TREE - L0S3-0F-FLOW ACCIDENT, BLOCKAGE, FORCED CIRCULATION AVAILABLE "NT" 1.Û4E-Ô 1.91E-8 4-22 INITIATING REACTOR •REACTOR EMERGENCY REACTOR CORE EVENT PROTECTION COOLING COOLING BUILDING RELEASE SYSTEM SYSTEM SYSTEM ISOLATION STATE LOCA SEQ . 1 NO CO MO R 2 NO CD NO R 3 NO CD NO R 4 NO CD NO R 5 NO CD NO R 5 CD NO R 6.1E-11 7 CD R 3.2E-11 FIG. 3 EVENT TREE - LOCA "N~ 6.1E-11 3.2E-11 INITIATING REACTOR REACTOR EMERGENCY REACTOR EVENT - LOSS PROTECTION CORE COOLING COOLING BUILDING RELEASE OF FLOW SYSTEM! SYSTEM SYSTEM! ISOLATION STATE SEQ . 1 NO CD NO R 2 NO CD NO R -7 O NO CD NO R 4 CO NO R 3.38E-8 5 CD R 3.4E-8 FIG. 4 EVENT TREE -• LOSS-OF-FLOW ACCIDENT i FORCED CIRCULATION UNAVAILABLE "\~~ 3.38E-S 3.4E-8 4-23 2.3 FAULT TREES For each, event tree heading an appropriate fault tree is constructed -the system failure modelling is carried out. The models depend on each other, on the supporting systems they have in common, on the operator s commands, etc. To avoid the multiple modelling of the same systems, which occur in many sequences, for each system a fault tree is constructed. Later, it is reduced for the use in each event tree sequence. The construction of the common fault trees are shown. They are reduced depending on the initiating event and the sequences. Fault tree of reactor protection system Typical of this tree is multiple redundance, specially in the part of the signal generation and the electromagnet power supply interruption. On the upper levels of the tree human errors are present as well as the common cause failures. The shown tree includes in its configuration all expected signal channels. At the processing of sequences the tree is reduced to the possible occurrence of the signals, which are produced by the postulated initiating event. Thus tire fault model of this system is not qualitatively changed according to the initiating events, but only reduced. Fault tree oS core cooling system The fault tree of the system for the core cooling (the reactor pool cooling) contains the failures of the primary and secondary circuit. The failure of the heat removal function from the reactor pool to the environment is the top event of the tree. As the supporting systems occur the power supply and the source of the secondary cooling water. This system is placed in the heading of the event tree, but it is not placed in a single sequence. Anyway, the safety of this system is included in the analysis, because it is one of the vital systems for the safe reactor operation. Fault tree of the emergency core cooling system There are also some elements which belong to the cooling system. Since the cooling, system does not appear in the sequences, it is supposed that the two systems(f unctions) do not dep'end on each other. According to the assumption that the water tower mhalt is sufficient to perform its function, the water tower supply is left out in the model. The shown basic tree is reduced regarding the initiating event. For example, at the loss of the power supply the emergency 4-24 supply from the water tank by means of the electric pump is switched off; or the secondary cooling system source is not required in the discussed initiating events. Fault tree of reactor building isolation system The model treats the functions of the reactor building isolation- the removal of the radioactive gas and fluid. To prevent the removal of the radioactive gas the isolated reactor building is required as well as the purification (reduction) of the radioactivity in the reactor building air. This is necessary since the complete isolation of the building is not achieved by the isolation of the ventilation openings, i.e. the release of radioactivity is possible through different gaps. As in the previous trees, the reduction of the fault tree depends on the initiating event. At the loss of the power supply the manual manipulation with the isolation valves is planned; the recirculation phase is not taken into account. 2.4 DATA BASE Initiating events The probabilities of the postulated initiating events are assessed by frequency of the events or failures, which cause the initiating event. For the initiating event of the power supply loss the experience is used. i. Loss of coolant accident: a)-Tank damage 1.3E-5 per year b) Leakage at penetration places 6.95E-4 per year 7.1E-4 per year 2. Loss of primary coolant circulation: a) Failure of both pumps 2.2E-4 per year b) Valve plugging 1.2E-3 per year c) Flow meter plugging 1.23E-1 per year 1.22E-1 per year 3. Loss of electric power supply: One grid disintegration in three years,according to the Bayes method 4.3E-1 per yeai 4-25 4. Loss of (natural) circulation through the core: On the basis of the reference documents IAEA-TEC DOC-400 2.5E-2 per year Basic events in the fault trees The data sources at the establishing of the „data base are: the generic data (IEEE Std 500 Reliability Data), the data from the analysis on other research reactors,' and in certain cases the data based , on the experience. The reference times for the unavailability assessment are defined on the basis of the test period and the mission times. 1. Reactor protection system: The test period is once per weeK., and for some exceptions once per year. 2. Reactor protection system and the reactor building isolation system: Mission time 24 hours 2.5 MODEL PROCESSING The reactor model processing is performed by means of the shown critical sequences. The following programs are used: FTAP 2 for the reduction of the fault trees, IHPORT for the fault trees quantitative evaluation. To start with, the processing of the fault tree at the system level is carried out. 2.6 SEQUENCE PROCESSING For each critical sequence two issues are important: the core state and the radioactivity release to the environment. The sequence model is presented by the fault tree connected with the initiating event over the AND gate. The successful parts of the event tree branches are not taKen into account, because their contribution to the sequence issue can be neglected. The probabilities of the core damage and of the radioactivity release to the environment for the critical sequences are shown in the event tree tables. The final result of the possibilities of the core damage and of the radioactivity release to the environment presents the sum of the critical sequence issues at each initiating event. 4-26 Probabi1i ty No Initiating Event. CQre Damage Radioaci:ivi Re I ease 1 LOCA 3.2E— i i 2 LOFA 3.38E-8 Loss of Electrical Power S Li p D 1 y 1.3SE-8 • iio LOFA - Fuel Channel bl cjckage 1.Q4E-6 1 . «?•! P-R 1.09E-6 b.32E-3 FIG. 5. Final results of the probabilities of the core damage and of the radioactivity release 3.0. FIHAi, COHCLUSIOHS Given analysis of the risk source and of the reliability of the research reactor TRIGA gives results which can be used for the comparison with studies done on similar reactors or with the reference assessment. It also gives some idea of the reactor safety. The results can also be used for the maintenance of the established safety level or perhaps for it£s increase b-y improving maintenance, operation or even changes in the design of safety systems. 1. Critical sequences: For the undesired event - the core damage - the most critical sequence is found for the initiating event "Loss of natural circulation in the part of the core". Two out of three sequences lead to a partial core damage. The most critical as far as the radioactive release is concerned is the sequence 5 for the initiating event "Loss of flow accident" , and sequence 3 for the initiating event "Loss of flow accident - blockage". 4-27 2. Critical events: The most critical events found were: -loss of electric power connected, with, loss of coolant accident -common cause events for control rods and relay contacts for all initiating events and specially in the case of the loss of circulation through the core 3. Safety system design The automation of the reactor building isolation and the ventilation switch over to the recirculation mode would reduce the possibility of the radioactivity release to the environment. 4.0 REFEKEHCES /l/ IAEA-TEC DOC-400, "Probabilistic Safety Assessment for Research Reactors", Vienna, 1980 /£/ Dimic,Dr Viktor, "Opis reaKtorja TRIGA II" , Ljubljana /3/ IEEE Std 500, "Reliability Data", Mew York, December, 1983 4-28 PROBABILISTIC SAFETY ASSESSMENT OF THE VIENNA TRIGA REACTOR C. Kirchsteiger, : H. Bock Atominstitut der Österreichischen Universitäten, Vienna, Austria 1. INTRODUCTION The TRIGA Mark-II reactor Vienna achieved its first criticality on March 7th, 1962 and is in operation since that time without major undesired interruption about 220 days a year. It has a maximum continuous power output of 250 kW and may also be operated in the transient mode with a peak power level of 250 MW. As the fuel consumption is rather low at this power level, the core still contains 54 aluminum clad fuel elements from the first criticality as well as 18 stainless steel clad fuel elements, either with 20% enrichment or with 70% enrichment (FLIP-elements). Therefore, the core is considered as a mixed core with the stainless steel elements concentrated in the two center core rings (B- and C-ring), while the Al-clad elements occupy the outer core positions. The reactor is heavily used for neutron and solid state physics experiments (utilizing three beam tubes, one thermal column and one neutron radiography facility), as well as for radiochemistry and neutron activation analysis, (one beam tube and in-core position). Details of the reactor utilization has been published elsewhere /l/, but it has to be mentioned that experimental work around the reactor or in the reactor tank poses an additional non-negligible risk in view of reactor accidents. The intention of this work is to identify, as an example for a TRIGA reactor, the various ways of possible and hypothetical accident sequences using the system characteristics of the TRIGA reactor Vienna. As mentioned before, the undesired top event is the radiation exposure of persons above international accepted limits. Each accident sequence which finally leads to this top event involves a different amount of radioactivity which has to be identified. Compared with a complete risk analysis like WASH-1400, a similar classification of release categories is not possible however, "consequence categories" were introduced to distinguish among different failure event sequences. 2. METHODS AND ASSUMPTIONS First of all, the external limit which confines the reactor system has to be defined. Within the scope of the constructed Master Logic Diagram (MLD), the external limit is defined by the reactor hall which includes in this case also the ventilation system, primary cooling system and equivalent boundaries. Radiological accidents which are associated with the handling of radioactive products out of the above defined area (i.e. radiochemistry laboratories) are not taken into consideration. The group of persons who would eventually be exposed to radiation is divided into two subgroups called internal (facility staff) and external (public) persons. Both groups can be exposed as a consequence of accidents which result from the logic of the MLD-scheme and the therein included initiating and primary fault events. The exposure level depends on the amount of radioactivity released. To set an international limit, it was decided that the whole body dose limits for occupational staff and public are taken as recommended by ICRP. 4-29 For this study, the usual PSA methods as i.e. discussed in /2/ were applied. However, because of the special aspects of a research reactor instead of an event tree analysis many individual fault trees to be combined in an MLD-scheme were used. Failure of components to be assumed in the fault trees were always taken as complete failures. Even if the component could still partly perform its function, it was considered as a complete failure. Failure rates were always assumed to be constant quantities in time. Each fault tree event is marked with a consecutive number to denominate possible common events in a standardized way. For the Boolean logic statements describing the fault trees the symbol Ej is used for the event number i. The symbol A: stands for the events in the MLD. Due to lack of data some cases only can be analyzed qualitatively; quantitative data were obtained from /6-16/. The fault tree evaluation was performed with computer programmes /3,20/. 3. FAULT TREE ANALYSIS OF THE TRIGA REACTOR VIENNA 3.1 The Master Logic Diagram (MLD) (Figure 1) Assumed top event A, is the "radiation exposure of persons" (occupational and/or public). The shaded events together with conditional event A4 are considered as important events and are treated in further detail in this study. FIGURE 1: MASTER LOGIC DIAGRAM AND TABLE OF USED SYMBOLS ZJ ZS" ^ A 5 A6 7£Y A 7 Aio An A 12 z±y Au 14 15 c& A 16 Al7 ia 19 4-30 rectangle fault event; it is usually the result of the logical combina tion of other events circle independent primary fault o event diamond fault event not fully developed as to its causes; it is only an assumed primary fault event OR gate the union operation of £B events, i.e. the output event occurs if one or more of the inputs occur AND gate the intersection operation of Ä evens, i.e. the output event occurs if and only if all the inputs occur INHIBIT gate output exists when X exists and condition A is present; this gate functions somewhat like an AND gate and is used for a secondary fault event X triangle-in triangle symbols provide a tool to avoid repeating sections of a fault tree, or to transfer the tree construction from one sheet to the next. The triangle-in appears at the bottom of a tree and represents that branch of the tree (in this case "A") shown someplace else. triangle-out appears at the top of a tree and denotes that the tree "A" is a subtree to one shown someplace else. LIST OF EVENTS IN THE MLD-SCHEME s.j radiation exposure of persons (public and occupational) ^2 radiation exposure of public ^3 radiation exposure of facility staff i.4 fault in ventilation system of the reactor building i-5, A5' radiation inside the reactor building due to release of activation products >.g radiation inside the reactor building L7 massive release of fission products from fuel elements ig release of activity from other sources 19 disturbances in reactivity LJQ fault of fuel element transfer cask m direct gamma exposure from reactor pool 122 direct gamma exposure from other sources H3 release of activity from experiments H4 release of ion exchange resin LJ5 release of activity products through loss of coolant L26 loss of primary coolant from reactor tank including experimental facilities ^17 loss of primary coolant by failure of primary circuit components other than tank Hg loss of primary coolant due to faults of the reactor tank itself 49 loss of primary coolant through faults of beam tube 4-31 3.2 Aj - Massive Release of Fission Products from Fuel Elements (Figure 2) Quantitative values for some relevant events in A7: The fault tree A7 was split into four main branches due to the possible locations of fuel elements, which is either - in the reactor core - in tank storage racks - in fuel storage pits - during manipulation, transfer etc. The frequencies per year for some specific events were assessed and the following results were obtained: release of fission products F(22) = 5.61x10* - during transport F(24) - l.SôxïQ- - from the core F(25) - 5.35x10" - from storage rack F(26) = 4.87x10' - from storage pits (mechanical causes) F(27) = 3.33x10 -If- - from storage pits (chemical causes) It should be emphasized that the above values are based on the assumption of massive fission product release (i.e. in excess of the maximum permissible whole body dose) as consequence of the particular top event. FIGURE 2: A7 - Release of fission products from fuel elements ED 4-32 A7 - RELEASE OF FISSION PRODUCTS FROM FUEL ELEMENTS AJQ fault of fuel element transfer cask 21 release of fission products from the reactor tank 22 release of fission products during transport 23 release of fission products from fuel storage pits 24 release of fission products from core 25 release of fission products from fuel element storage racks in the tank 26 damage of the stored fuel elements by mechanical causes 27 damage of the stored fuel elements by chemical causes 28 mechanical damage of the fuel elements 29 chemical damage of the fuel elements 30 thermal damage to the fuel elements by excess fuel temperature and consecutive release of fission products 31 accidents during experiments or during fuel loading 32 debris of the reactor building falling into the tank 33 chemical reactions due to experiments (Cu, Hg) 34 corrosion products in the primary coolant 35 excess fuel temperature above specified limits (the trigger temperature for an RSA8 and an RSA9-SCRAM is 375 °C, respectively) 36 fuel element manipulation due to experiments (explosion, implosion) 37 fuel element manipulation due to changes in reactor core configuration 38 fuel element manipulation to measure fuel element elongation and bowing 39 explosion (sabotage); EXTERNAL EVENT 40 aircrash; EXTERNAL EVENT 41 seismic event; EXTERNAL EVENT 42 corrosion of reactor components 43 core totally or partly uncovered (i.e. by plastic foil, glass plate) 44 bad positioning of fuel elements in the core 45 corrosive chemicals in primary coolant due to experiments 46 instruments for surveillance of water chemistry not calibrated (conductivity, pH- value) 47 instruments for surveillance of water chemistry are indicating properly, but operator does not react or misinterprets (human error) 48 mechanical damage of fuel elements during transfer from core to storage racks in the tank 49 accidents during experiments effecting the stored fuel elements 50 corrosion of the stored spent fuel element in the storage pits 51 inadequate surveillance of the water chemistry (human error) 52 corrosive chemicals in the water of fuel storage pits 53 normal corrosion of the stored fuel elements in the storage pits 54 RSA11-SCRAM (manual SCRAM) is failing resp. is not initiated 55 RSA8 resp. RSA9-SCRAM (automatical SCRAM) fails 4-33 3.3 A16 - Loss of Primary Coolant from Reactor Tank (Figure 3) FIGURE 3: Loss of primary coolant from reactor tank EXPLANATION TO FIGURE 3 A18 loss of primary coolant due to defects in the reactor tank Aj9 loss of primary coolant through failures of beam tubes or thermal column 39 explosion (sabotage); EXTERNAL EVENT 40 aircrash; EXTERNAL EVENT 41 seismic event; EXTERNAL EVENT 111 beam tube doors not closed due to experiments 137 mechanical damage 138 damage due to beam tube installation (i.e. collimator loading) 139 damage of beam tubes due to external events 108 parts of the reactor building crash into the tank and damage the core (Debris falls into reactor tank and damage beam tubes) 147 heavy object falls accidently into reactor tank (i.e. fuel transfer cask) The frequency of top event Ajg was calculated to 5 ,-1 F(A16) = 1.55x10-° a 4-34 4. QUANTITATIVE RESULTS 4.1 Frequencies and Consequencies of Relevant Failure Events Although more than 50 TRIGA reactors are in operation worldwide and they have several common properties such as U-Zr-H fuel, design of the control system, tank construction etc., it would be expected that appropriate data on failure rates of important components exist. However, during this study it was found that component failures are poorly documented, not accessible and no centralized data collection has ever been established. Therefore, the present study relies on data collected during 26 years of operation of the TRIGA reactor Vienna and on some data obtained through personal communication from other stations. This makes this study very specific for the Vienna TRIGA reactor but, as a model, it could enhance similar studies at other stations. In order to give the obtained results a rating, all individual fault trees which have relevant consequences to the top event Aj ("radiation exposure of persons") were classified in three major "consequence categories" listed below. a) Low consequence: releases with very low or no consequence for facility staff or public which includes 109.Elcj (= inadequate storage of radioactive sources together with failure of hall area monitoring system) llO.Ej^ (= access to restricted areas together with failure of hall area monitoring system) A|g (= loss of primary coolant from reactor tank) b) Medium consequence: release with low consequences on facility staff and no effect to public which includes 24 (= release of fission products from core) 25 (= release of fission products from fuel storage racks) 26 (= damage of the stored fuel elements by mechanical causes) 27 (= damage of the stored fuel elements by chemical causes) c) High consequence: release with consequences for facility staff and possible exposure of public. In fact, event 22 (= release of fission product during transport leads together with A4 (= fault in the ventilation system of the reactor building) to a possible exposure of public. Other pathways are AJJ (= direct gamma exposure from reactor pool) or 112.113.EIcj (= failure of the reactor hall area monitoring system together with failure of the thermal column door interlock and opening of this door at power levels greater 0.025 W). The computer code SAMPLE /20/ calculated frequency distributions for these three categories by combining the median values and the corresponding error factors of the individual failure events with the Boolean equations for LOW, MEDIUM, HIGH. The frequency value for A2 (radiation exposure of public) was calculated by quantifying the relation F(A2) = P(A4)-F(A10) In summary, the following values for the three consequence categories using the SAMPLE code were obtained for the Vienna TRIGA reactor. 4-35 FIGURE 4: Probability values of three consequence groups for the TRIG A reactor Vienna frequency consequences [a" M low medium high not appl. A7 release of fission products from fuel elements •l(T4-l(r7 x A9 disturbances in reactivity AJQ fault of fuel element transfer cask -5x10" A|j direct gamma exposure from reactor pool -8x10" Aj2 direct gamma exposure from other sources .10-4-10-6 x Aj3 release of activity from experiments Ai4 release of ion exchange resin A ig loss of primary coolant from reactor tank -10"5 x Ai7 loss of primary coolant by failure of primary circuit components other than tank -3x10 -2 A4 fault in ventilation system (under pressure, filtering) -0.4 d_1 (-85% HE!) From these values, SAMPLE determined a frequency distribution by Monte Carlo simulation shown in Figure 5. Pointvalue Upper bound value Sensitivity analysis **************** Danger to the environment of the reactor facility 4-36 FIGURE 5: Probability of three consequence groups for the TRIGA reactor Vienna 10 10 w to" u O >> !_ V Q. 10" • -O o 10" **************** **************** **************** 10" 10" LOW MEDIUM HIGH CONSEQUENCES 4.2 Interpretation of Results In spite of generally very conservative assumptions it has been demonstrated that the TRIGA reactor Vienna fulfills the generally accepted safety criteria for a research reactor to a high degree. The calculated off-site exposure probability being approxi mately 1x10" a" justifies this optimistic assessment. The calculated frequencies for on-site exposure range between 10"4 to 10~3 a"1. When taking these values into account it hs always to be kept in mind that "exposure" is defined here as an exposure exceeding the international accepted ICRP values of 5 rem/a whole body dose for occupational staff and 0.167 rem/a for public. Any other limit would be unjustified and may lead to undesired speculations. 4-37 4.3 Examples of Accidents Involving Radiation Release 4.3.1 Failure of a single fuel element in the reactor tank To analyze this accident and its radiological consequences to persons, it is assumed that the fuel element with the highest power density (2.6% of total power) fails after a continuous operation of four years at nominal power (250 kW). This is equal to 1. MWy of power production. Under these assumptions the inventory of volatile fission products accumulated in the fuel has been calculated. The release of fission products from U-Zr-H fuel has been investigated as a function of fuel temperature /18/ by General Atomic and it was found that in the temperature region below 400 °C the release is controlled by fission product recoil processes. This means that only fission fragments produced in a thin surface layer of fuel may be released because of the high energy of the fission process. The amount of released fission products is proportional to the surface of the fuel. Starting at about 400 °C this recoil release is more and more covered by a temperature dependent diffusion process of fission products from the inside of the fuel material. As in the TRIGA reactor Vienna the maximum fuel element temperature reaches only 360. °C during a 250 MW transient, only about 1.5xl0~3% of the volatile fission product inventory is released to the gap between fuel meat and cladding. Further, it was assumed that the fuel element with the highest power density (2.6%) releases 100% of the noble gases and 50% of the halogènes. For the calculation of the radiological consequences several assumptions had to be made which are summarized in /19,21/. The results are shown in Figure 6. In the reactor building the following activity concentrations were calculated assuming no ventilation and no leakage of the reactor hall noble gases 6.1xl0~12 Ci/cmr halogens 2.1xl0~*3 Ci/cm3 (I-131 equivalent) According to radiation protection limits for Austria the maximum permissible concen tration (168-hours value) is 1-131 IxlO-13 Ci/cm3 Kr-85 3x10~12 Ci/cm3 Xe-131m 4xl0"12 Ci/cm3 Comparing these values it can be seen that the activity concentration inside the reactor hall is in the range of the maximum permissible values for the institute's staff and preventive measures such as controlled ventilation or access restrictions have to be taken. 4.3.2 Loss of coolant accident The probability of the loss of primary coolant was calculated in chapter 3.3. In the present chapter the radiological consequences of such an accident are investigated. With the low power density of the fuel elements in the Vienna TRIGA (6.5 kW for the element with the highest power density) the maximum fuel temperature upon loss of coolant is 275. °C. Therefore, there is no temperature induced fission product diffusion and no melting of the fuel or the cladding. 4-38 FIGURE 6: Whole body dose through different pathways as a function of distance from the reactor Assumption: Failure of one single fuel element in the reactor tank -5 1o DOSE FROM GROUND CONTAMINATION 1o -7 1o INHALATION DOSE - FROM RESÜSPENDED NUCLIDES 1o a O Q :» Qo 03 W -9 tj 1o O _L- J_ 1 X i 1oo 2oo 3oo 4oo 5oo 6oo loo 1ooo DISTANCE FROM REACTOR FACILITY IN /m/ According to /4,5/ the dose rate of out the empty reactor tank as a function of time after the hOCA is listed below. direct radiation level above scattered radiation at reactor time after 1 LOCA empty tank in [R.h-1] hall floor level [R.h" ] J 10 s 2.5xlO 0.65 1 day 3xl02 0.075 1 week 1.3xl02 0.035 1 month 3.5X101 0.01 4-39 Therefore, in case of a LOCA the direct radiation beam must be avoided, the control room and the reactor hall area is accessible for a limited time several hours after the accident. The public will not be affected in this case as the direct beam is collimated upwards towards the roof the building, the scattered radiation is even lower as inside the reactor building and the institute's compound is not accessible for public without permission. 4.3.3 External events External events on the TRIGA reactor Vienna^have been analyzed on the basis of the complete destruction of the reactor building and of the fuel elements. This means that the volatile fission products are released immediately to the environment without delay by the cooling water or the ventilation system. The total fission product inventory after 1 MW-year is about 2x10^ Ci. If only volatile nuclides with a half life above 0.5 hours are considered the inventory is reduced to lxlO5 Ci. For the accident analysis it is assumed that 100% of the noble gases and 1.5xl0~3% of the halogens are released. The noble gases account mostly for the gamma-submersion dose while the halogens are responsible for the thyroid dose. For the atmospheric dispersion model the following assumptions were made: stable weather conditions (stability class F) wind speed 1 m/s release duration 3 hours release height 0.5 m The following nuclides were considered for this analysis: Kr-83m, 85, 85m, 87, 88 1-131, 132, 133, 134, 135 Xe-131m, 133, 133m, 135, 135m, 138 The release factors are ÎNG - 1 fH - 1.5x10° With these factors and data the atmospheric dispersion program /19/ calculated the following results: The total dose (immission + inhalation + gamma contamination) within 24 hours after the accident is about 2.7x10"' rem in distances between 100 to 200 m to the reactor building in the main wind direction. The dose decreases below 1x10"* rem at the distance of the nearest appartement houses (see Fig. 7). It is evident that the radiation levels will be much higher at the destruction site, but it is almost impossible to calculate doses or dose rates as they depend on the particular arrangement of debris. One important fact is that no fuel melting will occur due to the fuel power density. 4-40 FIGURE 7: Whole body dose through different pathways as a function of distance from the reactor Assumption: Complete destruction of the reactor building, the shielding construction and all fuel elements INHALATION DOSE FROM CLOUD 1o DIRECT DOSE FROM CLOUD 10-V -3 10 1o- 4 DOSE FROM GROUND CONTAMINATION 1o -5 u en o 10- 6 Q » a o INHALATION DOSE FROM RESUSPENDED NUCLIDES a a 1o -7 oj 1o J. _L 1oo 2oo 4oo 3oo 600 îoo 1ooo DISTANCE FROM REACTOR FACILITY I>!'m ' 5. SUMMARY In the present report a probabilistic safety assessment of the TRIGA reactor Vienna was performed investigating different pathways of radioactivity release to the environ ment As top event a radiation exposure of persons (staff, public) was defined A master logic diagram was developed and individual fault trees which may lead eventually to radiation exposure were .established. Many of fault trees are facility related depending on the design of systems such as primary coolant or ventilation system. Generally, it was found that there is a great lack of relevant data which could be improved by better information exchange among the TRIGA community, with assistance or coordination of General Atomic, the TRIGA manufacturer. In chapter 4 three types of accidents with activity release or potential radiation exposure have been analyzed using facility related data to illustrate the consequences of such accidents analyzed in chapter 3. 4-41 REFERENCES /l/ H.Böck et al.: Use of a Low Power Research Reactor in National and International Research. Int.Symp. on the Utilization of Multipurpose Research Reactors and Related International Cooperation. IAEA-SM-300/048, Grenoble 19.-23.10.1987 /2/ PRA Procedures Guide. A Guide to the Performance of Probabilistic Risk Assessments for Nuclear Power Plants. NUREG/CR-2300, U.S. Nuclear Regulatory Commission, Review Draft, September 1981 /3/ ORCHARD, A Fault Tree Analysis Program. Draft Program Manual, 1986 /4/ Hazards Report for the Austria 100. kW TRIGA Mark-II Pulsing Reactor, General Atomic, January 1961 /5/ Sicherheitsberichts des TRIGA Mark-II Reaktors, Atominstitut der Österreichi schen Universitäten, Wien Mai 1988 /6/ Credible Accident Analysis for TRIGA and TRIGA-fueled Reactors. U.S. Nuclear Regulatory Commission, NUREG/CR-2387, April 1982 III Reactor Safety Study. An Assessment of Accident Risks in U.S. Commercial Nuclear Power Plants, WASH-1400 (NUREG-75/014), U.S. Nuclear Regulatory Commission, October 1975 18/ Betriebsdokumentation für den TRIGA Mark-II Reaktor des Atominstituts der Österreichischen Universitäten (Zeitraum 1977 bis 1985) 191 Fehlerereignistabelle für den TRIGA Mark-II Reaktor Ljubljana (Zeitraum 1966 bis 1986), aus: IAEA Contract No. 4395/RB, Progress Report, March 1987, "Jozef Stefan" Institute, Ljubljana, Yugoslavia /10/ A.Swain, H.Guttmann: Handbook of Human Reliability Analysis with Emphasis on Nuclear Power Plant Applications, Final Report. NUREG/CR-1278, August 1983, U.S. Nuclear Regulatory Commission /ll/ Combining Risk Analysis and Operating Experience. IAEA-TECDOC-387, IAEA, Vienna 1986 /12/ IEEE Guide to the Collection and Presentation of Electrical, Electronic and Sensing Component Reliability Data for Nuclear Power Generation Stations, IEEE Std-500 (1977) /13/ Risikostudie SAPHIR: Probabilistische Analyse einiger Sicherheitsaspekte des Schwimmbadreaktors SAPHIR, Eidgenössisches Institut für Reaktorforschungs, Würenlingen, Dezember 1980 /14/ Applications of Probabilistic Analysis Techniques to a Typical 10 MW MTR. Safety and Reliability Directorate, UKAEA, IAEA-TECDOC-400, p. 97 ff., Vienna 1986 /15/ Siting of Research Reactors. IAEA-TECDOC-403, IAEA, Vienna 1987 /16/ C.J.Everline, M.G.Stamatelatos: Probabilistic Frequency Assessment for Radiological Releases from a 1 MW(t) TRIGA Mark-II Reactor under Seismic Conditions. GA Technologies Inc., San Diego, May 1987 4-42 /17/ N.McCormick: Reliability and Risk Analysis. Methods and Nuclear Power Applications. Academic Press, New York 1981 /18/ N.L.Baldwin, F.C.Foushee, J.S.Greenwood: Fission Product Release from TRIGA-LEU Fuel. GA-A-16287, November 1980 /19/ M.Strubegger "STRISK" - A Program to Calculate Radiological Consequences due to Emission from Nuclear Facilities. Diploma Work, TU Wien, September 1978 /20/ SAMPLE: A Code for Determining the Distribution and Confidence Limits by Simulation. Reactor Safety Study, App. II, WASH-1400, October 1975 /21/ H.Böck, C.Kirchsteiger: Probabilistic Safety Analysis for the TRIGA Reactor Vienna, AIAU 88305, July 1988 4-43 RECENT RADIATION, PROTECTION PROBLEMS AT THE VIENNA TRIG A REACTOR Manfred TSCHURLOVITS, Hans SLANETZ Atominstitute of Austrian Universities, Vienna, Austria Abstract The work of the radiation protection group at the Atominstitute TRIGA- reactor since the last report in these meetings is described in brief. The work of this group was focussed to the following issues: + occupational monitoring + release monitoring + environmental monitoring Although no substantial changes in the programs were required in the final licencing procedure, which was carried out last year, some minor changes have to be incorporated into the routine monitoring program. These changes are described in the paper. No unplanned events happend in the reporting period. 1. Introduction Radiation protection issues of the Vienna TRIGA reactor are reported already in some meetings of this series / 1,2 /. This paper continues the reporting of radiation protection question arising in reactor operation and gives some status report. The changes in the radiation protection program were not very substantial and were mainly caused rather by formal requirements than real needs for improving protection. However, two major events happened in the reporting period which had some consequences for the work of the radiation protection group: * maintenance work close to the reactor core * realisation of the final licencing procedure Where the first issue included some aspects to routine operation, as associated with fuel element handling, installation of a new irradiation unit etc. , the second point lead to some minor changes in the routine program as occupational monitoring, release monitoring and environmental monitoring. 2. Radiation protection at maintenance work In 1985, some work was carried out to install an equipment for neutron radiography to replace the source of cold neutrons no longer in use. Since this took place very close to the core, all fuel elements had to be removed from the core to provide a radiation level 4-45 suitable for work. At this occasion, some maintenance work was done by the reactor group which is reported elsewhere. Since this was not the first time that all fuel elements were moved from the core to a storage position, experience gathered in the first time ( August to October 1966 ) / 3 / was used advantageously. In particular, it was obvious that a time interval of about 4 weeks is required to allow for sufficient decay of a fuel elements to an activity of app. 370 GBq ( 10 Ci ) of fission products, which can be handled in the reasonable effort with the available transport container ( external dose rate at the surface of the container: .2 to 1 mSv/h ). After this work, the source for cold neutrons, which was installed 1966, was removed from the irradiaton position. The dose rate at the beam hole with the equipment still in was up to 1,5 mSv/h. The dose rate at the surface of the dismantled material was up to 1,2 mSv/h, mainly Co 60. The dose rate at the open beam hole was ranging from 2o mSv to 1 mSv, dependent from the distance from the reactor core ( Fig.l ). Some work has to be done inside the beam hole as decontamination and inspection as well. Obviously, this work was carried out with additional protective clothing. After this work, the new radiography unit designed by the experimental group was put into the beam hole. / Fig.2 /. Eventually, the fuel elements were returned into the core. Regarding radiation exposure at this work, the collective external dose of the staff ( a total of 16 persons was involved in this work over 4 month ), as assessed by personal dosimeters, was 13 mSv, ranging from 0,1 to 2 mSv. This is clearly well below the limit of 4 mSv per month. No incorporated activity was assessed. 3. Radiation protection at routine operation The routine work of the radiation protection group can be seperated into the following fields. * occupational monitoring * release monitoring * environmental monitoring * calibration and maintenance of protection instrumentation Regarding occupational monitoring, measurement of external exposure of the reactor staff ( operational and experimental staff )is carried out by TLDs, where both Li 6 and Li 7 cristals are used, which are evaluated by a dosimeter service. The measurement of the fast neutron component is done by means of rem counter in regular intervals. This is necessary because it is obvious that the measurement of neutron dose with TLD is rather an indicative method than real personal dosimetry, but experience has shown that this is a reasonable compromise. Some results of the measurements are shown in fig.3. Release monitoring consists of the routine monitoring of noble gases, iodine and particulates. A schematic diagram of the exhaustive system is shown in fig.4. In addition 4-46 to the assessment of the activity concentration / 4 /, it was necessary to assess in addition the time integrated activity. This is useful for estimation of the total release of Ar-41. Estimates on the total release lead to a figure for the normalized release of about 1,5 GBq/MWh. This is substantially lower than the figure given in / 4 / but some changes of the installation in the beam holes took place since 1978. Another program was concerned with the assay of C-14 released from the TRIGA reactor, and the results are mainly reported / 5 /. However, it might be worth to note that the release differs substantially dependent from the operation conditions ( Fig. 5 ). The handling of the liquid releases is done by the radiochemical unit. The environmental monitoring program, which takes into account that the reactor is located in an urban area, was slightly changed by the licencing procedure. The extension was mainly directed to include gammaspectroscopic measurements of vegetation grown in the vicinity of the institute. In addition, it was required that TLDs become located at 4 sites around the reactor, the exposure interval being 2 month. This interval was changed to six month, because the pre-dose of the dosimeters was in the same order than the the expected dose. The results of these measurements show obviously only statistical deviations . This applies also for the measurements of the environmental samples. However, the only real response of the monitoring system was the detection of the Chernobyl fallout by means of instrumentation operated for the assessment of releases of our reactor. Figs. 6 and 7 show the external dose rate as well' as the activity concentration in air at the AI- site. Regarding the regular checks, some data sheets were prepared to make a reasonable documentation of radiation protection issues as required by licencing. A monthly radiation protection report is prepared including the following issues: external dose rate monitoring instrument check calibration of instruments adjustment of thresholds contamination check activity concentration of primary water measurements airborne releases assessment results of environmental monitoring check of the emergency equipment result of personnel monitoring of the reactor staff unplanned events. 4-47 References / 1 / E. Tschirf, M. Tschurlovits: Proc. 4th TRIGA Users Conf. Vienna ( 1976 ) / 2 / H. Bock, J. Hammer, E. Tschirf, M. Tschurlovits: Proc. 8th Europ. Conference of TRIGA Reactor Users, Espoo, 1984 / 3 / W. Jeschki, E. Tschirf: Arbeiten am Core des TRIGA Mkll und dabei auftretende Strahlenschutzprobleme. Internal Report AI, 1966 / 4 / E. Tschirf, M. Tschurlovits: Proc. 5th TRIGA Users Conf. Portoroz ( 1978 ) / 5 / M. Tschurlovits, K. Pfeiffer, D. Rank: Proc. 6th TRIGA Users Conf. Mainz 1980 / 6 / K. Buchtela et al: Preliminary Report on Measurements of Environmental Radioactivity of the Chernobyl Accident: AIAU Report 86602, May 1986 / 7 / M. Tschurlovits, E. Unfried: Proc. XIVth Regional Congress of IRPA, Kupari, 1987 r ^ « m •m * • 0 * « (J * » o • • CCNCRCTE SHIELD 0 Q i CCfCROÎ « CORE j % O C a 1 10 20 *& • 4> • n •* 0 dose rate mSv/h *• a » 1 • • • .\ •* , 1 * •*-*.* o * * - » • • •* •J ,0 ,1« Fig.l: Dose rate [ mSv/h] at the open beam hole 4-48 rCore conical Si-filter collimator Fig. 2: New radiography unit f [ ; i i Ht3 -Ht1- Misch* iihltr POLARISIERTE NEUTRONEN TRIOA- REAKTOR WIEN Juli MB -TVT Fig.3: External dose rate in ( neutrons and gamma ) the reactor hall 4-49 ABLUrT "PHYSIK" 4- _,jAerosol- und Ae V^H Jotjsamml er Jo e Filter tf. y Gebläse Gebläse filter Edelgas- cessung Abluft Abluft Filter Absaugleitung Tankoberfläche Gebläse Messung A Aerosol(«ssunrW—e ' REAKTOR Strahlrohr- /*_ Zuluft Entlüftung Zuluft Abluft Abluft Absaugleitung für Argon <*1 Zuluft Zuluft "Chemie" "•*"•.• è«. blase Zuluft Filter Fig. 4: Exhaustive system of the reactor hall lb is ' 26 ACTIVITY CONCENTRATION Ul«"') Fig. 5: Lognormal distribution of specific activity of C-14 in exhaust air 4-50 activity concentration (Bq/cubic meter) time from 29.4.1986, 0:00 (hours) Fig.6: Activity concentration in air from Chernobyl releases ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^;j Fig 7.: External dose rate at the reactor site from Chernobyl releases 4-51 RADIOLOGICAL ENVIRONMENTAL IMPACT RELATED TO REACTOR ACCIDENTS i , V / X * 1 S. Altieri , P. Pedroni , S. Meloni 1) Laboratorio Energia. NucUare Applicata, Univeraità di Pavia, Pavia (Italy) 2) Itiituto Nazionale di Fiâica, Nucleare, Sezione di Pavia, Pavia (Italy) INTRODUCTION In the 1986 European Triga Users Conference, held in Rome, a review of two reactor accidents was reported [1-2], namely loss of reactor coolant and unattended and sudden reactivity insertion. The review was performed in connection with the renewal of the reactor operation license. In both cases the temperature rise is not so high to induce a damage in the fuel cladding. The present paper deals with a new accident hypotesis: fission products release from an unobserved cladding failure in a fuel element after a long reactor run at full power, while one of the previously cited accidents is occurring. FISSION PRODUCT ACCUMULATION AND RELEASE It is assumed that a fission product release may occur from a failure in a fuel element, while the 250 Kw LENA Triga Mark II reactor is in the conditions described in Table 1, which were prudentially chosen conservative. The evaluation of the amount of cumulated fission products in the hottest fuel element was performed by means of the ORIGEN calculation code [3j. TABLE 1 Fission product accumulation in the hottest fuel element: reactor and element parameters Quantity of 'JiiäU 39 g Thermal neutron äux l.lO^cm'* s *J Burn-up 122 Mwd Reactor run at 250 Kw 488 d The fraction of fission products gathering in the gap between fuel and cladding has been evaluated according to Baldwin et al. [4j: -1.3<-10 + * fi(T) = 1.5-10-5 + 3.6-10+3-e-"T (l) where T is the fuel temperature (#). 4-53 In the calculation of fx it was conservatively assumed that maximum temperature is reached by the whole mass of the fuel element. Table 2 gives the values of /x for the hypotized reactor accidents. TABLE 2 Fission products released in the gap Reactor accident Max. fuel h temp. (°C) reactivity insertion 447 4.48 • 1CTÖ total coolant loss 405 2.44 • HT6 In case of cladding failure, however, only a fraction, /2, of gaseous fission products accumulated in the gap may be released outside the plant. If the cladding failure oc curs during the loss-of-coolant accident, gaseous fission products could directly diffuse in the reactor room. However, in this emergency condition, an auxiliary air ejection system, which includes also absolute and activated carbon filters, would be powered. Absorption efficiency of activated carbon is 99 % for iodine,, and therefor /2 is taken as 0.01 for iodine radionuclides and as 1 for the other gaseous fission products. The airflow of the emergency air ejection system is 900 m3 • h~l: it takes 6 hours to change the air of the reactor room. In the fission products release calculations, it was assumed that release occurs at constant rate during 6 hours. In the case that the cladding failure is induced by an unattended and sudden reac tivity insertion, a fraction of the gaseous fission products would dissolve in the water of the reactor tank. It was assumed that noble gases are all released outside the plant (/i2=l), whereas for halogens f2 was taken as 0.5 except for iodine radionuclides for which /2 was taken as 0.005 because of their absorption on activated carbon filter. In Table 3 the activities of most abundant gaseous fission products are given for the chosen fuel element, as well as the activities that might be released outside the plant in the case of the two hypotized reactor accidents. 4-54 TABLE 3 Activities (Bq) of the most abundant gaseous ßssion products in the chosen fuel element Radionuclide cumulated released out released out released out in the element (loss of cool. ) (reactiv. ins. ) (dose calcul.) 06 131mX« 9.1 • 1010 2.22 • lO06 4.07 • lO 2.9 • 10ü7 u 07 07 ÜS 133m Xe 6.2 • 10 1.51-10 2.77 • 10 1.9 • 10 133X« 2.6 • 1013 6.26 • 10us 1.15 • 10U9 8.1 • 1009 13ÔX« 1.0 • 10i3 2.52 • 10us 4.62 • 10U8 3.3 • 10ü9 ^Cs 4.2 • 1011 1.02 • 10ü7 9.44 • lO06 1.3 • 1008 l2äCs 1.9 • 1011 4.73 • 10ue 4.35 • 10U6 6.1 • 10U7 i37CV 1.6 • 10« 3.92 • lO07 3.60 • 1007 2.6 • 1008 l2*C$ 2.2 • 1013 5.43 • lO08 4.98 • lO08 7.0 • 1009 82 Br 6.2 • lO09 1.53 • 100& 1.41 • 10o& 1.9 • 1006 ™Br 2.0 • 10« 4.S3 • 10ü7 4.43 - lO07 6.3 • 10ÜS S45r 3.6 • 10« 8.71 • 1007 7.99 • 10°' 1.1 • 10U9 a3mÄ> 2.0 • 10« 4.83 • 1007 8.86 • lO07 6.2 • 10oa **mKr 4.8 • 10« 1.17 • 1008 2.15-1008 1.5 • 10U9 11 06 06 07 nKr. 1.8 • 10 3.97 • 10 7.29 • 10 3.9 • lO a7Kr 9.7 • 10« 2.36 • lO08 4.32 • 1008 3.0 • 1009 s*Kr 1.4 • 10" 3.38 • 10üa 6.19 • 10oa 4.1 • 10ua i3ij ' 1.1 • 1013 2.77 • 1006 2.54 • 1006 3.6 • 1009 132j 1.7 • 10i3 4.14 • 10ü6 3.80 • 1006 5.2 • 10U9 133 J 2.6 • 10i3 6.26 • 10üe 5.74 • 1006 7.8 • 10ü9 134J 3.0 • 1013 7.38 • 10üe 6.77 • 10ÜG 9.3 • 10üö L36J- 2.4 • 10i3 5.77 • 10üe 5.29 • 1006 7.4 • 10ü9 For dose calculation the activities reported in the last column of the Table 3 were used, which were chosen according to the following conservative assumptions: a- the evaluation of the fraction of gaseous fission products gathering in the gap was worked out at the highest temperature, 477 °C, in connection with the unattended 5 and sudden reactivity insertion reactor accident. Therefore fx = 2.44 • 10~ ; b- it was assumed that all the gaseous fission products in the gap be released out side the plant, without considering the effect of activated carbon filters on iodine radionuclides; therefore /2 = 1 for all gaseous fission products; c- the evaluated activities were further increased by considering a final safety factor, /3 = 7. DOSE CALCULATION The calculation of the environmental radiological impact due to gaseous fission prod ucts release at the plant stack, was carried out by following the radioactive cloud diffu sion according to the Pasquill- Gifford model [5-6]. The concentration at soil level (total activity) of the cloud in the absence of wet deposition, at a P(x,y,z) point downwind with respect to the release point, is given by: Q(t) ïïr+£? r(x,y,0) c (2) •Kffy 4-55 Q{t)= released activity in Ci for an instantaneous release or in Ci/s for a continuous release u= average wind speed in m/s along the i axis h= height (m) above ground of release point y= distance (m) from the cloud axis ay= Pasquill-Gifford coefficient (m) for the atmosferic diffusion along y axis CT,= Pasquill-Gifford coefficient (m) for the atmosferic diffusion along z axis. In case of wet deposition, the concentration along the x axis is given by: vy(x)-T{x) (3) where: *(*) 2iî) (4) w 1 + 0.5 • i(s) -ü-S ">& = * «<**-<' (5) ft • Ä — '»com — — • Vabb (7) and, taking into account the building effect: *"•+% • (9) where: Vimp^ cloud depletion speed (m/s) «ow.= deposition speed of the cloud particles (m/s) 2 S— Led • -ff«r-the building cross-section (m ) H~ height above ground (m) of the basis of inversion layer. Equation (3) has been numerically solved using the DRKSTP of CERN library. Once that the T{x, y,Q,t) value is known, the concentration of the i-th radionuclide may be obtained by: [CnteOli-rfo «)•$,(*)• «-*<•'- (10) where: Qi(t) = 0 for t < tr Qi{t) = [Rate]i • At for tr < t < tmax + tr Dose and contamination values at the ground level were obtained by: SteO-ÇlCfrOH^ (H) Cn(ar1j)=^[C7„(2)i)]<'k]i (12) 4-56 where: [-Fdl»= conversion factor for unit concentration for the i-th radionuclide [vg]i= deposition speed at ground level for the i-th radionuclide. TABLE 4 Input parameters for cloud diffusion Meteorological conditions F Category Wind speed 2 m/s Depletion speed 0.003 m/s Deposition speed 0.01 m/s Building size (17 x 17)m Release height: Om (ground) ISm (stack) Radioactive source: continuous release for 6 hours The following quantities were evaluated for all radionuclides reported in Table 3 as a function of time and distance from release point: a) Effective dose equivalent from external irradiation for people directly exposed to cloud; conversion factors from unit concentration in air are reported in Table 5 [7-8]; b) Effective dose equivalent commitment from inhalation for child (1-10 years age) and for adult (greater than 18 years age); conversion factors from unit concentration in air are reported in Tables 6 and 7 by assuming the following respiration rates: 6.94- HT5™3*-1 (infant), 2.17- i c) Effective dose equivalent commitment from inhalation to thyroid for child and adult; conversion factors from unit concentration in air are reported in Table 8 [7-8]; d) Collective effective dose equivalent from external irradiation and inhalation for adult and child living within 3 km from plant; e) Soil contamination for caesium and iodine radionuclides. 4-57 TABLE 5 Conversion factors for effective dose equivalent is te from external irradiation 3 l 3 Radionuclide v • 3 • Bq~l • m~ rem • s • Ci~ • m~ i3lmXe 3.58 • 10ie 1.33 • 103 l33mXe 1.45 • 10iô 5.36 • 103 ™Xe 1.69 • 10" 6.26 • 103 ™>Xe 1.26 • 1014 4.68 • 102 i34 7.26 • 1014 2.69 • 101 l36c* 13 x C3 . 1.02 • 10 3.77 • 10 i37Ca 4.59 • 1014 1.70 • 101 13*Ca 1.19 • 1013 4.41 • 101 "Sr 1.26 • 1013 4.66 • 101 835r 3.53 • 10i6 1.31 • 103 S4Sr 9.06 • 1014 3.35 • 101 *3mKr 9.86 • 1019 3.65 • 106 *àmKr 8.19-1010 3.03 • 102 ihKr 1.28 • 10lb 4.74 • 104 87#r 3.92 • 10i4 1.45 • 101 i&Kr 9.77 • 1014 3.69 • 101 13ij 1.81 • 1014 6.71 • 102 132j 1.07 • 10i3 3.96 • 101 " ' ""133 J 2.86 • 1014 1.06 • 101 i34j' 1.24 • 1013 4.58 • 101 135j 7.81 • 1014 2.88 • 101 TABLE 6 Conversion factors for effective dose equivalent commitement from inhalation for children Radionuclide Sv Bq~l Sv-3- Bq'1 -m-3 rem • s • Ci~l • m-3 7 134C3 .86 • 10 .1866- 1010 69.042 i3ÖC3 .14 • 107 .3038 • 10u 11.241 137C3 .50 • 107 .1085 -10iu 40.145 i3&Cs .20 • 109 .4340 • 10i3 0.161 131J .87 • 107 .1888 • 10iü 69.856 i32j .84 • 109 .1823 • 10i2 0.675 133 jr .15-107 .3255 • 10u 12.044 134j .26 • 109 .5642 • 1013 0.209 135J .31 • 10a .6727 • 1012 2.489 4-58 TABLE 7 Conversion factors for effective dose equivalent commitement ' from inhalation for adults Zadionuclidt i Sv • Bq~l Sv-3- Bq'1- -m-' rem -3 • Ci~l • m-3 13*Cs .10 • 10' .33 • 1011 12.32 i36Ca .16 • 10a .53 • 10" 1.97 a u MC3 .75 • 10 .25 • 10 9.24 i38 131J .87 • 10s .290 • 1011 10.72 132 j .89 • 101U .296 • 1013 0.11 i33j .15 • 108 .499 • 10" 1.85 i0 14 ... i3i^... .30 • 10 .999 • 10 0.037 i3Sj .31 • 109 .103 • 10" 0.38 TABLE 8 Conversion factors for effective dose equivalent commitement to thyroid \ ladionuclid J Sv-Bq-1 Sv-3-Bq~x -m-3 rem • 3 • Ci~l • m 3 i3ij .29 • 10e .966 • 101U 357.31 132 J" .17-10a .566 • 1012 2.09 iSZj .49 • 107 .163 • 1010 60.37 134J .29 • 109 .966 • 10i3 0.36 """ Ï35"j" .85 • 10a .283 • 1011 10.47 131/ .29 • 10s .629 • 109 2328.4 ~I32j .17-107 .369 • 1011 13.649 TSS'f .49 • 106 .106-10" 393.31 1341 .29 • 10a .629 • 10" 2.328 1SSJ .85 • 10iü .185 • 10io 68.265 131'j .43 • 100 .298 • 10s 1104.15 132j- .25 • 107 .174 • 10u 6.419 ' 133 j """' .73 • 106 .507 • 10iü 187.44 134j ' .43 • 10a .298 • 10" 1.104 l35j- .13 • 10e .902 • 1011 33.38 RESULTS Doses and contamination. Dose trends vs. time and distance from release point are reported in Fig 1 for dif ferent population age classes. Soil contamination by caesium and iodine radionuclides is given as well. Release at soil level and at chimney (15 m) were considered. Collective dose. Collective dose to population living in the eight circle sectors was evaluated at dif ferent distances from source point by using the effective dose equivalent commitment data. Each sector was assumed to be 100 % downwind; dose was evaluated for child (1-10 years age) and adult (grater than 18 years age). The distribution of population in the sector is reported in Table 9. Total collective dose to population living within 3 km from the reactor is shown in Fig 2. Figures 3-6 show collective doses to children and adults living either within 500 m or 3 km from reactor. All doses were calculated assuming a radionuclides release at ground level and include only contributions from external irradiation and inhalation. In order to obtain an overall evaluation of computed dose estimates, the reported 4-59 data may be compared with derived reference limits given by the Italian Government for nuclear emergency [9] as shown in Table 10. As far as doses are concerned the computed data are quite below the derived refer ence limits; for instance the dose to thyroid due to inhalation in about 60 times lower. Should the hypotized accidents occurs no emergency condition needs to be started. As far as soil contamination is concerned , the limits for 137Ca are not overcome at any distance, whereas in the case of 1317 contamination, the derived limit for eggs is overcome within 100 m from reactor, and the one for milk within 2000 m from reactor. At the moment no egg or milk production is going on at the reported distance limits from the LENA plant. TABLE 9 Population distribution in the eight circle sectors around the LENA plant Sector Inhabitants 1 454 2 4303 3 22204 4 32014 5 4040 6 2270 7 1100 8 50 TABLE 9 Derived reference limits for nuclear emergency Hose Exposure Critical Population Individual Collective description organ group ref. dose ref. dose (mSv) (man Sv ) Any Exposure: •whole body children 150 pregnant mothers 150 2. • 103 adults 250 137Ca (introduction) whole body whole population 100 2. • 103 1311 (introduction) thyroid whole population 250 1. • 104 Soil contamination Radionuclide milk vegetables eggs ~ Bq • m-2 Bq • m~2 Bq • m~2 adults 131j 1.04 • 106 7.77 • 107 3.15 • 106 - 137 Ca 3.26 • 106 5.18 • 10a 5.18 -107 - children i'iij 4.07 • 104 2.26 • 107 7.03 • 10° - 137 Ca 8.14-106 8.88 • 105 7.03 • 107 - 4-60 O RELEASE AT SOIL A RELEASE AT CHIMNEY (15m} -3 £10 B~ CHILDREN THYROID o -» » » ê * 1 O 10 m Ai 500} m 2 4 6 8 10 2 4 6 8 10 hours from the re!, beg. Km from the reactor 2 4 6 8 10 2 4 6 8 10 Km from the reactor Km from the reactor £ 1 -1 °10" -2 10 -3 10 •4 10 -5 10 2 4 6 8 10 2 4 6 8 10 Km from the reactor Km from the reactor 7 o- 10' 10 ö 10 £ic , 5 • o* ** .5 10 m 10 10- CAESIUM CONTAMINAI. 10! 10; 10' io; 10 io: 1 2 4 6 8 10 2 4 6 8 10 Km from the reactor Km from the reactor Fig 1 Dose equivalent commitement by inhalation and external irradiation for child and adult and soil contamination (cSv = Sv • 10~2j 4-61 Fig 2 Collective dose (man Sv) for total population living within 3 Km from reactor 4-62 Fig 3 Collective dose (man Sv) for children (0-10 years age) living within 500 m from reactor 4-63 Fig 4 Collective dose (man Sv) for children (0-10 years age) living within 3 Km from reactor 4-64 Fig 5 Collective dose (man Sv) for adults (greater 10 year3 age) living within 500 m from reactor 4-65 Fig 6 Collective dose (man Sv) for adults (greater 10 years age) living within 3 Km from reactor 4-66 REFERENCES [1] S. Altieri, P. Pedroni - L'incidente di perdita di refrigerante nel reattore Triga Mark II da 250 Kw (aprile 1988) [2] S. Altieri - L'incidente di inserimento incontrollato di reattività nel reattore Triga Mark II da 250 Kw (maggio 1988) [3] M. J. Bell - ORIGEN - The ORNL Isotope Generation and Depletion Code 4628 [4 N. L. Baldwin, F.C. Foushee, J.S. Greenwoord - Fission product release from Triga - Leu Reactor fuels. General Atomic, G.A. - A 16287 (novembre 1980) [5] D. Dini, L. Pucciarelli LIDIA Un programma di calcolo per la valutazione delle conseguenze radiologiche dei rilasci accidentali. [6] A. Cuoco, R. Galvani, L. Pucciarelli, R. Scafè Determinazione delle concentrazioni di una nube radioattiva e delle contaminazioni del suolo secondo il metodo di Pasquill- Gifford. CNEN RT/PROT (65)10. [7] D.C. Kocher Dose rate conversion factors for external exposure to photon and elec tron radiation from radionuclides occurring in routine releases from nuclear fuel cycle facilities. Health Physics vol. 38 n. 4 (aprile 1980) [8] F. Breuxer, C. Brofferio, A. Sacripanti - Equivalenti di dose impegnata da intro duzione di attività unitaria, per quattro classi di età relative agli individui apparte nenti alla popolazione per lo studio dell'impatto ambientale degli scarichi radioattivi - ENEA -RT/PROT (83)24. [9] Ministero degli Interni-Circolare n. 70 8/8/1973 4-67 FISSION PRODUCTS DISTRIBUTION FOR ISTANBUL IN AN HYPOTHETIC REACTOR ACCIDENT A. R. BAYÜLKEN M. C. BARLA I.T.U. I.T.U. Institute for Nuclear Energy Faculty of Aeronautics and Astronautics ABSTRACT As it is well known by all TRIGA users, Istanbul TRIGA Mark-II reactor is situated on a hill, in the outside of the populated city area. Its geogra phical situation is 41°06'32" north lattitude, and 29O01'44" east longitude. In the reactor Safety Report..and in some other publication, the safety as pects of the reactor and I isodose distribution have already been inves tigated. In this study, the downwind distance of the maximum concentration will be investigated for different meteorological conditions of Istanbul. The particle size distribution effect will also be investigated. 1- Introduction Some of the many contributors to uncertainty in reactor accident consequence assessments are the wind speed, the wind direction and the height of the stack. The particle size distribution is also another major contributor to this un certainty. The wind conditions and the stack height are important for the determination of the maximum dose deposition distance . The maximum dose of the released radioactivity is also calculated accordingly. Secondly, current consequence models generally do not include the effect of a particle size distribution (PSD) in their calculation, despite the fact that wet and dry deposition in the environment, as well as deposition in the lung airways and hence internal exposure resulting from inhalation of radionuclides, all have a dependence on particle size. In what follows, all these factors are investigated for a concrete case, the hypothetic Istanbul TRIGA Mark II reactor accident. The distances of maximum dose deposition for different wind direction and speeds are calcu lated, according the statistics of all meteorological data concerning the reactor side. The particle size effect will also be investigated. 2— Release of Fission Products The fission products occured in the life of a reactor can be released to the environment in various ways. The events which are abnormal incidents that should be taken into account are as follows: 4-69 1- Cladding rupture 2- Reactivity accident 3- Loss of coolant accident 4- Human errors 5- External factors as sabotages or aircraft crashes If one of these events is occured, some amounts of gaseous or solid fission products are released to the atmosphere and dispersed into the environment. A clear explanation of some of these incidents was done in a previous inves tigation, (1) . For human error, the most famous experience is the Tchernobyl Accident. The results of this accident influenced many European countries as well as some countries in Asia. Even the human health was not seriously suffered, the economics of many countries received some negative impact from this acci dent. In the actual world of uncertainty, external factors are of great importance. An aircraft crash occured on a research reactor can produce very serious damage, because of the absence of safety buildings as in power reactors. On the other hand, sabotages became also a routine operation, specially between countries which are not very friend to each other. With all these possibilities, a release of radioactive products has not a zero probability to be occured. As an exemple, the list of noble gases and halogens produced in a 250 kW TRIGA reactor during a continous four year operation period can be given, (Table-1) , (2). TABLE - I NOBLE GAS AND HALOGENS IN THE REACTOR Isotope Quantity (Ci) Isotope Quantity (Ci) Br-83 1.020 1-132 8.850 Kr-83m 1.020 1-133 14.350 Br-84 2.060 Xe-133m 350 Br-85 2.150 Xe-133 14.3 50 Kr-85m 2.150 1-134 16.100 Kr-85 113 1-135 13.400 Kr-87 5.400 Xe-135m 4.050 Kr-88 7.700 Xe-13 5 13.850 Kr-89 9.750 1-136 12.950 Kr-90 10.850 Xe-137 12.550 Kr-91 7.350 Xe-138 11.700 1-131 5.930 Xe-139 11.800 Xe-131m 48 Xe-140 8.100 If one of these events occurs, one of the major questions to be answered is to know the maximum dose concentration and its deposition area. To see this problem for the Istanbul case, all the meteorological data available were analyzed and (Table-II) was obtained for wind characteristics . 4-70 TABLE - II WIND CHARACTERISTICS FOR ISTANBUL TRIGA MARK-II REACTOR SITE OVER 30 YEARS PERIOD p - occuring probability v - average wind speed at 10 m from ground level (m/sec) N NNE NE ENE E ESE SE SSE S SSW SW WSW W WNW NW NNW Calm Max. W.aver. JAN. P 0,09 0,11 0,18 0,03 0,02 0,01 0,01 0,02 0,05 0,07 0,12 0,03 0,03 0,01 0,03 0,05 0,12 SSW/31,5 4,30 V 5,2 5,8 7,7 4,6 3,6 2,0 2,3 3,1 2,4 3,7 4,0 3,3 3,0 2,7 4,3 4,9 — FEB. P 0,11 0,11 0,19 0,02 0,02 0,01 0,02 0,02 0,04 0,08 0,13 0,04 0,02 0,02 0,03 0,03 0,11 WNW/30 4,40 V 6,2 4,6 7,1 4,6 5,0 2,8 2,6 3,2 2,5 3,9 4,3 4,1 3,0 3,4 4,1 4,6 — MAR.* p 0,11 0,13 0,27 0,02 0,01 0,00" 0,01 0,01 0,03 0,06 0,10-0,03 0,01 0,01 0,02 0,04 Ü,13 WSW/27 ,3 4,50 V 5,1 5,1 7,2 5,0 3,7 2,0 1,8 T,9 2,1 3,8 3,9 4,4 2,5 2,6 4,7 4,4 APR. p 0,06 0,11 0,29 0,01 0,02 0,00 •u7oi""o*,or"o7o5"o;u5 ü,ii 0,05 0,02 0,01 0,01 0,03 0,15 SW/26,4 3,80 V 3,8 4,3" 6,4 3,4 5,3 1,1" 1,4 1,8 1,7 3,6 3,4 3,8 3,0 2,5 3,0 4,ü ^1AY p 0,08 0,12 0,32 0,02" 0,03 0,01 'Ö;Ö3 O,Ô6™Ô,03 0,01 0,01 0,02 0,02 WNW/ 24,3 3,40 "~4 0701 o,or"ö7oi" 0,19 V 3,4 4,0 5,5 3,7 5,1- 1,3" 1,4 1,4 1,6 3,0 2,8 3,7 2,6 1,9 2,6 3,4 — JUNE p 0,09 0,11 0,38 0,04" 0;04" "0,01" 0,01 0,01 Ö,-(T2'",O,"01'Ü,O4 0,02 0,01 0,01 0,02 0,02 0,16 NE/25,5 3,90 V 4,2 3,8 6,1 3,2 4,2 1,2" 1,2 T,0~r;"5 2,4 3,0 3,6 2,3 2,4 2,8 3,1 JULY p 0,10 0,12 0,47 0,06" "0,05J "0", 01 b7oï"07oïï""o,oi o.oi 0,61 0,01 0,01 0,01 0,01 0,02 0,10 NNW/29,4 4,80 V 4,3 4,3 6,8 5,4 4,6 1,0 1,3" O 170""1,8 "1,9 3,6 2,2 2,3 3,0 4,4 AUG. p 0,10 0,13 0,41 0,08 0,06 0,01 0701'"ö","o"i~ö7or 0701 O,ÛI O,OI0,0 1 0,01 0,02 0,02 0,10 WNW/24 ,4 5,00 V 4,4 4,7 7,1 5,3 4,3 2,8 'i',"o' " r,'f' T'X" T;'ff 2;'2 i,5 3,5 2,2 5,3 , 3,4 SEP, p 0,08 ïï,To"o,3"6''ïï;r(r'o7o7~o,oi 2.Q.X XF-2 C = ir.C ,C .u exp 4 ) (1) It should be noted from (1) that C -s- 0 as X -*• 0, that is, asthe base of the stack is approached. Thus the maximum ground concentration C occurs at some distance downwind of the source. The maximum concentration C , as well as its downwind distance X , may be obtained by maximizing the equation (1) with respect to x, leading to 2.Q C = (2) m ( ) e.ir.u.h ( h .2/2-n k C ; (3) m z . According to these equation, X and C were calculated with respect to Q, h, and n values. For the variation of Cm, the above formula was put in the following form: m C = K.Q (4) m where K is a constant which is variable according to u, h, and diffusion coefficients. Generally, wind speeds are measured at 10 m. from the ground level. To ob tain the values of these wind speeds at the height h, the following formula with respect to the stability constant of the atmosphere, is used. / h v n (5) uh " U10 ( ÏÏÏ } In the calculation of C , an angular distribution of a = 10 for wind blow , m was assumed. 4-72 The values of X and C , for a stack height of 25 m. and for n = 0.25 are on Table III. m m From this Table, it is easy to see that the minimum downwind distance from the reactor is 458 m. with an ESE wind in July and the maximum distance is 622 m. with a NE wind in January, If one take into account the yearly weigh ted average, these distances are 498 m. and 605 m. But, as it is easy to see from Table II, the dominant wind for the reactor site is from NE direction; so, the distance of 605.m.can be taken as the distance where the maximum dose rate will be collapsed. This.value is also in good agreement with the value obtained in Reference 1. To know the radioactivity deposition in this place, one has to know the Q value of the gaseous radioisotopes, and make calculations with the K values of the Table III, according to the equation (4). 4- Dry Deposition of Fission Particles When the fission products are not in gaseous form, but are small particles, the diffusion theory is not applicable in all case. For small particles of some ym of radius, the maximum concentration and its downwind distance X can also be calculated by using the diffusion deposi tion equations. For particles having 60 um of radius or more, the diffusion theory is not admitted as reliable, and the dry deposition velocity is then calculated according to the gravity forces. This deposition velocity will then permit to calculate the impact point of the particles . This distance X of a particle of radius r can be calculated by using the following formula: X = 0,812.10"6. u,h (6) S.r where, u = wind speed at the 2/5 of the stack height (cm/sec) h = stack height (cm) <5 = density of the fission product particle r M particle radius (cm) In this equation, the coefficient was calculated according to the viscosity of the atmosphere at 10 C. For atmospheric temperature changes, the value of X obtained by the help of equation (6) must be multiplied by 1,002 for each increase of 1 C of temperature or divided by the same factor for 1 C of each temperature decrease, (4). For different particle radius and for a stack height of 25 m., X values are calculated and plotted on Figure-1 for Cesium, and on Figure-2 for Strontium. The figure 1 shows clearly that, for any event occured in July, which has the highest NE wind blow probability, the cesium particles of 50 urn. of radius will be deposited around 170 m. away from the reactor for a stack height of 25 m. If the reactor accident event occurs creating a stack height of 100 m., this distance becomes 1100 m. For a tempest case, if the wind speed is 31,5 m/sec, the 50 Jim. Cesium par- 4-73 TABLE - III THE MAXIMUM CONCENTRATION CONSTANTS AND THE DOWNWIND DISTANCES FOR h= 25 m. and n= 0,25 X = (m) C = 10 6 K . Q m » N NNE NE ENE E ESE SE SSE S SSW SW WSW W WNW NW NNW AVR. Max. JAN. to 3,68 3,30 2,49 4 ,17 5,32 9,58 8,33 6,18 7,98 5,18 4,79 5,81 6,39 7,10 4,46 3,91 4,46 0,56 588 598 622 578^ 558 513 524 ' 546 527 560 ""567 551 544 536 572 583 572 761 m FEB. K 3,09 4,17 2,70 4 ,17 3,83 6,84 7,37 5,99 7,66 4,91 4,46 4,67 6,39 4,67 4,67 4,17 4,35 0,59 k 603 578 615 578 585 538 533 549 530 564 572 568 544 553 568 578 574 755 MAR. Km 3,76 3,76 2,66 3,83 5,18 9,58 10,6510,09 9,13 5,04 4,91 4,35 7,66 7,37 4,08 4,35 4,26 0,65 586 586 616 584 560 "513 505 509" 516 5"62~ 564 574 529 532 579 574 576 745 m - : k?R. 6C '5',Ï8'T/4'6" 2,99' 5,8'l"l'"i5"2 l7^2 "13>68'TÔ^ " U^7 5,3"2' 5,81 5,18 6,39 7,98 6,39 4,91 5,18 0,67 X 560 570 603 551 587 469 485 503 "499 555 551 560 541 527 541 564 560 738 m yiAY K 5,99 4,79 3,68 5,38 3,76 14,74 13,68 13,68 1198 6,39 6,84 5,38 7,37 10,65 7,37 5,99 5,99 0,72 X 549 561 588 555 581 478 "483 . 483 " 493 539 534 555 528 505 528 549 549 727 m JUNE K 4,56 5,38 3,14 5,99" 4 ,56 15,97 15,97 19J.6 12,77 7,9'tf 6,39 "5,81 8,33 7,98 6,84 6,18 4,91 0,69 X 563 ,555 594 542 563 471 471 T59" "4"86 "520" 537 551 517 520 531 539 557 7 29 m JULY K 4,91 4,91 2,82 3,55 "4,17 19,16 14,74 15,97 19J.6 10,65 10,09 5,32 8,71 8,33 7,10 4,35 3,99 0,60 X 564 564 602 582 569'"" 458 475 470 458 498 502 550 512 516 531 566 573 742 m AUG. K 4,35 4,08 2,70 3,76 4,91 7,66 19,16 17,42 13,69 " 10,65 •8,71' 7,66 5,47 8,71 3,76 5,64 4,26 0,72 X 566 ' 571 606 581 564 530 458 464 "480 498 512 522' 547 512 581 545 576 722 m SEP. K 4,56 4,08 2,66 3,19 3,55 10,65 1127 17^2 13,69 7,98 "7,66 6,39 7,37 7,66 4,08 5,47 4,35 0,76 X 563 571 608 593 582 503 495 465 481 520 523 544 526 529 571 547 574 719 m 3CT. K 4,08 3,62 2,66 3,19 4,26 "11^8 17,47 11,90 10,65 6,84 7,37 6,39 7,10 7,66 4,91 4,67 4,35 0,64 X . 571 587 611 595 576 492 4'6"7 .492 5Ö5 53X 528 544 536 525 559 553 569 740 m NOV. K 4 ,35 3 ,99' "2 ,82"T,Î7"6Ï8"4""W"10,65"(0,65 "6*,W^"V9" 5,81 4,91 6,84 7,66 4,67 4,46 4,91 0,65 X 569 573' ""602' ' 578"' "5'3"$"" T68""5Q'2' ""498/ " 1535"" '5'61"' '5"48 "'557 538 523 565 569 557 745 m DEC. {. 3,99 3,76 2,66 3,14 5,47 10,65 10,09 6,18 7,10 4,08 5",32 6,18. 6,39 7,10 4,08 3,68 4,91 0,64 X 580 586 611 600 547 504 508 545 534 571 556 545 542 534 571 586 563 744 m YEAR. te 4,17 4,17 2,86 3,68 "4,26 11,27 11^7 8,33 9,12 5,32 5.Ï7 5,32 6,.84 7,37 4,79 4,46 4,56 0,56 AVER. X 574" 575 605 584 572 "498 ~49"8 520 513 554 552 554 534 529 562 568 566 761 m Lunj 100- tOOkm figure! The downwind distances versus particles radius for Cesium rjum] Partide radius 10m 100m tOO km Figure.2 The downwind distances versus particles radius for Strontium tides will be deposited at 1360 m. from the reactor. 5- Conclusion Generally, the diffusion theory is adopted for the safety calculations of nuclear reactors and, in these calculations, 1-131 is almost taken as an exemple. But one other thing which is as important as the gaseous fission products, is.the particle release from a reactor accident. The most spec tacular exemple of this release is the cesium and strontium distribution from the Tchernobyl Accident. The deposition of these particles, among various others,' on the tea leaves, contributed to the radioactive pollution of the turkish tea. In the present day, around 50 thousands tons of tea leaves are waiting for a final deposition area. In this investigation, one can see the importance of this factor visualized by Figures 1 and 2. If. the particle radius of the released fission product is known, then the impact point of this isotope can be calculated from a similar figure with respect to the wind speed and direction. One thing to take care about is to know that up to 60 ym of particle radius, the diffusion theory can also be utilized for the maximum concentration cal culations, (4) . In all these calculations, the particles were supposed to have an exact spherical form. For more realistic investigation, a roughness factor must. be included into the-equations. It will also be interesting to notice that these investigations were done according to the pure fission product particle. But, if a small fission product particle is adhered on a bigger dust particle, these calculations and distances are not truth. In that case, one have to know the density of this carrying particle, e.g. soil or concrete particle, to be able to cal culate the deposition distance. The importance of particle size distribution was the object of some recent publication, (5). From these investigations, the Figure 3 was obtained, where it is easy to see the variation of X distance with the wind speed, with respect to the n factor, according to the diffusion theory. In this figure, the stack height was taken as 25 m. The Figure 4 shows the same variation for a stack height of 100 m. Acoording to the Figure 3, one can see that for the most probable wind di rection, NE, the impact point of the maximum concentration will be at 6600 m far from the reactor, in July, for n = 0,50. But for a stack height of 100 in. for the same ground velocity of 6,8 m/sec in July, this distance will be 53.000 m. As it is easy to see from this investigation, for n = 0,20 or n - 0,25, the downwind distance.of the maximum concentration is not very long, and is si tuated in the Campus of the University, but for n = 0,5 0 and for a high stack the populated area of the city of Istanbul will receive some amount of radio- 4-77 n=aso s n=a2S n=0.20 12 3 4 5 6 7 8 u,wind speed (m/sec) Figure.3 Maximum con can tra Hon distances versus wind speed for stack height 25m 4-78 ) \ 7 ; Xm Gnl n=0.5Q 1 no 9 « 7 s 5 A n=0.25 3. Z n-0 20 00 6 7 8 u,wind speed (m/sec) Figure.'» Maximum concentration distances versus wind speed for stack height 100m 4-79 activity with the most probable wind direction which is NE. The Tables and Figures showed.us also that with the SW wind, the danger of radioactive pollution of the populated area has no importance, because, with the same distances we are on the Black, Sea zone and only the environ mental pollution must be taken into account. REFERENCES 1- N. AYBERS, H-. YAVUZ. " About the safety analysis of Istanbul TRIGA Mark II reactor ". V. European TRIGA Users Conference, 4-6 September 1978, Portoroz. 2- General Atomic. " Safety analysis report for the TRIGA Mark II reactor for the Technical University of Istanbul ". E-117-478, October 1975. 3- 0. G. SUTTON. " Micrometeorology ". McGraw-Hill Book Company, 1953, 4- Smithsonian Meteorological Tables. Smithsonian Institution, 1951. 5- W. NIXON, D. J. ALPERT. " Effect of particle size distribution of released material on consequences of PWR accidents ". UKAEA, Safety and reliability directorate, SDR R 370, August 1987. 6- Israel Research Reactor I Hazards Evaluation Report. Israel Atomic Energy Commision, IA-689, December 1961. 4-80 EVALUATION OF THE EFFECTIVE DOSE EQUIVALENT TO TBE PUBLIC OF PAVIA AFTER THE CHERNOBYL NUCLEAR ACCIDENT S. AUieri , A. Berztro , N. Genova , S. Meloni , G. Roati 1) Laboratorio Energia Nuclear« Applicata, Université di Pavia, Pavia (Italy) 2) Dipartimento di Chimiea Generale, Università di Pavia, Pavia (Italy) INTRODUCTION In the last Triga Users Conference held in Rome in 1986 a report was presented on the concentration of radionuclides, deriving from the Chernobyl nuclear accident, as measured nearby the LENA plant, in air particulate, water, grass, milk etc. The Chernobyl radionuclide monitoring campaign in air particulate and foodstuffs was carried out and continued up to June 1987. On the basis of collected data estimates of the collective effective dose equivalent commitment to te public of province of Pavia, by external irradiation or by inhalation, were carried out and are reported in the present paper. RESULTS AND DISCUSSION The environmental radioactivity surveillance system at the LENA plant consists of four air pumping stations, continuosly collecting air particulate on filter paper (SS589 type, 50 mm diameter). Average air volume passing through the filter is about 200 m3 per day. Filters collected during the Chernobyl radionuclide monitoring campaign were_sub mitted to total beta activity counting and to direct gamma-ray spectrometry using a Ge(Li) detector connected to an analyzer-computer system. The area considered in the present investigation is the province of Pavia (2965 Km2), located in the south-west corner of Lombardia in Northern Italy. The population esti mât of the Province of Pavia was taken from the last census (1981): the total number of inhabitants was 185,375. They were divided in four population classes according to age: infant (age between 0 and 2 years), child (age between 3 and 10 years), teenager (age between 11 and 18 years), and adult (age greater then 18 years), as shown in Table 1. TABLE 1 Age distribi.ition of the population in the t province of Pavia Class Age(years) Population INFANT 0-2 4576 CHILD 3-10 16639 TEENAGER 11-18 20722 ADULT >18 143438 Radionuclides considered in the present dose evaluation are reported in Table 2: they are either the most abundant species (1317, for instance) or long lasting species in the environment (137Cs, for instance). 4-81 TABLE 2 Considered radionuclides for dose calculation Radionuclide Half-life l37C* 30.17y 134C* 2.06y l40La 40.3A ™Ru 39.4d 13ij 8.04 The following dose assessments for each, population class were worked out: - effective dose equivalent from external irradiation - effective dose equivalent from inhalation - effective dose equivalent to thyroid from inhalation - collective effective dose equivalent from external irradiation and inhalation. The effective dose equivalent for each class has been obtained by the following relationship (1): Deq = Hii:\il-Qi{t)-[F\i-ù,tj (i) where: Qi{t)= i-th radionuclide concentration, considered constant over 24 hours until May 31, 1986 and over a week in the time period June 1-December 31, 1986 [F]i= conversion factor for the i-th radionuclide for the considered doses At,-= 24 hours until May 31, 1986 ad 7 days from June 1-December 31, 1986. The concentration factors, namely the corresponding dose for unit concentration of radionuclides, are reported in Tables 3 to 5. The following respiration rate (m3 • s'1) were also considered: adult = 3.33 • 10"4, teenager = 3.33 • 10-4, child = 2.17 • 10-4 and infant = 6.94 • io-5. A-82 TABLE 3 Effective dose equivalent from external irradiation: conversion factors for unit radionuclide concentration in air (Sv • a-1 • Bq-1 • m3) Radionuclide factor i37C* .46 10-" l3*Cs .73 io-13 lwLa .11 io-12 iö3 £t* .22 10-13 ' '" 131 j .18 io-13 132J .11 io-12 ^Ct .89 10- » ™Cs .10 io-12 lü6i2u .97 io-14 9bZr .34 io-13 13 132Te .11 io- 9t >Nb .35 10-13 "Mo .75 io-14 133J .29 10-13 141Ce .36 io-14 TABLE 4 Effective dose equivalent from inhalation: conversion factors for unit radionuclide concentration and for different population class (Sv • s~l • Bq-1 • m3) Radionuclide Adult Teenager Child Infant 137Cs .75 • HT8 .13 • 10~7 .50-10-'' .74 • 10-7 134Ca 6.2 • 1011 1.51 1007 2.77 • 1007 1.9 10ü8 13 "*La 2.6 • 10 6.26 10Ü8 1.15 • 1009 8.1 10ü9 m 13 Ru 1.0 • 10 2.52 1008 4.62 • lO08 3.3 10Ü9 11 i3ij 4.2 • 10 1.02 10ü7 9.44 • 10üe 1.3 10Ü8 132j 1.9 • 1011 4.73 1006 4.35 • 10oe 6.1 1007 144c« 1.6 • 1012 3.92 ioü7 3.60 • 1007 2.6 lO08 13 08 08 Ö9 ™c3 2.2 • 10 5.43 IO 4.98 • lO 7.0 10 106 12 ü7 07 iïu 2.0 • 10 4.83 10 4.43 • 10 6.3 1QÜ8 96 12 Zr 3.6 • 10 8.71 10" 7.99 • 1007 1.1 1QÜ9 12 07 07 08 132Te 2.0 • 10 4.83 10 8.86 • 10 6.2 10 9& 12 08 Nb 4.8 • 10 1.17 10 2.15 • 1008 1.5 1009 11 "Mo 1.6 • 10 3.97 1006 7.29 • 10ü6 3.9 1007 i2 08 08 133j 9.7 • 10 2.36 10 4.32 • 10 3.0 10U9 141 Ce 1.4 • 1013 3.38 1008 6.19-1008 4.1 1009 4-83 TABLE 5 Effective dose equivalent from inhalation to the thyroid: conversion factors for unit radionuclide concentration and for different population class (Sv-3-l-Bq~lm3) Radionuclide Adult Teenager Child Infant ™Cs .75•KT8 .13 • 10-7 .50 • 10"7 .74-KT7 i34Ca 6.2 10" 1.51 • 1007 2.77 • 1007 1.9 10ü8 U0La 2.6 10" 6.26 • 10ü8 1.15-10ua 8.1 10°9 3 08 ™ Ru 1.0 1013 2.52 • 10us 4.62 • 10 3.3 10Ü9 131 j " 4.2 10" 1.02 • 10U7 9.44 • 1006 1.3 10ÜS T32j ' ' 1.9 10" 4.73 • 10oe 4.35 • 10oe 6.1 1007 144c« 1.6 10i2 3.92 • 1007 3.60 • 1007 2.6 1008 13 ua 08 Ü9 13ÖC3 2.2 10 5.43 • 10 4.98 • lO 7.0 10 iüöÄu 2.0 10" 4.83 • lO07 4.43 • lO07 6.3 10U8 96Zr 3.6 10i2 8.71 • 1007 7.99 • lO07 1.1 1009 i2 07 07 üa iwTe 2.0 10 4.83 • 10 8.86 • 10 6.2 10 Sbm 4.8 1012 1.17-10U8 2.15 • 10U8 1.5 10ü9 **Mo 1.6 10" 3.97 • 1006 7.29 • 10ue 3.9 1007 133 j 9.7 1012 2.36 • lO08 4.32 • 1008 3.0 1009 141 i3 us 08 C« 1.4 10 3.38 • 10 6.19 • 10 4.1 1009 Results are shown in Figures 1 and 2. Figure 1 shows the effective dose equivalent to the thyroid from inhalation to the different population class, micro sievert vs. number of day since April 29, 1986, for about 1 month. Figures on top of the curves indicate the cumulative dose values in 32 days. The child population class results to be the most exposed aliquot of general public. Figure 2 shows the effective dose equivalent from inhalation to the different population class. For the adult class, the effective dose equivalent from external irradiation is also given. Figures on top of curve have the already cited meaning. Also in this case the child population class results the most exposed one. The reported data allow the evaluation of the collective dose equivalent from external irradiation and inhalation: dose values are reported in Table 6 as sievert man for the different population classes and for the general public. TABLE 6 Collettive dose equivalent from external irradiation and inhalation in the time period April 29-December 31, 1986 Population classe Collettive dose man Sv mFANT 0.4 CHILD 3.2 TEENAGER 1.2 ADULT 4.5 GENERAL PUBLIC 9.3 The dose estimates reported in the present paper refer to dose contribution above natural radiation background. A global evaluation of the exposure increase in the Province of Pavia, due to Chernobyl nuclear accident in 1986, is not yet available because another important contribution is yet missing, i.e. the contribution due to ingestion. Calculations are currently carried out to derive, after a carreful evaluation of diet habits and of radionulcide persistence in soil, the important and longlasting contribution to collective effective dose equivalent commitment due to radionuclide ingestion. 4-84 CO S ADULTS to TEENAGER £ o o O 3 3 Ë10 ^O 488.2 E 242.3 -r 2 10 ~f 10 2 i 10 r 10 1 sr- 1 1 io- s-| I I I I I I in-1 f i i I I I I 8 16 24 32 8 16 24 32 days after 29/4/86 days after 29/4/86 I U CO CHILD o t. o 1233.1 I ,« 3 10 r 10 2 i 10 1 ; i"i i I I I I I 0 8 16 24 32 0 8 16 24 32 days after 29/4/86 days after 29/4/86 Fig. 1 Effective dose equivalent to the thyroid from inhalation 4-85 ,>1Û ADULTS en TEENAGER to 2 2 o E 2 £ 2 52.2 o 10 t 10 31.2 10 0.5 1 J -1 10" EXT. IRRADIATION f l 1 1 i 1 m" -2 ~" l I I I I I I 8 16 24 32 8 16 24 32 days after 29/4/86 days after 29/4/86 >10 er CO o CHILD w Ü 191.7 V2b 10 = .-1 10 -2 10 J I I I I L 0 8 16 24 32 0 8 16 24 32 days after 29/4/86 days after 29/4/86 Fig. 2 Effective dose equivalent from inhalation and external irradiation 4-86 REFERENCES [l] D.C. "Kocher Dose rate conversion factors for external exposure to photon and electron radiation from radionuclides occurring in routine releases from nuclear fuel cycle facilities". Health Physics 38 (1980) [2] F. Breurer, C. Brofferio, A. Sacripanti - "Equivalenti di dose impegnata da intro- duzione di attività unitaria, per quattro classi di età relative agli individui apparte- nenti alia popolazione per lo studio delFimpatto ambientale degli scarichi radioat- tivi" - ENEA Report RT/PROT (83)24(1983). 4-87 SESSION V Experiments with TRIGA Reactors THE STATIONARY NEUTRON RADIOGRAPHY SYSTEM: 1 A TRIGA-BASED PRODUCTION NEUTRON RADIOGRAPHY FACILITY j Robert H. Chesworth, Director TRIGA Reactor Division and Dean B. Hagmann, Project Manager SNRS Project Manager General Atomics - San Diego - California - USA ABSTRACT General Atomics (GA) is under contract to construct a Stationary Neutron Radiography System (SNRS) - on a turnkey basis - at McClellan Air Force Base in Sacramento, California. The SNRS is a custom designed neutron radiography system which will utilize a 1000 KW TRIGA reactor as the neutron source. The partially below-ground reactor will be equipped with four inclined beam tubes originating near the top of the reactor graphite reflector and installed tangential to the reactor core to provide a strong current of thermal neutrons with minimum gamma ray contamination. The inclined beam tubes will terminate in four large bays and will interface with rugged component positioning systems designed to handle intact air craft wings, other honeycomb aircraft structures, and pyrotechnics. The SNRS will be equipped with real-time, near real-time, and film radiographic imaging systems to provide a broad spectrum of capability for detection of entrained moisture or corrosion in large aircraft panels. GA is prime contractor to the Air Force for the SNRS and is specifically responsible for the TRIGA reactor system and a portion of the neutron beam system design. Science Applications International Corporation and the Lionakis-Beaumont Design Group are principal subcontractors to GA on the project. 5-1 INTRODUCTION Neutron radiography is a mature non-destructive inspection technique and has been used successfully for many years to detect the presence of hydrogen-containing materials inside of, or behind metal structures. Accordingly, the use of neutron radiography to inspect aircraft wings and control surfaces for the presence of moisture or corrosion in aluminum honeycomb assemblies is straightforward and has been demonstrated on a piece-parts, low throughput basis. The use of real-time imaging with x-radiography and neutron radiography does not have as long a history of demonstrated performance, but has emerged within the past several years as an alternative to film imaging where high throughput of objects to be radiographed is desired, and this type of imaging permits rapid scanning of objects and concentration on specific areas of interest. The real-time scanning of objects through a fixed neutron beam requires a programmable and precisely reproducible component positioning system. With the current state-of-the-art of robotic systems, this capability is currently available with modest development. The combination of a high intensity TRIGA reactor neutron source, real-time imaging and data storage and retrieval, and a ruggedized and reliable programmable component positioning system form a solid technical basis for establishing the Stationary Neutron Radiography System (SNRS). The primary mission of the SNRS is to non-destructively inspect F-lll aircraft wings to detect moisture or corrosion on a relatively high throughput production basis. The SNRS (figure 1) consists of a shielding and containment system, TRIGA Reactor System, Neutron Beam System, Component Positioning System, Neutron Imaging System and Image Interpretation System. Each of these systems will be briefly described in the sections which follow. SHIELDING AND CONTAINMENT SYSTEM The SCS is a rectangular enclosure (1370 M^) incorporating the TRIGA 5-2 reactor neutron source, four neutron radiography bays, staging areas, equipment areas, offices and control rooms (figure 2). Although the SCS is designed to ultimately accept and interrogate space shuttle solid rocket boosters, intact F-lll aircraft wings will be the largest items to be inspected initially within the facility. The SCS is designed with shielding and shutters to permit manned access to a radiography bay while the reactor is at full power and the other bays are in use. The staging area shielding walls are designed such that the increase in dose ratio due to the operation of the facility does not exceed 0.2 mR/hr, and radiation monitors are provided to measure the radiation level in the gaseous effluents from the facility (figure 3). A drain line and sump 'are provided to monitor the amount and activity of any water collecting as a result of reactor tank leakage (figure 4). A hatch is provided in the ceiling of the reactor room to allow long items to be inserted or removed from the reactor. TRIGA REACTOR SYSTEM The reactor system is a standard design 1000 kW, natural convection-cooled TRIGA reactor with the graphite reflector modified to accept the source ends of the four neutron radiography beam tubes which terminate in four separate neutron radiography bays. The reactor is located near the bottom of a water-filled aluminum tank 7 ft in diameter and about 26 ft deep. The tank is surrounded by concrete shielding on the sides and bottom. Access to the core is through the water from the open top of the tank (figure 5). The water provides adequate shielding at the top of the tank. The control rod drives are mounted at the top of the tank on a bridge structure spanning the diameter of the tank. The reactor is monitored and controlled by a state-of-the-art computer-based instrumentation and control system featuring color graphics display, self-calibration, and automatic logging of vital information (figure 6). Both manual and automatic control options are available to the operator. TRIGA fuel is characterized by inherent safety, high fission product retention, and the demonstrated ability to withstand water quenching with 5-3 no adverse reaction from temperatures to 2012° F. The inherent safety of this TRIGA reactor has been demonstrated by the extensive experience acquired from similar TRIGA systems throughout the world. This safety arises from the large prompt negative temperature coefficient that is characteristic of uranium-zirconium hydride fuel-moderator elements used in TRIGA systems. As the fuel temperature increases, this results in a mechanism whereby reactor power excursions are terminated quickly and safely. A plan view of the reactor is shown in figure 7. NEUTRON BEAM SYSTEM The NBS consists of a beam tube, aperture, shutter and beam stop for each of the four radiography bays. The NBS features an optimized "source end" design of the graphite reflector and beam-extraction "hole" to ensure maximum thermal flux, minimum gamma contamination, and minimum non-uniformity at the image plane. Beams with an L/D = 100:1 will provide thermal neutron flux of approximately 6.4 x 10" n/cm^.s at an operating power of 250 kW. All four beams are capable of simultaneous operation. The apertures, located at the edge of the reactor graphite reflector are replaceable in all beam tubes. Aperture change-out is accomplished by remotely removing a section of the reactor reflector and replacing the in-tank tube section with one with the desired aperture. The reflector section is then replaced. The in-tank beam tube section also provides a means of effecting future beam upgrades (e.g., liquid-cooled filters) by providing space along the beam tube for such upgrades. Two shutters are installed at the output end of each beam tube. A massive "biological" shutter has an "open" position and a "closed" position to allow safe personnel access to the exposure bay. A fast-operating, lightweight shutter, attached to the biological shutter, allows the thermal neutron beam to be attenuated for accurately controlling film exposures. Beam stops built into the walls are provided in each bay to minimize the thermal neutron albedo. An elevation view showing the above-identified 5-4 design features of a typical NBS for the SNRS is shown in fig. 8. Illustrated is the neutron radiography (NR) port and its biological shield, the massive personnel beam shutter, and the TRIGA reactor. COMPONENT POSITIONING SYSTEM The CPS is designed to position airplane components in each of the four inspection bays. The CPS in bays 1, 2, and 3 will be totally automated and will provide five independent axes of motion. Each of these bays is optimized to provide positioning of components by size. Bay 1 is designed to handle the largest components; the FB-111 wing being the largest (figure 9). The bay 1 CPS will accommodate all of the smaller components with proper fixturing. Bay 2 is optimized to handle other large components. The F-lll horizontal stabilizer is the largest. This bay can also accommodate all of the smaller components with proper fixturing. Bay 3 is optimized to handle small components up to 5 ft by 5 ft. Bay 4 is sized to accommodate NASA's solid rocket booster (SRB) components. The bay is equipped with fixturing to hold pyrotechnics and film cassettes that are moved in and out of the inspection area manually. NEUTRON IMAGING SYSTEM The NIS consists of components which transform the neutron distribution into a video image. It also contains the sensors to measure the incoming neutron flux and a system to mark defects observed by the operator. These items are included In a single subsystem since they are physically combined in the NIS module that extends from a telescoping arm (figure 10). The heart of the NIS is a 9-in.-diameter neutron-sensitive image amplifier. This device used an internal Gd£02S converter screen and deposited photocathodes to image and brighten the original neutron distribution to 5-5 form an image on an output phosphor. This amplified image is coupled to a high-performance Plumbicon camera through a pair of ultra-fast collimator lenses. This arrangement provides a very sensitive and wide dynamic range analog imaging system capable of producing excellent real-time radiography over two orders of magnitude of flux from ICH n/cm^-s to greater than 107 n/cm^-s. Built into the neutron imager are two subsystems that make its use more quantitative. These are the neutron current gauge and a defect marking system. The neutron flux gauge consits of two small U-235 fission chambers, one located at the beam exit port and the other located beside the image amplifier. This pair of counters are read out in the counter mode by a dual sealer-timer so that their rate can be directly interpreted in current units. The two flux measurements can be used to monitor correct system operation, determine average neutron attenuation through a part, and provide a means of calibrating the neutron absorption measured by the imager. Using a remote display system, the counter outputs will be displayed on the operator console where they are easily accessible to the operator. The defect marker system is a spray-jet ink marker. When a defect location is to be marked, the operator sprays a spot. He then moves the part to additional locations for marking. This can be done in real time while viewing the monitor, since the ink is quite visible in the neutron image. A single spot can be marked, or a series of spots to form a line around the damaged area. Neutron opaque marks along the movable arm assist the operator in describing the location of the flaw relative to the encodercoordinates of the image center. These marks can be used in the audio record as well as being entered onto the image if needed. The entire NIS package provides a complete set of input data to the image interpretation system which then aids the operator in interpretating the image appearing on the monitor. 5-6 IMAGE INTERPRETATION SYSTEM The purpose of the IIS is to collect, process, and display data from the NIS in a manner which best aids the operator to interpret realtime images for detecting the presence of corrosion or other flaws. Each system operation has been optimized with this goal in mind. The choice of hardware has been, in particular, specifically selected to provide the unique characteristics required for effective real-time detection and inter pretation of corrosion products on aircraft components. In addition, the IIS is capable of storing image sequences during the inspection and retrieving them later for comparison or review. The IIS functions by digitally storing and processing images to improve their contrast and effective resolution. The IIS performs all of the standard image processing functions of noise reduction, contrast stretch, and edge enhancement through its flexible pipeline processor and multiple frame buffers. In addition, the proposed IIS uses a unique field-flattening procedure to permit the sensing of very faint quantities of corrosion in real-time images. This is accomplished by special organization of the components. This organization permits the simultaneous correction for the imager response and neutron distribution to the image, as well as image integration and processing. This is a critical feature because the indications of corrosion within the image are often smaller than the instrumental effects of the system. The system is further enhanced by an interactive contrast stretch and pseudo-color display which permit the windowing of typical neutron images into color groups that represent particular areas of the part. On a typical image, the aluminum skins and honeycomb may appear blue, the corrosion pink and the sealant nearly white. This fine separation of image contrast in a smooth and controllable way not only eases the operator's task of interpretation but further provides a quantitative means of determining the level of the corrosion problem. Data is stored on a 1/2-in. industrial VHS video recorder. Additional capabilities are provided for storage onto a 1-gigabyte optical disk as a system upgrade. Hard-copy video images are also available within seconds to 5-7 provide a convenient means of referencing images for rework orders. Video cursors, as well as keyboard alphanumeric titles, can be overlayed on the image to augment the audio commentary on the video tape record. The images from the system are displayed on three monitors on the operator console (figure 11). The video processor can display two different images simultaneously. The third monitor is used to retrieve stored analog images from video tape to be compared with the processed data. A video switcher permits the switching of live and processed images between the monitors. All components and software used in the IIS are modular and can easily be expanded and upgraded as needed. In particular, the IIS is very well suited for advanced frequency domain processing and automated pattern recognition tasks. The proposed IIS represents the most advanced image processing equipment commercially available and will provide a long service life of high-performance operation in the SNRS project. SUMMARY GA is prime contractor for the Stationary Neutron Radiography System, which will be used to inspect aircraft wings and control surfaces for moisture or corrosion on a high throughput production basis. Construction at the McClellan Air Force Base site began in September 1987, and will be completed in late 1988. Figures 12 and 13 show progress at the site. It is expected that the TRIGA neutron source will start up in the first quarter of 1989, and the facility will be turned over to the Air Force by the end of 1989. 5-8 STATIONARY NEUTRON RADIOGRAPHY SYSTEM STATIONARY NEUTRON RADIOGRAPHY SYSTEM ENCLOSURE SNRS SECOND FLOOR PLAN VIEW SNRS ELEVATION SECTION B-B VIEW FROM TOP OF REACTOR TANK REACTOR CONTROL CONSOLE PLAN VIEW SNRS TRIGA REACTOR BEAM TUBE ASSEMBLY COMPONENT POSITIONER SECTION OF 1/48 SCALE MODEL TELESCOPING NIS ARM INSPECTION SYSTEM CONTROL CONSOLE SNRS CONSTRUCTION SITE MODERNIZATION DESIGN OF NEUTRON RADIOGRAPHY OF ITU TRIGA MARK-II REACTOR B. TUSRUL, A.N. BILGE Istanbul Technical University Institute for Nuclear Energy ABSTRACT ITU TRIGA MARK-II Research and Training Reactor has a power of 250 KW and has three beam tubes. One of them is tangential beam tube used for neutron radiography. In this study, the neutron radiography set in the tangential beam tube is described with its problems for ITU TRIGA Reactor. Afterthat modernization of the system is designed and the applicability of the direct and indirect methods is evaluated. Improving the ratio of length to diameter for the beam tube, elimination the fogging on the film and constructive design for practicle and secure application of the technique is developed. INTRODUCTION ITU TRIGA Mark-11 Research and Training Reactor which is placed in Istanbul Technical University-Institute For Nuclear Energy, has a power of 250 KW. Tte reactor has three beam tubes and thermal column which can be used for neut ron radiography like the other TRIGA reactors (1 ,2,3). Fig.l shows the, ge neral view of the ITU TRIGA Reactor. 5-21 -RAOIAL BEAMPORT RADIAL PIERCING - SEAM PORT -ALUMINUM CASING -SORAL PLATE \'-~~-i'. '•*'- !-'«.•;•<••>'V^-y^v GRAPHITC REFLECTOR S'5Ät> •THERMAL COLUMN OOORPLUG ;'S REACTOR TANK- Fig.l The General View of the ITU TRIGA Reactor One of the beam tubes is tangential used for neutron radiography. Neutron radiography is well known technique and effective for some special materials which have low atomic number (4). Thus, the neutrons can be attenuated by the light materials like water, hydrocarbon, boron and can be penetrate through heavy materials like as steel, lead, uranium. THE OLD NEUTRON RADIOGRAPHY SET IN ITU TRIGA REACTOR Thermal neutrons coming to the tangential beam tube, after passing through the water and graphite reflector due to position of the beam tube as it is seen from Fig. 1. Therefore, tangential beam tube is convenient for neutron radiography. For the qualified application of the neutron radiography should be arranged the angular spread of the neutron beam that reachs the object. It depends upon the source size and distance. Thus, the main parameters are length and 5-22 diameter of the beam tube. Collimator is used for suppliying of the aiming the angular spread. The length to diameter (L/D) ratio for a collimator effects both the resolution and the collimator efficiency. It can be said that, the characterisation of the collimator is determined by L/D ratio. Lenghth to diameter ratio of the tangential beam tube of the ITU TRIGA Re actor was originally 17. Maximum flux for the horizontal channels of the reactor was around 4.1 x lo"'' n/cm sn (5,6). The tangential beam tube had been arranged for the neutron radiography in three times . A conical collimator had been placed in the beam tube. It consists of four pair of ring which their diameters larging through the inner to outer directions. Fig.2 shows the old collimator separately. This part of the collimator had been placed near the reactor side in the tube. The bismuth filter had been used at the first ring of the collimator for eliminated the gamma rays. The diameter of the filter is 20 mm and length of the tube is 2582 mm. In this case, L/D ratio is about 130. First rings of the pairs in the conical collimator are made of cadmium and lead alloy. The weight ratio of the cadmium is 35 % and lead is 65 %. Second rings are made of lead and tin alloy with weight ratio 91 % and 9 % on recpectively (7,8,9). For the outer-side of the tube, a soller slit collimatorwas made for impro ving the geometrical unsharphess of the system. Itwas consist of aalvanized pipes in 1/2" diameter and stainless steel plate around the pipe bundles. Fig.3 shows the old neutron radiography set for ITU TRIGA Reactor. 5-23 2 3 Û ^3 m S22 $7* JCL -O- i I $2 Ä m -cx SE 290 275 275 18 22 I 6 Wire Pincers Stainless Steel 5 Rivet Aluminium 4 Tube Steel 3 Rino % 91 Ph % q sh 2 Filter Rings % 35 Cd % 65 Pb 1 Gamma Plu a Bismuth No Part Name MatPrifll OLD CONICAL COLLIMATOR Fig.2- Old Collimator Fig.3- Old Neutron.Radiography Set Outside of the beam tube, the concrete blocks have been placed as around a small area which has exposure car and related rails and object in it. We call this place as neutron room. The concrete blocks are in the same quality concrete with the reactor shielding wall. With this system neutron radiog raphy was realized in 30 minutes on 100 KW powered irradiation. PROBLEMS OF THE OLD NEUTRON RADIOGRAPHY SYSTEM The soll er slit collimator had not been functioning properly due to fabri cation and operational mistakes. For this reason, this part of the collima tor in the tube had not been used. Thus, only conical collimator in the tangential beam tubes had been used. As it is well known; direct or transfer methods can be used for the neutron radiography. It was aimed to apply both techniques with the neutron radiog raphy systems. But, it was noticed that there was fogged" on the radiograph in the direct method. It was aimed to eliminate it more efficiently. 5-25 Furthermore, when the transfer method was used with dysporsium or indium foils, a big spot has been appeared on the upper side of the radiograph, '' Fig.4 shows an example of it. It may be occured due to unhomogenous scatte ring in the collimator. It was aimed to eliminate it completely. Finally, some leakage have been detected outside of the neutron room. The range of it is below the maximum permissible level, but to eliminate them were also necessary. Fig. 4- Radiograph of the Neutron Beam REVISED DESIGN OF THE NEUTRON RADIOGRAPHY SYSTEM At first, the dimensions of the orifice of the bismuth filter will have much smaller diameter and longer length comparing the previous sizes. The diameter at the, bismuth filter is made 17 mm. In this case, L/D ratio is reached 150, and resolution is increased. Furthermore, when the lenght bismuth of the become 40 mm, the gamma rays will be absorbed in it mostly. Therefore, the gamma-fogged effect decreases more efficiently. 5-26 A new conical collimator designed for the elimination of the heterogenity of the neutron beam. The new conical collimator consists of two parts essenti ally. The first part of the collimator is seen in Fig.5. It has two beginning plate which are made of cadmium-lead and lead-tin alloys with the central bismuth filter. The conical collimator is supplied by the acid boric in the conical aluminium container« The last ring of the first part of the colli mator is made from aluminium. The second part of the collimator has again conical aluminium container containg acid-boric. Fig.6 shows the second part of the collimator. Fig.7 shows the two parts of the collimators in the tangential beam tube schema tically. The conical angle of the collimator is approximately 2°. By modi fying this collimation, neutron radiography results became much more pro mising. Therefore, the system modernised and stabilised. Finally, the shape of the neutron room is changed and the concrete blocks rearranged for elimination of the radiation leakage. Furthermore, the door of the neutron room is rearranged and redesign in the point of practical working during the usage of neutron radiography applications. CONCLUSION The neutron radiography system of the ITU TRIGA Reactor was constructed in 1982 (7). Afterthat it is rearraged two times but same troubles were still being available (8,9). The main problems of the system are the gamma fogging and unhomogenous neutron beam in the tube. With the new design of the system, we eliminate most of the problems effectively. ACKNOWLEDGEMENT We are grateful to the Turkish Atomic EnergyAjthority(TAEA) for supporting this project. 5-27 OO 8 Stub Rinq Aluminium 7 Neutron Absorber Acid-Boric 6 Container Aluminium 5 Wire Pincers Stainless Steel 4 Rinq 91% Pb. 97 Sb 3 Filter Ring 35 % Cd, 65 % Pb 2 Stub Lead 1 Gamma Filter Bismuth No Part Name Material FIRST PART OF THE CONICAL COLLIMATOR ig. 5- First Part of the Conical Collimator 1 I 3 Neutron Absorber Acid-boric 2 Container Aluminium ] Aluminium No Part Name Material SECOND PART OF THE CONICAL COLLIMATOR Fig. 6- Second Part of the Conical Collimator New Collimator System For Neutron Radiography REFERENCES 1. Bock, H., Buckberger,. T., Bucktela, K.,Grass, F., Hammer, S., Kasa, T., Schindler, H., "Summary of Current Research Projects at the Atominstitu- te of the Austrian Universities", Tenth Biennial U.S. TRIGA Users Confe rence, Papers and Abstracts, TOC-18, pp.1-30, 1.45. College Station Texas, April, 6-9, 1986. 2. Altieri, S., Pelizzar, M„L., Sturini, M.C., "Approach To Neuronography at the Lena Plant", Ninth European TRIGA Users Conference, Papers and Abstracts, TOC-19, pp.VI-34, VI-44, Casaccia, Italy, October 7-9 1986. 3. Estes, B.F., Philbin, J.S., Thome, F.V., "Neutron Radiography Facility Annular Core Pulse Reactor", TRIGA Owners' Conference IV, Papers and Abstracts, TOC-7 Salt Lake City, Utah, March 1-3, 1976. pp. 5-70, 5.91. 4. Van Der Hardt, P., Ro'ttger, H., "Neutron Radiography Handbook", D. Reidel Publishing Company, Dordrecht, Boston, 1981. 5. "Safety Analysis Report For the TRIGA Mark-II Reactor for the Institute for Nuclear Energy Technical University of Istanbul", Istanbul, August 1978. 6. Bilge A.N., Tugrul, B. "Nuclear Applications With ITU TRIGA Reactor", In ternational Symposium On The Utilization of Multi-Purpose Research Reac tors and Related InternationalCo-Operation, Grenoble-France, IAEA-SM-3000/ 024, 19-23 Oct. 1987. 7. Arseven, K., "Instruction of A Neutron Radiography System and Radiographic Exposure" Istanbul Technical University Institute for Nuclear Energy, M .Sc. Thesis, Sep. 1982. 8. Erel, N„, "A Comparative Study On Neutron Versus Photon Radiography" Istan bul Technical University, Institute For Nuclear Energy. M.Sc. Thesis, June 1984. 9. Tekin, M„, "The Related Work On Neutron Radiography and Determination Of Neutron Beam Characteristics", Istanbul Technical University, Institute For Nuclear Energy, M.Sc, Thesis, Jan. 1986. 5-31 THE POSSIBILITY OF GAMMA RAY STERILIZATION BY USING ITU TRIGA MARK II REACTOR A.N. BILGE, B. TUÖRUL, H. YAVUZ Istanbul Technical University Institute for Nuclear Energy ABSTRACT Gamma rays are one of the effective method for sterilization which is pre ferred for a long time. Generally Co-60 radioisotope sources betatrons or accelerators are used for the sterilization. In this work, it was aimed to find the possibilities of the sterilization by gamma rays obtained in ITU TRIGA Mark-II radial tube. Radiation dosages are measured in the radial tube and several medical products are irradiated. Irradiation is arranged according to the desired dosages. Irradiated steri lized goods (mainly medical products) tested and checked at the Governmental Medical Health Center and results compared with literature. It can be seen that this kind of irradiation is a good tool for sterilizati on. Unfortunately, because of the stability of the radial tube and unpracti- bility of the system it is rather difficult to compete with industrial system using Co-60 and accelerators. Nevertheless, this type of irradiation is also applicable for the purpose of the sterilization by using ITU TRIGA Mark II. INTRODUCTION Sterilization with the radioactivity is improved for medical products and on some foods (1,2). Gamma rays are one of the effective method for steri- 5-33 lization which is preferred for a long time. Generally, Co-60 radioisitope sources or accelerators are used for the sterilization process. It was aimed to evaluate possibilities of the sterilization by using gamma rays obtain in TRIGA Mark-II Research and Training Reactor. Therefore, the study is to search whether or not the gamma rays sterilization could be re alized in TRIGA Reactor, ITU TRIGA Mark-II Research and Training Reactor has a power of 250 KW. It has three beam tubes and thermal column for the related research activities. Fig.l shows general view of the ITU TRIGA Reactor. Fig.-1- Shows General View Of The ITU TPIGA Reactor Naturally, in the beam tubes and thermal column the radioactivity is heterege- nous and mixed. However, the ratio of them are different according to geomet rical positions. It is desired to find strong gamma rays with neutrons in the piercing and radial tubes. The neutron beam must be more strong in the piercing tube than 5-34 the radial tube since the piercing tube is going through the reactor core. The radial beam tube has graphite block before the core. In the thermal column, the gamma rays must be lower than the piercing and radial tubes due to have large block of the graphite, and therefore thermal neutron population is naturally high. In TRIGA Reactor, tangential beairtube has a collimator system for neutron radiography (3,4), The gamma rays has being eliminated in a large amount in this tube. Therefore,the tangential beam tube couldn't used for gamma rays sterilization, EXPERIMENTAL WORK At the begining of the study, the important thing is determine the radiation dosages in the tubes. The different methods were tried to find the real do sages and some problems due to the high level of radioactivity have to be solved. The photographic emulsions method were tried at the beginning (5). The, ra diation dosages of the beam tubes were out of the sensitivity limits of this method. Therefore this method could not be satisfactory. Afterthat, Thermoluminescent Dosimeters (TLD) were used (5). The measured dosages for thermal column and tangential beam tube were within the limits but the dosages levels of the radial and piercing tubes were still being high for TLD. Therefore TLD dosimeters were saturated. Finally, the Fricke dosimeters were used for the measurements of the gamma doses (6,7). As it is known the Fricke dosimeters are the type of a chemical dosimeter and they don't have the sensitivity under the neutron doses. The refore, the neutron avalibility was a problem, and the neutrons were effec ted on the solution. The elimination of the neutrons, a neutron plug was constructed for each beam tubes. It was made from paraphine and acid-boric in a PVC container. 5-35 There is a cadmium plate at the end of the paraphine blocks. These plug thermalizes the neutrons by the paraphine and absorbed them by the acid-bo ric and cadmium. The ratio of the acid boric is approximately 5 % in the neutron plugs Fig.2 shows the plug schematically. 1 Î. i 0 , - 1 . soa 1-.-. „ 3 Plate Cadmium 2 Pluq R? Arid-ftni-ir 1 Pipe Poly vinel chloruKPVC •^to-i;.! . _ NEU T P. 0 N PL U G Fig. 2- Neutron Plug Using this neutron plugs, the gamma doses could also measured in piercing and radial tubes by the fricke dosimeters. Table : 1 shows the gamma doses for different beam tubes in ITU TRIGA Reactor. Table 1 Gamma Doses For Different Beam Tubes in ITU TRIGA Reactor Place of The Measurement Gamma Doses (R/h) Radial Beam Tube 2200 Piercing Beam Tube 3463 Tangential Beam Tube 640 Thermal Column 3850 (including the neutron dose) For the sterilization, the dose rate is necessary between the 0,05 KGray to 50 KGray according to the materials( 2^. Under this circumstances, it can be evaluated that the radial and the pier cing beam tubes may be used for the aim of the sterilization. STERILIZATION PROCESS Neutron availability is also problem for the sterilization process due to destroy the structure of the matter. In this study, polyetylen medical pro ducts were selected as the sterilization goods. The chemical conjuctions of the organic matter can be broken under the radio activity, especially by neutron doses. The neutron plugs which is made of paraphine and mixed with acid-boric were used for the elimination of the neutron effects. The plugs were tested with the irradiation the golden foils with the cadmium covered and alone and de cided that the neutron plugs has effected against the neutrons. After the placing of the neutron olugs in the beam tubes, the medical goods (e.g. enjectors and their needles are irradiated), Fig.3 shows the sterili zation system for the radial and piercing beam tubes. Irradiation were realized for difeerent periods in the beam tubes. These are between approximately 6.5 to 16.75 hours. The irradiation levels of them are between 142 Gray to 580 Grays. S-T7 Neutron Pluq Sample Wooden Sample Holder Fig. 3- Sterilization System CONCLUSION After the irradiatiation process, irradiated goods were send to the Govern mental Medical Health Center for testing and checking their sterilization The Results of the Sterilization Experiments are summarized in Table: 2. Table 2 The Results Of Sterilization Experiment Irradiation Material Irradiation Mean Results Irradiation Dose (Rad) (Gray) Polyuretan 6 h. 27 min. 14203 142 Sterilized Radial Beam Tube Metal 6 h. 27 min. 14203 142 Sterilized Polyuretan 6 h. 27 min. 22336 223 Sterilized Piercing Beam Tube Metal 6 h. 27 min. 22336 223 Sterilized Polyuretan 12 h. 27 min. 27415 274 Sterilized Radial Beam Tube Metal 12 h. 27 min, 27415 274 Sterilized Polyuretan 12 h. 27 min. 431.14 43.1 Sterilized Piercing Beam Tube Metal 12 h. 27 min. 43114 431 Sterilized Polyuretan 16 h. 45 min. 36884 368 Sterilized Radial Beam Tube Metal 16 h. 45 min. 36884 368 Sterilized Polyuretan 16 h. 45 mi n. 58005 580 Sterilized Piercing Beam Tube Metal 16 h. 45 min. 58005 580 Sterilized As it is seen from Table : 2 these doses are useful for the sterilization and for that reason one can conclude that radial and pearcing beam tubes are useful for the radiation sterilization not on the commercial scale. 5-39 ACKNOWLEDGEMENT We are grateful to the Turkish Atomic Energy Authority for supporting this project and Cekmece Research and Training Center for supply the Fricke and TLD dosimeters for the studies. We are thankful to M .Sc. Eng. Mehmet A. YA- VUZ for helping us during the our studies in the ITU TRIGA Mark-II Research and Training Reactor. REFERENCES 1- Etherington, H., "Nuclear Engineering Handbook", McGraw-Hill Book Company. 2- Mutluer, B., "History For Sterilization On Foods", (in turkish) , The Course On Food Sterilization, Turkish Atomic Energy Authority, Lalahan, ANKARA 13-17 June 1988. 3- Bilge, A.N., Tugrul, B., "Nuclear Applications With ITU TRIGA Reactor", International Symposium On The Utilization Of Multi-Purpose Research Reactors And Related International Co-Operation, Grenoble-France, IAEA- SM-300/024, 19-23 Oct. 1987. 4- Tugrul, B., Bilge, A.N., "Modernization Design Of Neutron Radiography Of ITU TRIGA Mark-II Reactor", 10 th European TRIGA Users Conference, Wien-Austria, 14-16 Sep. 1988. 5- Tsoufanidis, N., "Measurement And Detection Of Radiation", McGraw-Hill Book Company, Hemispher Publishing Corporation, Washington, 1972. 6- Spinks, J.W.T., Woods, R.S., "An Introduction To Radiation Chemistry", 2 nd Edition, 1976. 7- Jayson, G.G., Parsons, B.J„, Swallow, A.S., "The Mechanism Of The Fricke Dosimeter", Int. Journal Radiât, Phys. Chem. Pergamon Press. Vol. 7. pp.363-370, 1975, 5-40 Liquid Waste Processing from TRIGA Spent Fuel Storage Pits. Karl Buehtela, Atominstitute of the Austrian Universities. Spent fuel storage pits at TRIGA reactors. At the Atominstitute of the Austrian Universities and also at other facilities running TRIGA reactors, storage pits for spent fuel elements are installed. Fuel elements can be inserted in storage cans and deposited in underground storage pits using remote handling equipment. These pits are filled with water. Due to cladding defects of stored spent fuel elements fission products may dissolve in the water of the storage pits. The main contribution of the radioactivity in the contaminated pit water will be due to the content of cesium-137. Chemical compounds of this alkaline element are readily soluble in water and cesium-137 is therefore extracted by the aqueous phase in the storage pits. During the last revision procedure, the reactor group of the Atominstitute decided to refill the storage pits and to get rid of any contaminated storage pit water. Therefore this liquid radioactive waste had been pumped to polyethylene vessels for intermediate storage before decontamination and release. The activity concentration of the storage pit water at the Aominstitute after a storage period of several years was about 40 kBq/1 (l|iCi/l), the total amount of liquid in the storatge pits was about 0.25 m3. At that opportunity it was tried to find a simple and inexpensive method to remove especially the radioactive cesium from the waste solution. This method should involve a minimum of handling and should also avoid a risk of contamination. Ideally it should be something like a selective ion exchange procedure for cesium-137. Decontamination procedures for cesium-137 containing liquid radioactive waste. Decontamination of liquid wastes containing cesium-137 can be done by distillation, precipitation or ion exchange. Distillation is not specific and by that method the best decontamination is accomplished. It was not considered being economical, to set up a distillation just for this waste water from the storage pits. For 5-41 precipitation usually precipitates of hexacyanoferrates, like prussian blue are applied which specifically adsorb cesium. Cesium, adsorbing complex compounds for the decontamation of liquid waste. The specific adsorption properties of some heteropolyc acid compounds for alkaline ions are well known. This adsorption is due to the lattice structure and lattice distance of the compounds regarding the diameter of the alkaline ions. First observations in this field had been made long ago during analytical investigations. In case of phosphate determinations with ammonium molybdate, the weight of the precipitate, an ammonium-molybdato-phosphate, had been always too high, if the solution to be analyzed contained considerable amounts of alkaline ions, even sodium. It had been found that the uptake of cesium into the lattice of the complex compounds does not only occur during formation of the precipitate by the addition of the proper chemicals to the solution. Cesium is also specifically adsorbed by an already preformed precipitate. These materials can even be used to separate alkaline ions by thin layer chromatography. Investigations in this field had been done years ago in our institute (1). Later on these adsorption capabilities had been used to separate cesium-137 from large amounts of water in bulky samples e.g. from rainwater, river water, or sea water. Also some investigations in this field had been carried out in our laboratory. A special high pressure filter equipment had been designed and built in our laboratory for the separation and determination of cesium-137 in environmental liquid sample materials (2). Another group of complex compounds with good adsorption properties regarding cesium are the complex compounds of hexacyanoferrate (II), among these ammonium-ferric (III)-cyano- ferrate(II) (GIESE - salt; AFCF). This substance belongs to the group of "Prussian Blue" compounds which feature dark blue colour and a good colloidal solubility. The properties of this substance have been investigated by GIESE already more than 20 years ago (3). Fig. 1 shows the lattice structure of Ammonium-ferric(III)-cyano- ferrate(II). The formula of this compound is NH4FeIIIFeII(CN)6- Fig.l: Lattice structure of Ammonium-ferric(III)-cyano-ferrate(II) The structure of this system had been clarified already 1936. The comers of the cubes are occupied by Fe^ and Fe^ ions alternatively. The ammonium ions are placed in the centres of the cubes. The ions in the centres of the cubes can be exchanged by other ions, especially by alkaline ions, cesium has the most suitable ion diameter to fit into that space. In table 1 the ion radii of the elements of interest are listed. Table 1: Ion radii of alkaline and ammonium ions. Ion radii in pm Li+ 68 Na+ 97 K+ 133 NH4+ 143 Rb+ 147 Cs+ 167 5-43 According to theory, ion exchange due to lattice space is achieved most easily, if ions of similar size are exchanged, therefore ammonium is a good partner for the exchange of cesium. Ammonium-ferric(III)-cyano-ferrate(II) (AFCF) has found increasing interest again after the Chernobyl accident as a food additive for livestock animals. When ammonium-ferric(III)-cyano-ferrate(II) is added to mixed feed, pelleted feed or contaminated milk or whey, the crystalline particles disperse in the gastro-intestinal tract, forming a colloidal solution which binds radiocesium very efficiently. What is useful and considered as excellent properties for the above mentioned applications, is of disadvantage for application in ion exchange columns. Above all the good colloidal solubility is an unwanted properly for the use of that material in an ion exchange column. Frequently slurries are produced which block the filter system within a short period of time. Better results had been obtained by depositing these hexcyanoferrates at the surface of materials to obtain thin but mechanically stable layers of the adsorbent. Further experiments in our laboratory dealt with investigations of other complex compounds in order to examine their capabilities to adsorb cesium. Many heteropolic acids of phosphorus, silicon, molybdenum, and tungsten had been examined, various hexacyanoferrate complexes had been investigated. Regarding difficulties of preparation, cost of chemicals, stability and adsorption properties as well as behavior in solution of mixtures of various ions, good results had been obtained with cobalt-containing hexacyanoferrates. Ion exchange column for the purification of storage pit water. To make efficient use of the ion exchange properties of the cobalt complex compound, the material should be deposited on a carrier material providing a large surface. One of our main ideas for producing the adsorption material was not to deposit the precipitate on a material with large surface but to use a material with large surface for the production of an adsorption layer. A layer of a new material should not be added to a carrier material with large surface area, but an adsorption layer should be produced out of the carrier material itself. To achieve that aim, iron material, like wire, or chips, which are 5-44 materials with relatively large surface areas are corroded at the surface using suitable corroding chemicals, like acids. By that procedure, iron ions, or if the material consists of steel, also ions of the alloy compounds (e.g. cobalt ions) are available at the surface of the material. Finally a solution of potassium hexacyanoferrate is applied as a reagent to produce ferrous-ferric-cyanide complexes as well as cyanide complexes of the other components of the steel alloy, like cobalt. It is sufficient that this layer is very thin, because adsorption of cesium is only achieved at the surface of the lattice structure of the complex compound. For the preparation of the adsorption material about 500 g steel wool was filled into a plastic tube of 50 cm length and 5 cm diameter. From the beginning the tube had been filled with water to avoid trapping of air bubbles. Then the system had been rinsed with diluted hydrochloric acid (5 molar), afterwards with a solution of potassium hexacyanoferrate ( 0.1 molar). Finally the column had been washed with distilled water. A glass fibre filter had been placed at the outlet of the column in order to prevent small particles or precipitates from being washed out of the column, by that contaminating the effluent during operation with cesium-137. Fig.2 shows schematic picture of the experimental setup. It is not necessary to use any pumps or other moving parts, there is little risk of contamination during the whole process. The procedure is extremely simple and no skill in radiochemistry is needed to handle the equipment. Finally the liquid waste had been passed through the column at a low flow rate. Samples from the effluent had been taken to determine the residual cesium-137 -concentration. Measurements had been carried out using a Nal - well type detector combined to a single channel analyzer. Before the measurement of the effluent samples, one had to wait until Ba-137 has decayed to a level not to interfere with the cesium-137 - determination. Measurement had been carried out after a decay time of 20 min (Half life of Ba-137m. 2.55 min). This radioactive daughter product of cesium-137 is not adsorbed by the complex compounds in the column and can be found in the effluent. 5-45 Waste solution Fig. 2: Experimental setup for storage pit water purification. •5335 ssjÄ Steel wool column H •••••*: Effluent Activity of the effluent in % of initial activity 60- % activity 50 i 40- 30- 20 - 10- fraction no (5 litres) -T— i 40 50 Fig. 3: Cesium-137 content of the effluent during storage pit water purification 5-46 Fig 3 shows the results of a decontamination experiments for liquid waste from the storage pits. Fractions of 5 1 ware collected and aliquots of these samples had been measured. After about 200 1 the breakthrough of the cesium-137 could be observed After this waste purification procedure finally a calcium phosphate precipitation had been carried out to be sure that also other radionuclides, especially strontium-90 are removed. After that treatment the waste water could be released to the Danube Channel. The radioactive material had been concentrated at a rather small volume at the steel wool. Final treatment and conditioning can be done by removing the aqueous phase and storing the steel wool (e.g. embedded in bitumen) or by slowly rinsing the column with a solution of ammonium hydroxide, which will elute the cesium. The steel wool can then be reused after treatment with potassium hexacyanoferrate solution to produce a new layer of adsorbing material. Isotope generator for Ba-137m. Another use of these investigations may be seen not only in the adsorption of cesium-137 from liquid waste but also by the desorption of Ba-137m. The diameter of the barium - ion is much smaller than the diameter of the cesium-ion and therefore barium is not kept by the adsorbent. Moreover, the adsorbed Cs-137 is continuously producing the radioactive daughter product Ba-137m, which can be rinsed out of the column with practically no Cs-137 - impurity. By that this system can be considered as an isotope generator for Ba-137m. An isotope generator for Ba-137m can be a valuable tool for training purposes or may even of some use in industrial applications if the Cs- 137 contamination in the effluent can be kept sufficiently low. For this type of isotope generator, both for training and industrial application, the design has to be changed compared to the waste treatment system. For the applications mentioned above it is essential that the Ba-137m has to be at rather high concentrations, that means that the volume needed for the elution of Ba- 137m must be as small as possible. This can be achieved by producing the layer of the adsorbent material not on a material with a large surface being in contact with a surrounding liquid phase, but by producing an adsorbent layer at the 5-47 inner part of a capillary tube. By that the following advantages will be obtained: - the volume of the liquid phase can be kept extremely low (mikrolitres) - the diffusion length of the Ba-137m is short. - The layer of the adsorbent can also be kept thin - the apparatus and the procedure can be designed very simple. A capillary tube of stainless steel (as it is for instance used for gas chromatography), with a diameter of ca 0.1 mm and about 30 cm length is attached to a plastic syringe of about 2 ml. At first the inner surface of the tube is treated with 10 m HCl, then with potassium hexacyanoferrate solution. After rinsing with distilled water, one drop of carrier free Cs-137 solution (ca 100 kBq) is slowly passed through this column. The cesium-137 is mainly fixed at a region at the top of the column until approximately 5 cm from top. Then the tube is rinsed with 10~4 molar HCl to get rid of the Cs-137 which is not properly fixed at the layer of the complex compound.and the generator is ready for use. Fig. 4 shows a schematic picture of this isotope generator. At the plastic syringe 10"4 molar HCl is stored. By passing one drop of this solution through the column, Ba-137m is eluted and can be used for experimental demonstrations and practical students work. For example: Measurement of half life, demonstration of resolving time of a Geiger Mueller counting tube, absorption effects with monoenergetic electrons, volume determination by isotope dilution, and many other experiments. Elution of Cs-137 related to Ba-137m activity is very small, mostly lower than 1 to 10^. The preparation of this column had been described in detail by GRASS and KLIMA, also some experiments are mentioned which are suitable for training courses in radiochemistry (4). During the cooperation with the IAEA in the field of training courses in developing countries, this equipment has been applied successfully. 5-48 plastic syringe capillary tube imprgnated with KCoFe(CN)6 planchet for Ba-137m sample Fig. 4: Radionuclide generator for Ba- 137m References: ( 1 ) BUCHTELA K. und LESIGANG M., Dünnschichtchromatographie an Salzen von Heteropolysäuren MIKROCHIMICA ACTA 67 (1965) ( 2 ) AZADKHANI M. Thesis TU Wien 1976. ( 3) GIESE W. und HANTSCH D., Vergleichende Untersuchungwen über die Cs-137 Eliminierung durch verschiedene Eisenxyanoferrat- Komplexe bei Ratten. Symposium Radioaktivität und Strahlenbiologie in ihrer Bedeutung für die Veterinärmedizin, Hannover 1968. ( 4 ) BUCHTELA K., GRASS.F and KLIMA H.,to be published. 5-49 NEUTRON PHYSICS EXPERIMENTS EXPLOITING THE PULSING CAPABILITY OF THE VIENNA TRIGA-Markll Gerald Badurek Abstract: The combination of a pulsed high magnetic field system with the pulsing capability of the TRIGA reactor leads to the realization of a neutron diffractometer which opens new fields in solid state physics. It allows to expose the sample to magnetic fields of up to 25 T and hence to study field-induced phase transitions in ferri- and anti- ferromagnetically ordering materials. Neutron depolarization is another technique where such a pulsed high magnetic field is of great merit. 1. Introduction Due to its intrinsically negative temperature coefficient of reactivity the TRIGA reactor can be pulsed to about thousand times its stationary power of 250 kW without the need of any additional safety precautions. During the pulse the available neutron flux scales up linearly with power reaching a peak thermal flux of about 10 cm" s" at the center of the reactor core. The minimum pulse width and the maximum pulse repetition time are about 40 ms and 10 min, respectively, if the reactor is pulsed to its peak thermal power of 300 kW. These two pulse parameters are of the same order of magnitude than the corresponding time intervals which are typically necessary to charge a high voltage capacitor bank and subsequently to discharge it abruptly across a solenoid in order to produce an extremely high magnetic field for a short period of time . Therefore we were inevitably guided to combine the pulsing capability of the TRIGA reactor with a pulsed high magnetic field facility to open new fields of applications of neutron scattering in solid state physics. For that reason we have developed a pulsed field neutron diffractometer which allows to measure the spin structure of ferri- and antiferromagnetically ordering compounds in the vicinity of the critical region where the antiferromagnetic coupling between neighbouring spins is broken up and where non-collinear spin structures are induced by a magnetic field of up to 25 T. Due to its flexible architecture the same pulsed field facility can optionally also be used in combination with a neutron depolarization setup. There it serves to initially saturize extremely hardmagnetic samples (which are of particular technological importance) whose magnetic domain structure can subsequently be studied during its approach to thermal equilibrium by transmission of polarized slow neutrons. 2. The Pulsed Magnetic Field Neutron Diffractometer Fig.l shows a schematic sketch of both the pulsed field neutron diffractometer which is installed at a radial beam tube of the TRIGA reactor and the above mentioned neutron depolarization facility at the tangential tube. A graphite monochromator selects a wavelength of about 1.1  out of the thermal reactor spectrum. The monochramtic beam impinges on the sample that can be mounted in a quartz glass cryostat and thus cooled down to liquid 5-51 BEAM SHUTTER Fig.i: Sketch of the experimental setup of the pulsed field neutron diffractometer and depolarization facility at two different beam tubes of the TRIGA reactor. As indicated the high field system can alternatively be used also as a stand-alone 40 T magneti zation measurement instru ment. PULSED 40 T COIL POWER SUPPLY & helium temperature. The outer diameter of the lowest part of this cryostat is small enough to fit into the bore ( \ 5-52 FIELD COIL STORAGE dM OSCILLOSCOPE \ S /dt £> JLL r~W— 1 -*\ \är^ TT dt REACTOR START ^T DELAY o ni OET.H^ ni o PREAMP. Sync. TO ELEVATOR I/O-REGISTER Control \^BPC-AT / i v !6kx24lOKX^tan 8IT (Âddn) AUTO-INCREM. « \~~K. MEMORY Incr. 32 OET. PULSE GENERATOR V Fig.2: Block diagram of the pulsed field neutron diffractometer electronics. registration electronics which is realized in CAMAC standard. The amplified and discriminated signals from the 32 neutron detectors (fill gas: He, pressure: 10 bar, outer diameter: 1 cm, angular resolution: 0.29 ) are encoded to yield the 5 position bits of a 14 bit data word which specifies an address of a 16k x 24 bit CAMAC auto-increment memory. Each time a neutron is detec ted the content of the corresponding memory cell is incremented by 1. The other 9 bits of the address carry the time information which is provided by a programmable pulse generator and a latchable counter and which is referenced to the moment the reactor and the field pulse start. By doing so, the neutron intensity can be recorded both as a function of scattering angle and as a function of time, i.e. field strength. Control of the complete neutron registration system is achieved via an IEEE/CAMAC converter by an AT-compatible personal computer which also controls an magnetization measurement instrument via an IEEE interface. Thus, in combination with an axial N/N pick-up coil system surrounding the sample the magnetization of the latter can be measured even during the diffraction experiment. The neutron detector bank is placed within a heavy (-700 kg) shielding filled with boronated water. This shielding can be slightly lifted by a Hoover-craft platform and hence a small step motor is sufficient to rotate the detector bank around the sample within an angular range of 120 . If the sample is not poly- but single-crystalline no multidetector arrangement is necessary. In that case it can be replaced by a Li-glass scintillation counter to avoid a priori any dead-time problems. To establish the necessary Bragg-reflection condition the sample rod is mounted on a small step motor-controlled goniometer. An important detail of this special double axis neutron spectrometer is shown in fig.3 by the cross-sectional view of the cryogenic sample 5-53 environment. Since the magnet coil has to be cooled by liquid nitrogen to minimize Joule's heating by reducing the ohmic resistance of the copper wires, special precautions are necessary to keep the sample environment sufficiently transparent for the neutron beam. This is achieved by lowering the liquid nitrogen reservoir below the niveau of the coil gap a few seconds before the reactor pulse is fired. Since the aluminium vessel sourrounding the whole arrangement is filled with gaseous nitrogen there is no disturbing humidity which could freeze out, thereby strongly decreasing the beam transparency. QUARTZ-GLASS y^CRYOSTAT . rfi V////////A X/////////A •SPLIT-COIL -SAMPLE LIQUID-N- ELEVATING PLATFORM Al v/mymrmrn Fig.3: Cross-sectional view of the cryogenic sample environment. As mentioned above, the main purpose of this new type of diffractometer will be to study field-induced magnetic phase transitions in antiferro- or ferrimagnetic materials [2]. The measurement scheme can be summarized as follows: Suppose that at given temperature a single-crystalline anti- ferromagnetically ordering sample and the scintillation detector are adjusted to measure the maximum intensity of a specific antiferromagnetic Bragg- reflection peak. As long as the field has not reached its critical value above which the antiparallel spin coupling is broken up nothing will happen to that peak intensity. But at the critical field strength a sharp discontinuous reduction of intensity should be detectable due to the occurence of a canted spin structure within the sample. Depending on the size and reflectivity of the sample crystal we expect typical intensities of 40 - 100 neutrons/ms if the reactor is pulsed to its full power of 300 MW. For polycrystalline specimens the measurement scheme is analoguous with the exception that several Bragg peaks are measured simultaneously by the multidetector. But there the intensity will be much lower and the data of several "runs" have to be accumulated to achieve sufficient statistical accuracy of the results. Although many features of such field-induced magnetic phase transition can also be investigated by careful high field magnetization measurements [3], there is no doubt that neutron diffraction will yield the required information both much more directly and at a much more detailed level. 5-54 3. Neuton Depolarization Setup Its origin dating back in the early 'fourties [4] neutron depolarization is a quite well established technique to study the domain structure of magnetically ordered materials [5-10]. Since it is essentially a transmission method it can be applied successfully even at relatively weak neutron sources. Without going deeper into detail fig.4 shows the main components of the three-dimensional neutron depolarization setup which was installed at the tangential beam tube of our reactor a couple of years ago and which since then was routinely used for a variety of investigations of different classes of magnetic materials [11-13]. By this facility it is possible to reveal the influence of temperature, applied magnetic field and/or mechanical stress on the mean size and orientation of the ensemble of domains within the sample. A brief description of its function principle can be given as follows: The incident beam, which is monochromatized (As 1.5 Ä) and polarized (P = |P| =: 0.95) by Bragg reflection at a magnetically saturated Heusler crystal, passes a spin flip coil which allows to invert the initial orientation of the neutron polarization vector P at any desired moment of time relative to a weak (-1 mT) magnetic guide field. A combination of an tapered axial solenoid and two specially designed crossed coils, wound around a cubic frame in a manner also indicated in fig.4, produces a magnetic field that changes its direction slowly enough along the neutron trajectory for the polarization vector to follow the changing field adiabatically. Immediately behind this adiabatic spin turn device the neutrons enter a magnetically shielded soft iron box where the field strength falls off so rapidly that the polarization maintains its last orientation.Thus, depending on whether and which of the two cubic coils is energized by a DC-current the polarization of the neutron beam reaching the sample may be oriented along any of the three orthogonal spatial directions. In the lower part of fig.4 the variation of the field components along the beam path is indicated schematically. The result |CH ADVANCE SPIN-FLIP PULSED DELAY CONTROL POWER-SUPPLY /TIMING-MCA CRYOSTAT Ht) MAGNETIC PERM. MAGN. w„»m SHIELDING GUIDE FIELD ANALYZER POLARIZER FLIP-COIL /DETECTOR Fig.4:Schematic sketch of the three- dimensional neutron depolari zation setup. The variation of the magnetic field along the beam path is indicated, too. 5-55 of an exact calculation of the field distribution for the actually chosen geometry of the coils according to Biot-Savart's law of electrodynamics and a subsequent numerical solution of the Pauli spin equation of motion is shown in fig.5 [14]. There a beam initially completely polarized in z-direction is assumed to propagate along the y-direction, whereby its nonadiabatic transition to the field-free soft iron box takes place at y = 0. It is clearly seen that, as intended, the final polarization P(y=0) is almost exactly oriented along the x-direction. ROTATION z—x 10 15 Y (cm) Fig.5: Typical result of a numerical calculation of the field distribution and the components of the polarization vector to optimize the geometry of the adiabatic spin turn coils (for details see text). A second mirror-symmetrically arranged identical coil combination behind the shielded box containing the sample (and its environment) allows to project any of the three polarization components onto the z-direction which is analyzed by a second Heusler crystal in front of the detector. By this one can consecutively measure all nine elements of the so-called "depolarization matrix" D which relates the incident (P) and final (P') polarization vector via the relation P' = D P. This (3 x 3) matrix D is a complicated function of essential domain structure parameters as mean domain size, mean square direction cosines of the domain magnetization directions, mean magnetization of the sample and several types of correlations between adjacent domains. Under the restriction of some simplifying assumptions, in particular with respect to these correlations, sophisticated inversion algorithms have been developed [5,15] to derive the above-mentioned domain structure parameters quantitatively from the elements of the measured depolarization matrix. The temperature of the sample can be varied in the range from 4.2 K to about 800 K by insertion of either a gas-flow He-cryostat or a quartz glass furnace into the soft iron box. Due to its special construction it is possible 5-56 to combine this furnace with an electrodynamic drive system which allows to apply a mechanical stress to the sample, either statically or periodically with a frequency of up to about 1 kHz.. In addition the sample can be exposed to a magnetic field by means of a built-in split coil connected to a pulsed power supply. Measuring the depolarization matrix elements as a function of time starting at the moment the field or mechanical stress pulses are applied one can study corresponding relaxation effects of the domain structure. The accessible time scale ranges there from about 100 fis to several hours. Until recently the maximum field was restricted to a value of 0.25 T, which is sufficient for the investigation of softmagnetic materials but is by far too low for almost all hardmagnetic samples. Thus it will be of tremendous advantage if we combine this neutron depolarization setup with our capacitor discharge driven pulsed field system, which now can be done via a high-current coaxial cable. Due to the ultra high magnetic field that is produced thereby, any hardmagnetic sample can be saturated to achieve a well defined initial state whose subsequent decay then can conveniently be studied. Quite generally, the application of a magnetic field to the sample causes some problems to be overcome in a neutron depolarization experiment. At first, only that component of the polarization vector which is parallel to the field direction can be guided through without disturbing variation. The two other components undergo Larmor rotations which in practice lead to complete depolarization because of the finite velocity spread of the neutrons. On the other hand, it is not trivial to guide even the parallel component through a symmetric split-coil. Since the field inside the coil and the outside return field are necessarily oriented in opposite directions the neutrons have to pass through a region of zero field somewhere along their trajectory. Except ASYMM. = 1-4 , 1 = 50 A HO £ 1CL 0-N 0- a* o ÛL* 0 Fig.6: Typical result of a field and polarization calculation for an asymmetric Helmholtz-type split-coil. The beam propagates parallel to the y-direction; y = 0 denotes the position of the coil axis which is oriented along the z-direction (asymmetry: 1.4, bore diam. of smaller coil: 60 mm). 5-57 for an infinitely thin beam such a field node causes a strong depolarization because it leads inevitably to a violation of the adiabaticity condition n « w = "}B, where fi is the field rotation frequency in the rest frame of the neutrons, w the Larmor frequency in a field of strength B and 7 = -1.83 x 10 s" T" the gyromagnetic ratio of the neutrons. The well-known solution of this problem is to use an asymmetric Helmholtz split-coil instead of a symmetric one, where the field direction is adiabatically inverted along the beam path without the occurrence of any node. A computer program was written which exactly calculates the field distribution of such a coil and solves the equation of motion of the polarization vector [16]. It was found that the optimal compromise between the required adiabaticity and the highest achievable field strength and homogeneity is an asymmetry factor of 1.4 between the radii of the two solenoids of a split-coil. Fig.6 shows the typical result of such a calculation for an asymmetric coil pair with a bore diameter of the smaller solenoid of 60 mm. The separation between the two coils is 27 mm. Again the beam propagates along the y-direction, the z-direction coincides with the coil axis. It is seen that the polarization vector, which initially is oriented in the +z-direction (i.e. parallel to the outside return field), ends up aligned nearly completely along the -z-direction upon reaching the coil axis and hence it remains parallel to field over the whole particle trajectory. 4. Concluding Comments The implementation of a pulsed high magnetic field facility into some of the neutron scattering instruments at our TRIGA reactor offers a chance to make this type of small research reactor competitive and even superior to much larger reactors which cannot be operated in a pulsed mode. The main reason lies in the fact that such extremely high magnetic field "shots" in practice can only be produced at a very low repetition rate which fits well to that of the TRIGA pulses. One disadvantage still to be overcome by increasing the capacitor battery (expensive !) is the relatively short duration of the field pulses (-15 ms in the crowbar pulse mode) compared to those of the reactor (-40 ms) so that only a part of the available neutron intensity is actually used. Acknowledgment: The author wishes to thank the Austrian "Fond zur Förderung der wissenschaftlichen Forschung" (project S42/08) for generous financial support and the many colleagues and students involved in this project for their cooperation. References [1] R. Grossinger, H. Hummer, IEEE Trans. Magn. MAG-14 (1978) 554. [2] Y. Shapira, S. Foner, Phys. Rev. B7 (1970) 3083. [3] E. Stryjewski, N. Giordano, Advanc. Phys. 26 (1977) 487. [4] O. Halpern, T. Holstein, Phys. Rev. 59 (1941) 960. [5] M.Th. Rekveldt, Z. Physik 259 (1973) 342. [6] S.V. Maleev, V.A. Ruban, Sov. Phys. JETP 35 (1972) 222. [7] M.Th. Rekveldt, F.J. van Schaik, J. Appl. Phys. 50 (1979) 2122. [8] S.V. Maleev, V.A. Ruban, Sov. Phys. Solid State 18 (1976) 1331. [9] M.Th. Rekveldt, F.J. van Schaik, W. Kraan, Nukleonika 24 (1979) &09. 5-58 [10] N. Stüsser, M.Th. Rekveldt, T. Spruijt, J. Magn. Mat. 53 (1985) 278. [11] G. Badurek, G. Janeschitz, H. Weinfurter, J. Hammer, H. Rauch, W. Steiner, J. de Physique 43 (1982) C-7/57. [12] A. Veider, G. Badurek, R. Grössinger, H. Kronmüller, J. Magn. Mat. 60 (1986) 182. [13] A. Veider, G. Badurek, H. Weinfurter, K. Stierstadt, Phil. Mag. (1988), in print. [14] G. Janeschitz, Thesis, Techn. Univ. Vienna (1983), unpublished. [15] H. Leeb, G. Badurek, A. Veider, H. Weinfurter, submitted to Z. Phys. B. [16] W. Kristufek, Diploma Thesis, Techn. Univ. Vienna (1982), unpublished. 5-59 TRIGA OUT OF CORE GAMMA IRRADIATION FACILITY J.Rant, G.Pregl "J.Stefan" Institute Ljubljana, Yugoslavia September 13, 1988 Abstract A possibility to irradiate extended objects in a gamma field inside the shielding water tank and above the core of operating TRIGA Mark II Reactor has been investigated. The irradiation cask is shielded with Cd cover to filter out thermal neutrons. The dose rate of the gamma field strongly depends on the distance of the irradiation position above the core. At 25 cm above the core, the gamma dose rate is 2.2 Gy/s and epithermal neutron flux is ~ 8.106ncm~23_1 as measured by TLD (C0F2 : Mn) dosimeters and Au foils respectively. Tentative aplications of the gamma irradiation facility are in the studies of raditation induced accelerated aging and within the Nuclear Power Plant Equipment Qualification Program (EQP). A complete characterization of the neutron spectrum and optimization of the 7 radiation field within the cask has still to be performed. 1 Introduction Irraditions in strong mixed gamma/beta radiations fields with peak dose rates of a few Gy/sec are of interest for the research of radiation effects. Radiation testing of materials, parts and equipment within the accelerated aging studies and studies of consequences of accidents one nuclear power plant equipment and environment is an important part of the Qualification Testing Evaluation (QTE) programs in several developed countries. Normally most radiation testing is done using fixed-energy gamma sources (typically 80Co). As suitable irradiation facilities based either on fixed energy gamma sources or accelators are not always available, especially if larger parts of the equipment are to be tested, we investigated the possibility to exploit the high intensity gamma field of the operating TRIGA Mark II Reactor for the out of core irradiations of extended objects. A prototype Cd covered irradiation cask has been constructed. In this communication preliminary experiments using this cask for trial irradiations of several electrical and electronic components were performed up to the accumulated gama dose in excess of 2 MGy. Experiments to characterize the mixed n/7 irradiation field in the water tank above the core of the operating 5-61 TBIGA reactor will be described. Based on the few trial irradiations plans for future development of the irradiation facility will be discussed. 2 Irradiation facility and the geometry of irradiations For the irradiations in the water tank of the reactor shielding a cylindrical irradiation cask has been constructed. Dimensions and other details of the construction are given in the Table 1. The cask is sealed with a Viton seal and a special plastic putty to prevent the ingress of water when immersed 5 m deep in the water during several days long irradiations. The cask is completely wrapped into a removable Cd cover (1 mm) to filter out the thermal neutrons. The cask is lowered on an Al stand to a fixed irradiation position above the core. Normally a position on the rim of the core, near a reflector is chosen to avoid the interference with other experiments and to facilitate the handling with the cask. To achieve the unformity of the absorbed dose in the object due to the gradient of radiation intensity in the Z-axis and due to the attenuation of the radiation in the object itself the cylindrical cask can be rotated around its longer axis. The parameters of the njf radiation field depend primarily on the Z distance of the cask from the core and were experimentally determined. Table 1: Construction data of the irradiation cask dimensions: 3 Characteristics of the gamma and neutron radiation field The following preliminary experiments were performed: • measurements of the variation of the gamma ray dose rate, epithermal and fast neutron flux with the vertical distance from the core in the water tank • measurement of gamma dose rate and neutron flux distributions within the irradiation cask The gamma dose rates were measured using calibrated TLD dosimeters, based on CaF2 : Mn sintered pellets (<£18mm, 1.6 mm thick). TLD dosimeters and the 5-62 adapted dosimetric reader were the same as developed and used at IJS for accidental dosimetry (3). The reader has been modified to measure the gamma ray doses in the range from 0.01 kGy and up to at least 2 kGy. The TLD dosimeters (pellets within Pb energy correction filter) were calibrated in the radiation field of the point *QCo source (~ 10135g). The supralinearity of the detector response was verified up to the absorbed dose of 2 kGy and is apparently limited only due to the overloading of the reader. The TLD dosimetry has been found as very versatile and practical method to perform almost point like gamma dosimetric measurements in the remote and inaccessible irradiation cask, where other instrumented methods (e.g. gas flow ionisation chambers) are impractical. However, for the measurements of the inte grated gamma doses in excess of 1 MGy other dosimetric methods should be used as for instance various glass dosimeters or some fibre optic materials (4,5) . The variation of the gamma dose rate with the distance from the core has been measured by simultanecously irradiating TLD dosimeters and activation detectors in Al cans (<£ 5 x 12cm), covered by Cd (1 mm) and positioned along the Z-axis at selected distances from the core. The gama dose rate and expithermal neutron flux (measured with Au-foil) are given in Table 2 and presented in Fig. 1. Table 2: Characteristics of the radiation field Position above Gamma Dose.Rate. ^ 25 2.21 0.03 7.8.106 50 0.67 0.01 9.4.104 75 0.17 100 0.03 150 < 0.004 Variations of the gamma dose rate at 20 cm distance < 15 % across the 18 cm distance in x-y plane Dose rates in the range 0.1 to 4 Gy/s. can thus be achieved. Epithermal and fast neutron component of the radiation field are due to the attenuation in the water layer greatly reduced, however they are negligible only for the irradiation positions higher than 50 cm above the core. Trial irradiations of various electrical and electrical materials and parts 25 cm above the core (D-, ~ 2 Gya~l) for a period of 288 efp hours of irradiation time (interrupted irradiations) and after 2 weeks of cooling time revealed that the activation of all irradiated parts including the stator segment of an electric motor (1.1 kg) was sufficiently low to allow the testing of the equipment in a laboratory, which is not equipped to handle the radioactive materials. However, after testing the eventually present radioactive scrapp materials and active objects had to be collected and disposed as low activity radioactive waste. Contact (7,/?) exposure rates measured on the surface of typical irradiated test objects after 2 5-63 weeks of cooling time are quoted in Table 3. Table 3: Contact exposure rates of typical objects, irradiated 25 cm above the core (irradiation time 288.8 efp hours) after 14 days of cooling time Equipment £>(/? +7) D{i) (C/kgh,mr/h) (C/kgh,mr/h) stator segment of electro- 7.7 x 10"7, 3.0 5.2 x 1CT7, 2.0 motor (1.1kg) electric cable - silicon insu- 7.7 x 10~7, 3.0 2.6 x 10-8, 0.1 lation (02.5mm, 3m) araldit (100 X 50 x 4mm) 9.0 x 10-7, 3.5 5.2 x 1Û-8, 0.2 guideway (feed through) 3.1 x 10~7, 1.2 1.0 x 10"7, 0.4 Instead of Cd filter or in addition to it other neutron filters / ß ray converters could be used to increase the proportion of penetrating beta radiation. It has to be mentioned that due to the activation of the Cd filter the opening of the cask should performed with a hand manipulator beyond a Pb wall. At irradiation distances lower than 30 cm the fast neutron flux is of the order of 108ncm~23-1 as measured by In foil. Using tabulated kerma factors (6) and rather conservative calculation it can be shown that the contribution of fast neutrons to the radiation damage of simple organic materials as lucite or nylon is smaller than 0.01 MGy for the irradiation time of 10ôs (neutron ßuence 1014ncm~2). For organic materials as well as for most metals the radiation damage due to fast neutrons is in our case negligible in comparison with the radiation damage due to gamma rays. However,if the irradiated materials contain strong neutron absorbing nuclides the contribution of the neutron field to the total dose can no more be negligible. To accurately asses the radiation damage a complete knowledge of the neutron spectrum is needed and it has to be measured. Future plans are to construct somewhat larger ( 5-64 ~t).00 30.00 80.00 90.00 Z ( cm ) Figure 1: Variation of the gamma dose rate O and epithermal neruton flux o (Au foil) with the distance from the core in the TRIGA water tank 5-65 4 Conclusions • We have demonstrated, that irradiations of extended objects in an intense gamma field of operating TEIGA reactor in excess of 1 Gy/s and without sig nificant interference of epithermal and fast neutrons are feasible. The integral dose in excess of 2 MGy as required by some Nuclear Power Plant Equipment Qualification Procedures can be achieved in less than 12 efp days of TRIGA operation. The actual testing can be performed after about 10 days of cooling time. • Gamma dose rates of a few Gy/s can be achieved by properily selecting the irradiation position above the reactor core tank. • The tentative applications are within Nuclear Power Plant Equipment Qalifl- cation Program are foreseen. References [1] L.L.Barizon, Overview of NRC • Sponsored Radiation Effects Research at SNL, Trans. Am.Nucl. Soc, Nov. 1985, 162 [2] J.Rant, H.Udovc, Characteristics of Gamma Radiation Fields to be used in Radiation Effects Studies of US Report IJS-DP-3928, June 1985 (In Slovene) [3] Thermoluminescent Dosimetric System MTDL-710B, IJS internal report [4] E.Jeltsch, W.Graf, Zur Gammadosimetrie in Kernreaktoren KFA-Jülich - 108 RB, 1973 [5] H.Boeck, J.Siehs, N.Vana, Investigation of fiberoptic behaviour during gamma irradiation, Report AIAU 80303,1980 [6] R.S.Caswell, J.J.Cozne, M.L.Randolph Tables of Kerma Factors EUR 5629, Basic Physical Data for Neutron Dosimetry, App. I., Jan. 1977, 299-300 5-66 ANALYSIS OF TRIGA APPLICATION FOR IRRADIATION OF POWER REACTOR PRESSURE VESSEL SPECIMENS* I. Mele, M. Ravnik University of Ljubljana 9J. Stefan" Institute, Yugoslavia Abstract The possibility of using 250 kW TRIGA reactor for irradiation of power reactor pressure vessel specimens was analysed. Irradiation times for specimens at different locations in TRIGA reactor were calculated and influence of steel specimens on reactivity was estimated. It was found ont that prescribed fiuence lQianetn~* of fast neutrons above 1 MeV can be readied in realistic times also in small TRIGA reactor but the damages caused in the material are not equivalent because the spectrum in TRIGA reactor is harder than in power reactor pressure vesseL 1 Introduction It is generally believed that small TRIGA reactors are not applicable for long term irradiation of large amount of material. We investigated the possibility of using small TRIGA reactor (250 kW or 500 kW) for irradiation of power reactor pressure vessel specimens. In this report the results of our analysis are given. Using zone averaged fluxes it was estimated how long the power reactor pressure vessel specimens should be irradiated at different locations in TRIGA reactor operating at different power levels to reach the fluence10 19n«n~a of fast neutrons above 1 MeV [l,2,3,4j. The influence of reactor pressure vessel specimen on the reactivity of TRIGA reactor was calculated and according to that the maximal amount of irradiated material was defined. The analysis showed that from aspect of reactivity and irradiation time the reactor pressure vessel specimens can be irradiated also in research reactor of low power (like our). But material damage depends also on neutron energy and not only on its total fluence. Because of that it is necessary to compare the spectrum of neutrons in TRIGA irradiation channels and original spectrum to which the material of reactor pressure vessel is exposed. It was found out that the differences in these two spectra are significant and result in different average displacement cross-sections (< 5-67 2 Calculational method Neutron flux distribution and the influence of steel specimen on reactivity worth were calculated with programme TRIGAP. The detailed description is given in [5], here only those characteristics will be given which are important for our further explanation. - It is based on one-dimensional two-group diffusion calculation which is good approximation for TRIGA reactor with cylindrical geometry. - The absolute accuracy is about 10 about 1.5 % in multiplication factor. It was estimated by comparing calculational and experimental results. With simple method of adjusting the term of axial leakage [5,6] the accuracy of absolute critical calculations is improved to about ± 50 pern. The accuracy of relative reactivity changes is much better than absolute calculations (few pcm). The same is throe for neutron flux distributions. Because of that it can be concluded that the results which are presented here can be thrusted qualitatively in any respect. Figure 1: WIMS model of cluster with reactor pressure vessel specimen in the center. The data base (the library of two-group effective cross-sections for ail materials in the reactor) for diffusion calculation was generated with transport code WIMS [7,8j in 18-group approximation. For fuel rods unit cell approximation with homogenization of cross-section over the whole unit-cell 5-68 was used JTor generating the cross-sections of steel specimen in irradiation channel it was used cluster option with specimen in the center surrounded with six standard fuel elements as neutron source. The cross-sections were homogenized only over the central unit cell Le. for irradiation channel with specimen in the center. In this approximation also the neutron spectrum was calculated. On Fig.l the model of cluster is shown. Fig.2 shows typical Charpy specimen of reactor pressure vessel material together with dimensions and mass of the specimen. It is seen that typical specimens which are used for Charpy tests are smaller than the reactor core height. That's why some additional simplifications had to be made in calculations. It was assumed that the channel in which the steel specimen is irradiated extends from the bottom to the top of the core. The influence of the specimen on reactivity worth was taken into account by homogenizing the specimen over the whole reactor height. The number of irradiated specimens in the channel was varied simply by changing the density of the specimen and keeping the dimensions constant. The maximal number of specimens in the channel was 7. The density which corresponds to 7 specimens is almost natural density of the steel. 1cm Figure 2: Typical Charpy specimen of reactor pressure vessel material. 3 Time of irradiation and specimen influence on reactivity The influence of the specimen on reactivity worth and fluxdistributio n was calculated for homogeneous core composed of 75 fresh standard elements and for mixed core composed of 47 fresh standard and 28 fresh flip elements. In mixed core C and D ringswer e completely filledwit h flip elements (except two position for control rods). In both cases the core configuration was realistic. The effect of burn-up was not taken into account because according to our previous experience [9] it has not much influence on power redistribution. In Table 1 the multiplication factors of homogeneous and mixed core are shown for the cases when 1,2,...,7 steel specimens were inserted in the central channel (CK) in A ring. The effect on reactivity of a single specimen in CK is about 25 pcm. This is small perturbation in a 5-69 Table 1: Multiplication factors of homogeneous and mixed core with 1,2,...,7 steel specimens inserted in central channel. number of tpecimen* homogeneous core mixed core 0 1.0137285 1.0151919 1 1.0106954 1.0121825 2 1.0104539 1.0119852 3 1.0102284 1.0117967 4 1.0100207 1.0116217 5 1.0098271 1.0114617 6 1.0096505 1.0113126 7 1.0094820 1.0111737 reactor and it can be assumed that the reactivity is decreasing linearly with the number of specimens. We can explain that with small seHshielding of the specimen. It can be seen also from Fig.3 that thermal disadvantage factor in a specimen is very smalL In next step we analysed the possibility of irradiating the steel specimen at fuel elements locations. The geometry of irradiation channel was equivalent to the geometry of central channel. The results of our calculation are given in Table 2 and Table 3 and on Fig. 4 and Fig.5. Table 2: Multiplication factor of mixed core with 1,2,-J steel specimens inserted at different locations in the core (in B,C,D,E and F ring). multiplication factor of banc core configuration = 1.0151919 number of location in the core »peeitnen$ B G D S / 1 1.0070596 1.0086401 1.0102384 1.0111154 1.0123949 2 1.0067832 1.0084996 1.0101495 1.0110118 1.0123438 3 1.0065209 1.0083703 1.0100633 1.0109110 1.0122962 4 1.0062728 1.0082421 1.0099777 1.0108159 1.0122519 5 1.0060378 1.0081223 1.0099000 1.0107251 1.0122105 6 1.0058131 1.0080092 L0098299 1.0106395 1.0121713 7 1.0056040 1.0079045 1.0097594 1.0105587 1.0121335 Comparing with the results in Table 1 we see that the reactivity decrease is larger than in central channel. The main reason is that for insertion of irradiation channel into the core (in inner rings) one fuel element has to be withdrawn and that irradiation channel by itself replaces certain amount of water in the core. The effect of steel specimen itself is of the same order as in previous case (in A 5-70 O CO 1 1 —T • r i i (0 LT3 E =3« - - I I CM \ O CM I wat e in - « - o ^ - - o - - voi d a - - o • in stee l o o , 8T'0 SI'0 2T0 SO'O 90*0 SO'O oo-c° 3_oi* xnu Figure 3: Fast (0.821 - 0.08734 MeV) and thermal (0.1 - 0.05 eV) neutron flux in unit œil with reactor pressure vessel material in the center. 5-71 Table 3: Multiplication factor of homogewneous core with 1,2,..,7 steel specimens inserted at different locations in the core (in B,Cfi$ and F ring). mtitipiication factor ofbatic core configuration = 1.0181985 number of location in the eon specimen* B G D E F 1 1.0042434 1.0050844 1.0070366 1.0090692 1.0106182 2 1.0038749 1.0047477 1.0068566 1.0089319 1.0105637 3 1.0035276 1.0044346 1.0066272 1.0087980 1.0105119 4 1.0031962 1.0041294 1.0064067 1.0086734 1.0104636 5 1.0028830 1.0038453 1.0061976 1.0085551 1.0104164 6 1.0025833 1.0035737 1.0060039 1.0084416 1.0103724 7 1.0023023 1.0033172 1.0058148 1.0083368 1.0103304 A o(pcm) 0 o © O o • o o o • • O water O • O void o • steel - Figure 4: The difference in multiplication factor between reference mixed core configuration and core configuration with water /void /7 steel specimens inserted in different rings. 5-72 0 o O 0 S •3 • o £' 0 water • O 0 void O • steel • -120C • 1 i 1 i 1 B c D E F Figure 5: The difference in multiplication factor between reference homogeneous core configuration and core configuration with water/void/7 steel specimens inserted in different rings. ring) for both cases. According to our expectations the reactivity decrease is smaller if we move the specimen from inner rings to the periphery of the core. From the aspect of reactivity there is no difficulty for irradiation of reactor pressure vessel specimen. Bat from practical reasons the most probable place for irradiation is in central channel or in outer F ring. Irradiation of 7 steel specimens in central channel reduces reactivity for about 150 pcm and in F ring for only about 35 pcm because we could use spare positions which are usually available in this ring. The irradiation time is by definition the time in which the prescribed fluence(lO^ncro"* 2) of fast neutrons above 1 MeV is achieved. Using program WIMS we calculated the neutron spectrum in steel specimen and found that about 15 % of total flux present fast neutrons above 1 MeV. With this result the irradiation times for steel specimen irradiated at different locations (in A,B»C,D,E and F ring) in a homogeneous and mixed core operating at two power levels were calculated. The results are given in Table 4 and Table 5. The irradiation time was calculated also for irradiation in reflector where the effect on reactivity is negligible. Only slight differences were noticed between irradiation times in homogeneous and mixed core so that the following conclusions are valid for both cases. The irradiation of reactor pressure vessel material specimen is possible also in 250 kW TRIGA reactor, but the irradiation times are shorter if power level is higher. Irradiation times are increasing when the specimen is moved from the central position to the periphery and are about 2.5 times longer in F ring than in central channel In reflector an order of magnitude longer irradiation times were calculated than in a core and are very close to realistic irradiation times of reactor pressure vessel material in power reactor. In reality the irradiation of steel specimens in 250 kW TRIGA reactor is possible only in central channel or in irradiation channels in F-ring. About 60 EFPD (effective full power days) in central channel and almost 150 EFPD in F-ring would be necessary to achieve the prescribed fluence. 5-73 Table 4: The irradiation times for 7 steel specimens inserted at different locations (rings) in 250 kW and 500 kW mixed TRIGA core. Total fluxes(ncro -2*-1) and total fluxes above 1 MeV (ncm-3«-1) are also given. R stands for reflector. MIXED CORE P = 250 hW ring $total ${otoi x 0.148 time(iays) A 1.33778 x 10" 0.19800 x 10" 58.5 B 1.22920 X101S 0.18193 X 10" 63.6 C 1.03963 x 10" 0.15388 x 10" 75.2 D 0.88553 x 10" 0.13107 x 10" 88.3 E 0.76057 x 10" 0.11257 X 10" 102.8 P 0.54350 x 10" 0.08044 x 10" 143.9 R 0.10341 X 10" 0.01531 X 10" 756.2 MIXED GORE P = 500 kW ring $tolal $ A 2.70003 x 10" 0.39963 x 10" 29.0 B 2.48543 x 10" 0.36788 X 10" 31.5 C 2.10283 x 10" 0.31121 x 10" 37.2 D 1.79311 x 10" 0.26540 x 10" 43.6 E 1.54205 x 10" 0.22824 X 10" 50.7 F 1.10144 x 10" 0.16302 x 10" 71.0 R 0.20938 x 10" 0.03099 X 10" 373.5 5-74 Table 5: The irradiation times for 7 steel specimens inserted at different locations (rings) in 250 kW and 500 kW homogeneous TRIGA core. Total flaxes («cm-*«-1) and total fluxes above 1 MeV (ncm-3*-1) are also given. R stands for reflector. HOMOGENEOUS CORE P = 250 kW ring * A 1.41720 x 10" 0.20976 x 1013 55.2 B 1.34403 x 10" 0.19893 x 1013 58.2 C 1.25399 x 10" 0.18560 x 1013 62.4 D 1.06735 x 10" 0.15798 x 1018 73.3 E 0.82388 x 10" 0.12194 x 10" 94.9 F 0.56997 x 10" 0.08436 x 1013 137.2 R 0.10841 x 10" 0.01605 x 1013 721.3 HOMOGENEOUS CORE P = 500 kW ring Qtatal itotal X 0.148 time(iayg) A 2.88094 x 10" 0.42641 x 10" 27.1 B 2.74073 x 1013 0.40566 x 10" 28.5 C 2.56083 x 10" 0.37903 x 10" 30.5 D 2.18519 x 1013 0.32343 x 10" 35.8 E 1.68986 X 1013 0.25012 x 10" 46.3 F 1.16884 x 10" 0.17300 X 10" 66.9 R 0.22219 x 10" 0.03289 x 10" 352.0 5-75 .25 lax ©•fei- 9 .1 «»*** Äi a. gf @ t3fl @—gl—il—i|i ~ 1 - m«• - -*=i—13 t-"f>3- -M- 'Ht- ~i—r T" T—-, ! , ! , ! j , , j ! j r r—1 ! 1 r-^|!feœ - r_2p -JJ3 Q_ In (E/10 MeV) Figure 6: 27-group spectrum in reactor pressure vessel material in power reactor (calculated with DOT) and 18-group spectrum in steel specimen in TBIGA reactor (calculated with WIMS). 4 Spectrum It was found out that from aspect of reactivity and irradiation times TRIGA can be used for ir radiation of reactor pressure vessel specimen. For final conclusions it was necessary to analyse the spectrum in steel specimen in TSIGA reactor and compare it to the spectrum to which the material of reactor pressure vessel is exposed in reality. This spectrum was calculated in 27-group transport ap proximation with programme DOT [10]. The geometry of reactor and reactor barrel was realistically modelled. Both calculated spectra are shown on Fig.6. It is seen that significant differences in spectrum shape appear in the part important for the embrittlement (above few 100 keV). Main reason for neutron radiation damage in metals is the dis placement of atoms from their normal lattice sites. Number of displacements per atom (DPA) during an irradiation depends on neutron energies (neutron spectrum) and according to ASTM (American Society for Testing and Materials) can be used as a measure of difference between neutron spectra. DPA is defined as follows: DPA = / dt H The displacement cross-section (so called "damage function") for steel is given on Fig.7 [11]. Another measure of spectrum hardness, recommended by ASTM, is the average displacement cross-section < di >, where it is assumed that displacements are caused predominantly by neutrons 5-76 of energies greater than E0.< f£ Recommended value for Eo is 0.01 MeV. To show quantitatively the difference between the spectrum in steel specimen in TRIGA reac tor and the spectrum in reactor pressure vessel material in power reactor the average displacement cross-section and product of energy dependent neutron flux (spectrum) and damage function (the displacement rate) was calculated for both cases. The results are given below. See also Fig.8 and Fig.9. TRIGA REACTOR: < POWER REACTOR: < Significantly larger average displacement cross-section in TRIGA reactor confirms our previous conclusions that the spectrum in steel specimen in TRIGA is harder than the spectrum in power reactor. From Fig.3 and 9 we can notice that the contribution of thermal and epithermal neutrons to total DPA rate is small The main part of DPA rate is due to fast neutrons above 1 MeV. From our previous experience we know that fast flux can not be efficiently increased by incore fuel management so optimization of steel specimen irradiation in TRIGA with core rearragement is not possible. Because of harder spectrum in TRIGA the contribution of displacements induced by neutrons with energies above 1 MeV in total DPA is higher than in pressure vessel material in power reactor. The result is given below. TRIGA REACTOR: DPA < above 1 MeV > = 78.0% DPA < total > POWER REACTOR: DPA < above 1 MeV > = 53.9% DPA < total > 5-77 S N iz c 9 CT — y> o i_2: -A. •fH^ Figure 7: Cross-section for reactions inducing damages in material damage function). 5-78 DPAxlO 21 .25 (s"1) •w-4. spectrum DPA "»©*- -A—pi—I=M«I— -w- o. ; -*p "Ef |CJ' |."J' |i.l"^ r j ,.„—r__T ! r - 20 - 10 In (E/10 MeV) Figure 8: Displacement rate and spectrum for steel specimen irradiation in TRIGA reactor. 22 DPAx 10 .25 (s"1) DPA- Wi Ou .1 % spectrum (-J—._(:rf ti *. _S|—j|i—1|)__ g—g^ -IT _@—ö~ 0. •It . • • i.w • »i»i » i-, > ,T^*^T-r-r-r-r^ 10 In (E/10 MeV) Figure 9: Displacement rate and spectrum for reactor pressure vessel specimen in power reactor. 5-79 5 Conclusions Our analysis showed that the average displacement cross-section for steel specimen irradiated in TRIGA reactor is significantly larger than for reactor pressure vessel material in power reactor which confirms that the nentron spectrum in steel specimen in TRIGA reactor is harder than in power reactor. The contribution of DPA reactions above 1 MeV in steel specimen in TRIGA is about 25 % higher than in reactor pressure vessel of our power reactor. From this ascpect the irradiation of steel specimen in TRIGA reactor is not fully equivalent to the irradiation of pressure vessel steel in natural conditions. As a great practical problem we see the establishing and assurance of other physical parameters which are important for neutron damages of pressure vessel, such as temperature. It would be very difficult to maintain the operating temperature of PWR pressure vessel (about 290 °(7) in the irra diation channel of TRIGA for long term irradiation. The irradiation could be technically realized in the irradiation channel outside the core. If the irradiation can be realized at normal temperature also irradiation facilities in the core can be used. One of possible application of TRIGA research reactor is irradiation of steel specimen for annealing studies and even small (250 kW) TRIGA reactor could be used for these purposes. References [1] Standard Practise for Characterizing Neutron Exposures in Ferritic Steels in Terms of Displacen- ments per Atom (DPA), ASTM (2] NPP Krsko Technical Specifications [3] Final Safety Analysis Report for NPP Krsko, Section 4, Westinghoose, 1979 [4} U.S. NRC, Regulatory Guide, Office of Standards development, RG 1.99, Effects on Materials, Revision 1, April 1977 [5] LMele, MJtavuik, TRIGAP - A Computer Pragramme for Research Reactor Calculations [6j MJEtavnik, Experimental Verification and Adjustment of Calculation^ Procedures for TRIGA REactor, IAEA • Interregional Training Course on the Application of Small Computers to Re search Reactor Operation, Ljubljana, 1-26 June 1987 [7j KJKowalska, The WIMS-S Code for the CYBER-72 Computer, Nea cpl, France [8j WIMSD4 - Version 100 and Cataloged Procedure, NEA CPL, France [9j M.Ravnik,LMele, Burn-up Calculations of TRIGA Reactor Core, Progress Report on Work under IAEA Research Contract No. 3512/R1/RB at "J. Stefan: Institute, Ljubljana during the Time Period from 1984-04-01 to 1984-06-01 [10] WXRhoades, DOT TV Version 4.2 Two Dimensional Discrete Ordinates Radiation Transport Code System, RSIC Code Package CCC-320, ORNL, 1979 [11] D.E.Cullen^JCocherov.P.M.McLaughlin, The International Reactor Dosimetry File (B2DF - 82) 5-80 SESSION VI Radiochemistry, Radioisotope Production & NAA / ) / / ! \ THE EVALUATIOlT'OF THE RADIOISOTOPE PRODUCTION FOR INDUSTRIAL USES A.N. BILGE, H. YAVUZ, B. TUGRUL Istanbul Technical University Institute for Nuclear Energy ABSTRACT Radioisotope production is one of the most important functions of the reac tors. Importance of the uses of radioisotopes in medicine, industry and ag riculture is improving. Under this view, it is aimed to produce radioisoto pes can be used for industrial purposes by using ITU TRIGA Mark II reactor. The in core irradiation facilities that can be used for radioisotope produc tions are central thimble and rabbit systems. Continuous irradiation at 250 KW with a flux of 8,1 x 1012 and pulse irradiations 12C0 m and ~ 1015 flux used to irradiate the sodium. Irradiated sodium is chemically treated in order to be used as a trace element for the hydraulic applications. The purpose of this work is develop a simple industrial radioisotope produc tion cell and handle the radioisotopes for the applications. INTRODUCTION Radioisotope production is one of the most important functions of the reac tors. Importance of the uses of radioisotope in medicine, industry and agri culture is improving. The using of the radioisotope in the several fields with the different techniques also continue to expand in Turkey. 6-1 As known that, the system of the Lazy Suzan is an important system for the TRIGA Reactors for routine radioisotope production (1,2). But Istanbul Tech nical University (ITU) TRIGA Mark-II Training and Research Reactor has not this system. Therefore, the radioisotope production techniques have so me problems for the rautine production. However, several radioisotopes can easily be produced in ITU TRIGA Reactor for industrial applications. ITU Institute For Nuclear Energy is playing an important role with their academic training and improving activities in the industrial applications. One of the important application for industrial purposes with the radioiso tope is the radiotracing technique. RADIOTRACING TECHNIQUE The radiotracing technique is a metod for tracing the fluids using with the radioisotopes for different purposes. Using this technique, the knowledge can be taken about the system or some part of the system (3,4). Radiotracing technique essentially can be applied in the channels, canals or for the pools. Radiotracing technique is also applied for evaluation of the environmental pollution. The pollution and distribution of the wastes in the fluids are getting more importance in Turkey. The studies with the radiotracers have been started in our Institute for the related models. For the first stage of the this subject, the concentration was on the opened channel model. EXPERIMENTAL SYSTEM An opened tank was conctructed on 2 m length, 1 m width and 0,5 m height as a pool model. A 10 cm heights barrier is placed at 10 cm of the end of the tank. Three source valves are placed in the top of the beginning side of the tank. One of them is in the central position and the others are in the si de positions. 6-2 The opened tank is connected with a polyuretan storage tank used for the radioactivity cooling system. Nal (Tl) sintillation dedector and connecting with a multichannel analyser system were used for the measurements. A crane system was constructed over the tank for placing the detector on it. Therefore, the scaning measurements could be relazied in x and y directions over the tank. Fig.l shows the expe rimental system schematically. EXPERIMENT Sodium-24 (Na24) had prefered as the tracing radioisotope because of short half-life and similar properties as water. Table : 1 shows the some proper ties of sodium and combined elements (5). Table 1 Some Properties Of Sodium And Combined Elements Element Half-life Energy (KeV) Sodium-24 14.96 h 1368.60 2754.00 Oxygen 26.91 sec. 197.14 1356.84 Clorine-38 37.24 m 1642.69 2167.68 Sodium was used as sodium-carbonate (NaoC03). It could be used salt (NaCl). But, sodium carbonate was prefered due to carbon and oxygen are more convi- nient for the irradiation properties than the clorine. The properties of them are also shown in Table 1. 6-3 Val v e s .Water supply Discharge f r~i I—t>0—*» Power supply « Detector CT* 1 lir Experimental Storage ^lul tichannel Tank Tank analyser Set Fig. 1- Experimental System -o* Sodium carbonate was irradiated in the central thimble of ITU TRIGA Reactor 1? -1 -1 in 144 seconds at the flux value of 8,1 x 10 cm sec . It was also irra diated with the rabbit system. The experiments had been realized after two hours later from the irradiation time. After irradiation, normal solution was prepered in 10 cc which contains 0,23 gr sodium in it. Under this circumstances, the specific activity of sodium was 4.35piCi/qm approximately 1 mCi activity was used for each experiment. Based on the simple one-dimensional dispersion model, the allowable maximum amount of injected radiotracer can be evaluated by using the following formula (4). 2 a VnTTT Amax=—^ (MPC) (]) Allowable maximum amount of injected tracer (yCi) max k : Dispersion coefficient v : Mean linear velocity of the stream (cm/sn) MPC : Maximum permissible concentration of the radioisotope a- : Average cross-sectional area under the control of the experiments (yCi) x : Distance along the stream Dispersion coefficient (k) can be taken as for the streams with little or -3 3 no dispersion such as in canals (4). MPC was taken as 6.10 yCi/cm for Na-24 (6). With these values, our activity level is below the maximum amo unt of injected tracer. Thus, the condition of the experiment was safe. An important parameter of the radiotracing technique in the channels, canals or pools is the flow rate. In the first trial, only the central valve was opened with the 500 cm /minute volumetric flow rate value. After that, the experiment was repeated for the only the one side valve was opened with the same flow rate. Lower detection distance is important point to insure the healthy measure ments. There are some semi-emphirical formulas for estimating this distan ce (4). One of them which was used is; 6-5 1/3 Lmin-KQ (2) Lmmjm„ : Minimum detection distance Q : Flow rate K : Coefficient If the stream width is large, coefficient of K can be taken 200 for side injec tion and can be taken 50 for the central injection (4). In our cases, this coefficients were taken tenth value of than due to narrow width. Under this circumstances, first detection scanning line was selected 50 cm. far from the injection line that is to insure the necessary distance for both of sides and central injections. The second and third detection lines were selected at 110 cm and 170 cm. far from the injection line respectively. Detection measurements were taken at three points as the central and two si des positions on the each lines. Therefore, the experiment was realized with the scanning of nine points. Before the radiotracer injections, the counting values for the background levels were taken for each cases. Only the counting value of the two peaks for the sodium-24 were taken from the multichannel analyser as counting value. The experiments were also realized with the methylen (C,fi H17 N?S) that have strong blue color in the water. Therefore, the experiment could be controlled by the visual inspections. RESULTS OF THE EXPERIMENTS Continues measurements were taken for each types of the experiment. There fore, at least ten counting scanning sets were taken in two hours for cent ral and side injections separetely. Fig. 2-4 show the sacanning counting results for central injection and Fig. 5-7 show the sacanning counting re sults for side injection. If the Fig. 2-4 are investigated, it can be seen that the dispersion of the radiotracer is occured line by line to the downstream direction. Furthermo- 6-6 t\ 'r 9 t ' «C._ J iy Set 2. Set 3. Set 4. Set Fig„ 2- The Scanning Counting Results For The Central Injection "V,\ 5. Set 6. Set 00 7.Set 8. Set Fig. 3- The Scanning Counting Results For The Central Injection 9.Set 10.Set <3"> i .Set Fig. 4- The Scanning Counting Results For The Central I njection i. O /j re, some shape of the scanning counting results of the central injection are identical each other (e.g. 2,5,8 and 3,7). Last three sets of them are also identical. It shown that the dispersion reaches the regime at the half and hour. On the other hand, if the Fig. 5-7 are investigated, it can be seen that the dispersion of the radiotracer is occured by the circulations in the tank. It can be concluded that it is because of the closed and rigite side structure of the tank. At the begining of the experiment, the opposide side region according to the infection point was as a dead point. In the further dispersion stages, the radiotracer reached there. Also at the begining of the experiment, the sour ce point of the radiotracer is naturally the injection point (1. point). Half an hour later, the source point was changed to central point of the first line (2.point). After the third half hour, the source point was changed again to the central point of the tank (5. point). Fig. 8 and 9 show the counting rate of the scanning points respect to time. It could be concluded from these graphs that, the both type of the injections, the great change in the direction of the flow were occured in the first 30 minutes. Afterthat, the changing rate was decreased. Approximately, an hour later from the injection, homogeneous flow be reached and the stability was obtained. The visual inspection by using the methylen, was within the same pattern as the radiotracer experiment in the first 30 minutes. Afterthat, the homogenity in the tank and patterns couldn't be followed by visual technique. CONCLUSION With this study, the investigations of the fluid dispersion in the canals by using radiotraces was initiated in our laboratories. The differences of the injection points were evaluated for the same volumetric flow rate.These radiotracer experiments were easily used for the pollution studies of water for different models. 6-10 i 9 3^ -=r^- "X.—T < -"1 V c ,/ 4 17) 4 l.Set 2.Set I 4 3. Set 4. Set Fig. 5- The Scanning Counting Results For The Side Injection 5.Set 6.Set tr 5" \ V s^S> 7.Set 8.Set Fig. 6- The Scanning Counting Results For The Side Injectnjectioi n V t 9.Set 10.Set Fig. 7- The Scanning Countiny Results For The Side Injection Sayim ,, 6000 \ 5000 4000 i 3000 2000 1000 30 60 90 120 Zaroan (dak.) Fig. 8- Counting Rate Of The Scanning Points For The Central Injection -*^ Sayim 7000 6000 5000 4000 ON 1 3000 2000 1C00 30 60 90 120 Zama« (dak.) Fig. 9- Counting Rate Of The Scanning Points For The Side Injection v> This study is the first stage on this subject. It is already to investigate the effects of the parameters and the effects of the permutation of the injection points. As a conclusion, one could mention that the visual inspection with methylen is not sensitive and reliable like a radiotracing technique. Therefore, the radiotracer technique is more convenient and reliable method for the waste water managements. REFERENCES 1- Bock, H., "Operating Experience And Maintenance Of The 250 KW TRIGA Mark-II Reactor Vienna In The Period July 1980 To July 1982", Seventh European Conference Of TRIGA Reactor Users Conference Papers TOC-15, 1st. Turkey, Sep. 15-17 1982. 2- "TRIGA Mark-II Reactor Description", Golf General Atomic, GA-8848, 1968. 3- Gardner, R.P., Ely, Jr. R.L., "Radioisotope Measurement Applications In Engineering", Reinhold Publishing Corparation, Newyork, Amsterdam, London. 4- Gardner, R.P. et al, "Laboratory Manual For The International Training Course On The Use Of Radiotracer Technique In Industry And Environmental Pollution", North Carolina State University, U.S.A., July 1973. 5- Glascock, M.D., "Tables For Neutron Activation Analysis", University Of Missouri, May 1988. 6- "National Bureau of Standards Handbooks on Radiological Protection Exposure Limits", Washington, D.C., U.S. Government Printing Office. 6-16 / / EVALUATION OF SELECTIVE BORON ABSORPTION LN LIVER TUMORS 1 11 D. Ckiaraviglio , F. De Grazia , A. Zonta tf a o o o 5. Altieri , B. Brachieri , F. Fossati , P. Padroni , T. Pinelli A. Perotti , S. Specshiarello 5 5 G. Perlini , H. Rief 1) Dipartimento di Chirurgia, Università di Pavia, Pavia (Italy) 2) Istituto Nazionale di Fisica Nucleare, Sezione di Pavia, Pavia (Italy) 3) Dipartimento di Fisica Nucleare e Teorica, Università di Pavia, Pavia (Italy) 4) Dipartimento di Chimica Generale, Università di Pavia, Pavia (Italy) 5) CCR Euratom lepra, (Italy) This study is part of a project which proposes the introduction of NCT in the ther apy of Multifocal Hepatic Neoplastics (MHN) by means of Extracorporeal Treatment of the Liver (ETL). This method unites the advantages of NCT's antitumor selectivity with those offered by ETL's organ selectivity. The patient's liver can be explanted, pre served artificially and successively reimplanted by means of hepatic autotransplantation 5 This technique is made possible utilizing methods of extracorporeal preservation 2 designed and performed by the Department of Surgery of the University of Pavia. sblated irradiation of the liver is advantageous due to the possibility to plan and create a substantially uniform neutron field around the organ. Studies are in program in order to obtain such conditions in the facilities of the Triga Mark II reactor of the University of Pavia. The first step of our project was a pharmacokinetic study to identify substances which are good boron transporters and are therefore able to provide a high concen tration of the nuclide with respect to the healthy hepatic tissue in the MHN. For this purpose the tumor M5076/73 (M5) [ij, which matastasizes spontaneously in liver, was inoculated subcutaneously in a group of C57B1/6 mice. Thirty days after the inocu lation, when 90% of the liver was invaded by metastases, a boric acid 0.3 M solution enriched to 96% l0B was injected into the caudal vein of each mouse (50mg/Kg). The mice were then sacrificed, some after 1 hour and some after 6 hours and the same pro cedure was performed with an equal number of healthy control mice*. The liver of the mice was then removed and frozen; samples were then prepared to use for the measure ments. Boron concentration in the various samples was achieved by measuring the energy distribution of a paritcles produced in the nuclear reaction l0B(n, a)7Li induced by a thermal neutron beam extracted from the Triga Mark II reactor, a particle energy was measured by a spectroscopic apparatus including a surface barrier totally depleted Si detector 19/^m thick. The experimental data were eleborated by the following model (see figure 1). Experimental results [4] prove that in the case of a particles having energy larger than 1.1 MeV the range - energy relation : R{E) = aE 1.74 can be utilized. The above relation makes it possible to calculate the dispersion Ax of the quote x(E) inside the sample where the particles releasing the energy E on the detector have originated Az = a{{E + AE)1-7* - E1-74] 6-17 Ai? indicates the energy dispersion of the spectrometric channel having energy E. The value of the constant a is 1.34 • lO~sfxm/KeV when a particles are transmitted across biological tissues. The reaction rate N induced by he neutron beam in the sample layer included between x and I + AI results to be JV = n^cr$ = ^a[(E + AE)"* - E^\ Here n is the quantity of 10B atoms included inside the sample surface layer in front of the detector having thickness Ro which is the range of the most energetic a particle detected in the spectrometer apparatus (1.47 Mev). The rate C of a particles impinging on the detector in the energy interval between E and E+ AE is: C = r,gN where r\g is the geometrical efficiency of the detecting system that was evaluated by a Monte - Carlo computing program. The formula: Ro£ 1 74 x T)g T/N(lh) = 2.7 ±0.6 T/N{Qh) = 2.7 ± 2.6 The first of the above values achieved in this experiment seems to be close to that of brain tissues [3] and provides a rather favourable indication regarding the applicability of NCT to treat liver metastases and tumors. Due to a high background in the detecting system, the given data are affected by a relevant error; in the future the experiment will be repeated utilizing a thinner detector (about 10/xm thick) in order to obtain a drastic background reduction and more accurate measurements of T/N. * We would like to thank dr. G.Pratesi of the Istituto Nazionale Studio e Cura dei Tumor i-Milano 6-18 SAMPLE DETECTOR H-R0-H V. 10mm E + AE 20mm V i i X 0 X X+£X 70mm D — R0 = 8.5/W m 0 = 19|üm Fig 1 Model for the expérimentais a spectra elaboration 6-19 *10 en Ü 2.4 D O 0 TUMORAL LIVER o A HEALTHY LIVER 2. e • 1.6 1.2 0.8 * à. * * 0.4 i* ***** - 9* A 0. * i I- 0.6 1. 1.4 1.8 *10 ENERGY KeV Jig 2 Spectra obtained from samples drawn 1 hour following the injection of the borated solution 6-20 REFERENCES 1. Àlessandri,G., R.Giovazzi, P.Falautano, F.Spreafico, S. Garattini, A. Mantovani: A murine ovarion tumor with, urique metastasizing capacity. European J. of Cancer 17 (1981),651 2. Bonoldi,A.P., P.Dionigi, C.Tibaldeschi, A.Zonta: Extracorporeal circulation or hy pothermic liver plushing before normothermic liver perfusion with whole blood. Proc. Europ. Soc. for Artificial Organs 7 (1980) ,240 3. Hatanaka,H., WH.Sweet: Slow-neutron CT for malignant tumors IAEA-SM (1975),147| 4. Kobayashi,T., K.Kanda: Analytical calculation of 105 dosage in cell nucleus for neutron capture therapy. Radiat.Res. 91,(1982] ,77 5. Zonta,A., P.Dionigi, N.Del Ciotto, D.Chiaravigho, M.Alessiani, G.Bellinzona, C.FerrariJ C.Tibaldeschi: Impiego sperimentale di un nuovo tipo di protesi vascolare in corso di autotrapianto ortotopico di fegato. II Giornale di Chirurgia 7 (1986),631 6-21 A NEW HANDLING TOOL FOR IRRADIATED SAMPLES AT THE LENA PLANT L.Alloni, A. Venturelli Labofatorio Energia Nucleare Applicata, Universita di Pavia, PAVIA (Italy). INTRODUCTION The handling of neutron irradiated samples at the LENA plant has been so far carried out manually, thus exposing reactor and health physics operators and reactor users to radiation doses. It was then decided to develop an automatic system ope rated from the reactor console. The system was divided in two sections: one taking care of sample insertion and extraction and the other of the storage of irradiated samples- This paper describes the design and the installation of the storage section. It allows a fast removal of the irradiated samples from the reactor top and their storage in lead pits at the ground level.The extraction of irradiated samples comes out to be quite simplified and radiation doses to operators and users are strongly reduced. All work from design to construction has been carried out by the personnel of the electronic group of the LENA plant. DESCRIPTION OF THE DEVICE The new device is located on the east side of the Triga Mark II LENA reactor. A view of its location is presented in Fig. 1 The main parts of the device are: - main switchboard, located on the reactor top - storage box control panel, located on the ground floor close to the storage box - electronic equipment for the system automation, located on the ground floor - reversible motor and related scaling - two-way worm driving the trolley for sample deposition - conveyor for irradiated samples from reactor top to the storage box - storage box containing 10 pits for sample storage 1) Wain switchboard This panel, located on the reactor top, is shown in the drawing of Fig. 2. It consists of : - n. 10 light-emitting diodes (LED-A): they indicate if the storage pits are occupied or empty; - n. 10 light-emitting diodes (LED-B): they indicate the pit over which the trolley is operating; - n. 1 light-emitting diode (LED-C): it indicates if the trolley has reached the desired position (pit); 6-23 Main switch board Irradiated sample input Reactor top to -P- 6.95 in Storage box control panel Ground level Fig. 1 - VIEW OF-THE EAST SIDE OF THE LENA THIGA MARK II REACTOR. Led A o o o o o o o o o o Led B o o o o o o o o o o On off motor (j Led C Push - Push Led E button button Led F (?) o o o o Push - buttons (G) o o o (o o o o o o o 1 2 3 4 5 6 7 8 9 10 Fig. 2 - MAIN SWITCH BOARD. -n. 1 on-off switch for the reversible motor; - n. 1 light-emitting diode (LED-E) and push-button switch: the LED indicates the direction of the displacement of the trolley (right to left), the switch activates the proper rotation of the motor to ensure the wanted displacement of the trolley; (Fig.3) - n.1 light-emitting diode (LED-F) and push-button switch: their function is similar to the one just described and indicate the trolley displacement from left to right; - n. 10 push-buttons (G), with automatic return: they excite the memories (Fig.4) to displace the trolley over the desi- dered pit. 2) Storage box control panel This panel, located on the ground floor close to the storage box, is shown in the drawing of Fig.5. It const its of : - n. 10 light-emitting diodes (LED-H): they have the same function of LED-A in the main switchboard ; - n. 10 push-buttons (I) with automatic return : they reset the memories indicating which pit is occupied (Fig. 6). 3) Electronic equipment This equipment, located nearby the storage box, includes: - n. 10 memories for trolley positioning ; - n. 10 hold-on memories, activated by the insertion of samples in the pits. - n. 1 hold-on memory for rotation direction of the reversible motor. - information and power circuitry. 4) Two-way worm, motor and trolley The worm is located over the storage box and it is driven by a reversible motor,which is also provided with a scaling device. One end of the worm is directly connected to the scaling device whereas the other end fits in an autolubricating bearing ( Fig.7). A trolley is mounted by means of a volute on the worm and may be displaced left to right and viceversa. The trolley contains the terminal end of the conveyor. The lower part of the trolley contains half-moon devices which act on the limit switch of each position (pit). The latter are mounced on T-guides, set parallel to the worm. Two limit switches are located at both ends of the worm for automatic inversion of trolley displacement. 5) Conveyor The transfer of irradiated samples, usually in their irra diation containers, from the reactor top to the storage box, is made possible by means of a conveyor. It consists of a pipe (5 cm , I.D.), whose initial and terminal parts are made of rein forced rubber, whereas the middle part is made of alluminium. The path of the conveyor includes smooth curves that are useful 6-26 • 12 V dc FA 22 KO IN 4007 J o FI 22 KO Start -MW— IN 4007 47 Kfl -O O- 3,9 KO BC 337 ON •K I Beset •4 -O O- 1 Mfl IH 4007 100 yft~J -AW—H 3,3 (CO? O.IS pFT: >22KA Fig. 3 - MEMORY EXCITATION FOR RIGHT OR LEFT DISPLACEMENT OF THE TROLLEY. + 12 Vdc — Rl 22 Kß IN 4007 9- —'WW—| 0 22 Kß -M/VV— IN 4007 22 Kß _W1 vvvs— BC 337 IN 4007 K i M 00 '47 Kß H ft 47 pF ~ 0,1 pF rp 0,22 pF Trolley positioning Reset + 12 Vdc Fig. 4 - MEMORY FOR LOCATING THE TROLLEY AND RELAY FOR MOTOR EXCITATION. Led H O O O O O O o o O o Push buttons (I) 0 O O O o o o o 0 o 12 3 4 5 6 7 8 9 10 Fig. 5 - STORAGE BOX CONTROL PANEL. Reset 2 Fig. 6 - HOLD - ON MEMORY TO IDENTIFY OCCUPIED PITS. Conveyor Storage box Condenser control panel T - Guide jTOtffiföttftti^ Photo- Light receiver source Electronic Equipment ON I Liait switch Worm \ Half-Koon Bearinj Trolley Device trtr- n >j y y lDJ \ Q. H! . tflU im" • • ffltffi, J£R J n n n... n ,. n - t\ n n n n n ' Jf Fig. 7 - FRONT AND PLANT VIEW OF STORAGE BOX, WORM AND TROLLEY. to slow down the sample descent. In the upper part of the conveyor there is a funnel where irradiated samples, coming from irradiation facilities available on the reactor top, are deposited. The lower end of the conveyor consists of an allumi- nium conic nose, whose internal diameter is just a little bit larger than the diameter of the usual irradiation containers. Samples are thus drastically slowed down and gently dropped inside the prefixed pit in the storage box. 6) Storage box The storage box is actually a wooden box (150x40x30 cm) , internally divided in 10 sections (pits). The bottom of each pit is coated with rubber sponge acting as shock absorber for the dropping samples. Each pit is then coated with plastic film. The whole box is shielded with 5 cm lead bricks. On top of each pit there are a light source and a photore- ceiver, that provide information on the occupancy of the pits. The full circuitry and system logic are presented in Figures 8 and 9. OPERATIONAL INFORMATION The following procedure has to be followed for storing an i rrad i at ed samp1e- Before sample extraction from the irradiation facility, the operator gives a look at the main switch board and becomes aware, looking at LED-A, of which pit is empty. He is also informed on which position ( pits 1 to 10 ) the trolley is standing, by looking at LED-B . Then he selects the pit into which he wants to store the irradiated sample by depressing the proper push-button ( G ). This charges a memory and excites the relay for motor startup ( Fig. 3 ). By depressing push-buttom E or F it is possible to activate the proper rotation of the motor so that the trolley reaches the desired position as soon as possible (Fig.7). When the trolley is over the selected position the half- moon device depresses the limit switch of that position: the previously excited memory is resetted and the trolley stops. The operator is informed about this step when LED-C lights on. Now the sample can be extracted from the reactor and dropped on the funnel on top of the conveyor : the sample runs down into the conveyor and after being slowed down by the braking device drops into the selected pit. When the sample cross the beam of the light source in the upper part of the pit an analog signal excites the pit hold-on memory and the proper LED-A on main switchboard and LED-H on storage box control panel light on. Whenever the sample has to be withdrawn from the pit, the health physics operator monitors it for radiation dose and resets the occupancy memory by depressing the proper push button ( G ) on control panel : at the same time LED-H on control panel and LED-A on main switchboard will be switched off. In order to reduce further radiation doses to reactor 6-32 T^-cg tg cgjçg q a Cg eg çg ^ y Motor U \\ H \\ W « M « » W H « «TTQ l "' «' « '" » « " » » » " « H " I' I Stop Stop 110 V.a.c. 12 Vdc H.L.l 22 Xfl IN 4007 1 ^ 22 XO ;;= oil Start IN 4007 22 xn ac 337 IN 4007 1^ 0.22 1 Res«t •5Ht-J 1Qi- ! «47 JIT T jif 47 xn UJL P2 IN 4007 _il2 M.L. 2 —M— •o o— P10 IN 4007 22 Kfl |y H.L. 10 ..no -H- vW»V-r--KB •0 0— 47 Kfl 22 xn IN 4007 22 Kfl IN 4007 47 KO DC 337 47 Xfl BC 337 3.3 X(} ;2.2 J 47 xn .22 ? .22 rxn 47 xn T "F1 ** T fig. 3 - CENERAl SCHEME FOR MOTOR EXCITATION AND TROLLEY DISPLACEMENT. 6-33 4 ® i ® £ ® £ ® £ ® £ ® £ ® ^ ® i ® bp 10 12 Vde 2 x 2,2 KO Storage box control panel. ÇfCf) (j) 9 (j^ff ft fo Memory 0 N. 2 13 ta GO E) 0 (5J 13000 Main switchboard Al §f§ 12 Vdc A-Al - pit occupancy 0 - reset pit C - trolley position 320 Vac 0 - trolley In right position ~J 1000 yF Z - . trolley displacement •Fig. 9 - CENERAL SCHEME FOR PIT SELECTION ANO OCCUPANCY. 6-34 personnel, an automatic sample insertion and extraction to/from central thimble and rotatory specimen rack and its connection with the handling tool for irradiated samples described in this paper, has been accomplished. The lazy Susan connection construction is now underway. 6-35 IM ' / / ; •Vi t DETERMINATION OF ESSENTIAL AND TRACE ELEMENTS IN MILK AND MEASUREMENT OF SHORT-LIVED NUCLIDES USING FIMS v X •• X XX XX R.Demiralp , S.Kalayoglu , E.Unseren , F.Grass , H.Bock x * • Istanbul Technical University Institute for Nuclear Energy Maslak-ISTANBUL x »• « • * • Atominstitut der Osterreichischen Universitäten, Schütteistrasse 115, Vienna-AUSTRIA Milk is a slightly yellowish-white colored dense liquid with a level of nut rition and contains considerable amounts of almost all nutritional elements. Milk is consumed in its natural form or in the form of yoghurt, butter, cheese or similar milk products. The constituents of milk are water, fat, subtances soluble in fat (milk fat, lecithine, Cholesterine, carotines, A, D, K vitamins) , proteins (casein, lactalbumin, lactalglobulin), carbohydrates (lactose), salts, substances soluble in water (critic acid, vitamins Bx, B2 , B6, Bl2> vitamin C, miacine, pantothenic acid, choline), gases ('CO2 , N2 , O2) and enzymes. Its main nutri tional value lies in the fact that it is a source of protein, calcium phos phate and vitamin. The inorganic constituents are mainly alkali and alkaline earth chlorides, phosphates and sulphates. These substances are found in milk partially in the form of organic compounds, i.e., bound to casein and phos phatides , The most abundantly found elements are Ca, Cl, Mg, Na, K, S, P and N. Furthermore, trace amounts of Zn, Fe, Cu, As, Br, F, Co, Ni, Mo, Mn, Cr, Se, Pb, Si, I, Sb, Bi, Li and Eu are found (1,2). The pH of the fresh milk is 6.3-6.6 and the density is about 1.03-1.033. Pro teins are dispersed in the form of colloids whereas fat as droplets in milk. White color of milk comes from the light reflected by dispersed casein and calcium phosphate. Yellowish color is due to riboflavin and Carotine pignents. The most important factors effect the constituents of milk is temperature and microorganism. In order to preserve milk without being impaired , antiseptics , bor derivatives, formol, salisilic acid, perhydoxy benzoic acid are added. Milk is marketed in its natural form or after pasteurization and homogeniza- tion. Milk that a significant portion of its water evaporated in vacuum is called condensed milk. Sometimes milk is marketed by enriching with the ad dition of vitamin D, sugar and malt. Another important way of using milk is the milk powder. Milk is homogenized by passing through small pores under high pressure. Fat droplets found in milk are of 2-10 u in diameter. If milk is allowed to stand 6-37 these droplets accumulate. The diameters of the droplets are diminished to 1.5-2 u by homogenization. For sterilization milk should be homogenized. Pat hogen bacteria are destroyed by boiling milk, but boiling changes the taste of milk and destructs vitamin C. Pasteurization is made by immediately rai sing the temperature of the milk to an adequate level (70°-95°C) keeping there for a short time and promptly cooling to 4-8°C. In this way bacteria are destroyed without impairing the physical, chemical and biochemical cha racteristics of milk. Milk powder is made by evaporating water after completely removing its cream and contains dry substances of milk. Milk powder used to be prepared by eva porating milk at 10Q°C and grinding the dry residue. Milk powder obtained from milk dried at such a high temperature is difficultly soluble. Today cylinder or propelling methods are used which permit milk to contact heat for a very short time. Well prepared milk powder is a foodstuff with high nutri tional value and is very practical as it is easy to carry and to keep it. Milk powder is used in the various food industry, mainly preparing baby foods, in chocolate, and biscuit manufacturing as well as in pastry. Experimental Part The preparation of the Samples : In the experiments, Gülüm, Sek and Pinar brand of bottled milks and Pinar milk powder which are commercially available were used. Pinar milk is being processed in Izmir while Gülüm and Sek milks in Istanbul. Milk was first kept in deepfreezer for 12 hrs at -20°C, then dried for 48 hrs in freeze dryer (3). Pellets of 1.3 cm in diameter were prepared by 10 tons of press from the dry samples kept in dessicator for irradiation. Standards As standards IAEA Milk Powder A-ll, NBS-Orchard Leaves 1571 and for Cu single standard was used. Irradiation Samples and standards were irradiated together in the central thimble of I.T.U. TRIGA Mark-II reactor for 1-8 hrs and for 60 sec in the fast pneuma tic tube of Atominstitut der Österr. Univ .TRIGA Reactor. Depending on the nuclear characteristics of the isotopes to be analyzed, they are counted at different counting times (600 s-2 hrs) after 20 ms.-2 week cooling periods. Table 1 lists the energies of the gaam rays of radionuclides used in these determinations (4.5). 6-38 Table - 1 Nuclear Data for the Elements Determined Radionuclide Half-life Principal Gamma-ray Energy (KeV) 2.25 min. 1778.8 28A1 9.45 min. 1014.4 27Mg 5.11 min. 103 9 66Cu 4.54 d. 1296.8 49Ca 56v 2.58 hr. 1811.2 Mn 1.48 d. 554 80Br 12.36 hr. 1524.7 42K 15 hr. 1368 24Na 38ci 37.2 min. 2167.5 18.7 d. 1076.6 86Rb 26.3 hr. 559 76As 2.68 d. 564 124 Sb 3.75 min. 1434.4 52V 243.7 d. 11I5 65Zn 5.27 yr. 1332.5 60Co 33.8 d. " 1120 46Sc 45.1 d. 1099 59Fe 27:. 8 d. 3 20 51Cr 198 2.7 d. 411 A4 u Counting Systems Two counting systems were used : 1. The activities were measured with a high-pure Ge dedector coupled to Can berra 90 model multichannel analyzer in the ITU. The system has a resolu tion of 2 KeV for the 1332.5 KeV gamma ray of 60Co 2. In order to determine short-half life nuclides a very fast irradiation and measuring system (FIMS) has been used in the Atominstitut der Öster. Univ. The system has a Ge(Li) dedector whose resolution 2.8 KeV and a preampli- ficator which can operate at high count rates with an additional circuit. There was a PDP 11/34 computer with dual hard disc which automatically transfers and ejects the samples from irradiation position and which sto res the data. In this system the transfer process is made by nitrogen gas at about.10 atm. pressure (6). 6-39 Results and Discussion Tables 2, 3, 4 and 5 give the amount of trace and essential element concent rations in daily and pasteurized milks. Trace and essential elements in Pmar, Sek and Giilüm milk products and Pmar milk powder which are produced and consumed in our country are determined by instrumental neutron activation analysis. The average values obtained are gi ven in Table 6. When the average values of the 16 elements shown in the table are considered, it is observed that the amount of Na, As, Al, Mn, Zn, Rb , Co in the milk pow der is greater than that of milk, where as in Pinar milk, which is a durable kind of milk, the amount of Na, K, Br, Al , As, Zn, Co is larger than that in daily milk. When daily products of different brands are compared, it was fo und that the quantity of as, Fe, Cr, Mg is higher while the of CI, Sb, Zn, Rb, Co is less than that in Gülüm milk. The quantity of CI is about 10 times as much and Mg 7 times as less as that in Sek milk. It was not possible to investigate how the seasons and the regions from where the milk was collected effected on the quantity of the elements. It will be useful to continue the study in this field. The author (R.D.) is intebted to Mr.S., Mr.I.Saleh and Mrs.J.Dorner (Atominstitut) for technical assistance and also to the IAEA supporting partly this work. REFERENCES 1. H.Keskin, Food Chemistry, Istanbul University Publication, Nr. 1525, 1970. 2. A.R.Byrne, L.Kosta, A Study of Certain Trace Element in Milk, IAEA-SM- 227/45, 1979. 3. A.BaijSari, XRF Analysis of Milk, M.Sc. thesis. ITU, 1987. 4. G.Erdtmann, Neutron Activation Tables, Verlag Chemie, 1976. 5. M.Lederer, J.Hollander, I.Perlman, Table of Isotopes, John Wiley, 1967. 6. F.Grass, Short-time Activation Analysis with Steady State, J. of Radional and Nucl. Chem., Vol. 112, 2 (1987), 347-359. 6-40 Table-2 Trace And Essential Elements in Pmar Milks Determined by INAA Sample Name K K xxx H H xx c XXX r XX Na K* Br Mn** Rb** sb As*** AÏ** C1 Mg Cu* v Zn** Co*** Cr** Au*** tf a P-l 0.425+ 1.43+ 2.08+- l.6± 1.04+ 1.3+ 4.18+ 6.48+ 0.963+ 600 .8+ 0.91 0.36i 31.7+ 4.5+ 0.23f 0.19 17.9+ T.E 1.38+ 0.01 0.04 0.2 0.2 0.2 0.1 2.1 2 0.06 46 0.09 3.8 0.42 0.09 2.1 0.06 P-2 0.491+ 1.46± 1.06* 1.1 1 1.14± 1.22* 4.62+ 2,97 + 0.921i 528.9 + T.E 0.26t 44 î 4.2 + 0.239+ 0.25a 11.1± 14a 1.17 + 0.02 0.03 0.09 0.31 0.3 0.1 3 1.2 0.07 öl 0.05 2.7 0.02 0.03 1.7 0.05 P-3 0.436± 1.74± 2.29+ " 0.73+ 2.54* 1.38* 5.32 + 1.79 + 0.97 + 386.87+ T.E 0.19 + 44.9+ 4.52+ 0.487+ 0.32a 19.3s Ua 0.92 t 0.02 0.09 0.06 0.19 0.2 0.2 2.7 0.9 0,03 39 0.04 3.6 0.6 0.04 2.3 0.03 P-4 0.456+- 1.33* 1.33* 3.4 ± 1.56 + 1.16+ 4.6* 8.89* 0.867+ 755.13+ 0.74+ 0.6 + 56.5+ 3.71 + 0.5061 0.24± 17.3 + T.E 1.08± 0.04 0.06 0.08: 0.06 0.19 0.2 1.2 2.1 0.05 48 0.09 0.1 6.7 0.4 0.04 0.07 3.1 0.09 P-5 0.447+ 1.44* 2.17i 9.2 + 1.41 + 1.07+ 6.25+ 13.7 * 1.32 + 1370 f 0.85+ l.li 52.41 4.16+ 0.285+ T.E T.E T.E 0.911 0.06 0.03 0.09 0.1 0.2 0.09 2.4 1.0 0.09 23 0.2 0.2 9.6 0.3 0.01 Q.02 P-6 0.495t 1.37± 2.08* 0.43+ 2.26 + 1.062+ 6.43+ 9.08 + 0.89 + 746.12+ 0.82 + T.E 50.4 + 3.41+ 0.349a T.E 10.9i 8.9a 0.961 0.09 0.01 0.02 0.12 0.1 0.1 1.9 1.1 0.09 35 0.15 9.8 0.42 1.7 0.16 P-7 0.425t 1.38* 1.62+. 0.51 + 7.3 + 1.08* 6,95* 12.84* 0.758* 986.781 0.95+ 1.3 + 40.6a 4.02a T.E T.E 9.6+ 12.6a 0.93± 0.01 0.02 0.03 0.05 0.7 0.2 2.9 1.9 0.08 28 0.21 0.3 2.0 0.09 P-8 0.415+ 1.2 t 1.23 + 0.41+ 5.4* 0.94± 7.4* 11.87 + 0.924+ 253.04+ 0.6± 0.2 + 47.5! 3.38+ 0.852+ 0.58± 19± 4.5a 0.78 + 0.03. 0.02 0.05 0.2 0.5 0.2 2 2.2 0,09 46 0.14 0.06 6.5 0.5 0.07 0.1 2.8 0.03 a P-9 0.423* 1.34± 1.54t 0.68 3.46* 0.95* 5.52 + 5.2 i 0.89 + 806 .59+ T.E 0.73+ 36.5+ 3.39+ 0.105a 0.39 + 15 + 5.07a 0.94± 0.01 0.01 0.05 0.2 0.5 1.9 1.1 0.07 31 0.2 4.3 1 0.08 2.2 0.02 P-10 0.49 + 1.4 i 1.351 2.2>* 2.06 + 0.96+ 7.08+ 11.72 4 0.73± 832.37+ T.E 0.91± 28.9t 5.78+ 0.222* 0.54+ 19.3a T.E 0.86* 3.02 0.02 0.02 0.4 0.1 0.3 1.7 2.01 0.09 31 0.22 4 0.8 3.15 0.2 0.04 x : % xx : ppm xxx : ppb a : one measurement value T.E : not detected Table-3 Trace And Essential Elements in Giiliim Milks Determined by INAA Sample Name x xx xxx xx KH xx XXX Na KX BrX Mnxx Rb Sb Asxx> Al C1X Mg CuXX V Zn Co ScXXX |;r xx Fexx Auxx Cax G-l 0.38* 1.4* 1.7* 4.02* 2.1* 2.8* 1.4* 8.56± 0.733+ 145,8+ T.E 1.3± 39.lt 2t T.E T.E 12.2± T.E 1.006+ 0.02 0.13 0.04 0.3 0.2 0.8 2 3 0.3 13.2 0.1 6.7 0.09 2.4 0.04 G-2 0.352i 1.36i 1.99* 1.94+ 1.751 2.1 + 1.98a 3.47* 0,91 + 56,9+ 0.89a 0.71+ 34 * 1.6+ 0.361^0.26+ 15.7+ T.E 0.97 + 0.04 0.05 0.03 0.05 0.3 0.09 1.2 0.04' 10,1 0.1 5,9 0.08 0.03 0.09 2.8 0.06 G-3 0.379* 1.39s 1.66* 4.46t 1.88* 1.96* 2.09* 8.82 + 0.97 + 135.2+ 0.73+ 0.88t 41 i 1.5a 0.403j 0.29+ 14.6+ 6.85a 0.81 ± 0.08 0.07 0.05 0.9 0.3 0.03 0.9 2.6 0.1 26.1 0,09 0,08 4.3 0.06 0.06 3.4 0.03 * G-4 0.366+ 1.28s 1.84*. 1.62i 2.3* 4 ± 1.84i 1.81t 0,99 t 63.9+ 0.96+ 0.69± 40.3 + 3 * 0.96+ 0.43+ T.E 7.3a 0.73 + 0.05 0.06 0.09 0.33 0.2 0.06 0.8 0.09 0.24 13.2 0,08 0,07 6.2 0.7 0.1 0.1 0.09 G-5 Q.367± 1.37* 1.67* 1.14t 2.95* 3.63± 2.13± 1.39* 0.83 ± 40.9* 1.6t 0.921 29,3* 6 + 1.6a 0.51* 15.7+ 10.2a 0.92a 0.09 0.02 0.03 0.04 0.4 0.1 0.7 0.09 0.09 9.3 0,2 0,08 2.9 0,26 0.1 2.9 3 G-6 0.3651 1.38i 2.15» 2.001 3.21+ 2.94a 2.33+ 2.22* 1.006«- 65.5i 0.68* 1,01 + 37.2± 1.3+ 0.88* 0.34* 13.2+ T.E 0.86, 0.1 0.03 0.07 0.7 0.07 0.1 0.07 14,1 0.17 0.2 3.8 0.3 0.07, 0.08 2.5 0.02 G-7 0.371± 1.37* 1.87* 4.35± 2.82+ T.E 2.741 8.35+_ 1.13 -r 138.7+ 1.84t 0.96+ 49.1* 1.1 + 0.48+ 0.27+ 9.4* T.E 0.89 + 0.02 0.03 0.03 0.1 0.1 0.09 2.4 0.15 10 0.15 0.U3 6.8 0.09 0.09 0.09 0.9 0.08 a a a G-8 0.372+ 1.31* 1.73«-5.72* 5.6 + 2.01* 2.35+ 11.7 + 0.83 + 193.9+ 1.71 + 0.66 66 T.E "ÖT77 " 0.54 + 20.1+ 10a 1.11 + 0.03 0.09 0.04 0.19 0.2 0.5 0.05 3.1 0.08 29.2 0.21 0.12 3.7 0.09 Table-4 Trace And Essential Elements in SEK Milks Determined by INAA Sample XXX XXX XXX XX x Name Na' K* Br7 Mnx x Rbx x Sb As Al xx CV Mgx x Cux x ,xx Znx x Cox x Sc Cr Fe** Au** Ca* SEK-3 0.356+ 1.41 + 1.89* 3.97+ 2.3± 1.8* 3.82+ 10.87+ 0.81± 997.9± 1.2i 0.48+ 34.14+ 8.3± 0.71 + 0.95c 21.4+ .T.E 1.32+ 0.07 0.19 0.15 0.3 0.2 0.11 0.8 3.1 0.07 39 0.14 0.09 6.1 0.71 0.01 2.8 0.07 a SEK-4 0.309^ 1.2 ± 1.23 + 3.95+ 1.4* 2.1* 1.82+ 9.353t 1.12+ 820.13i| 1.8" 0.21 + 33.38+ 5.1± 0.36*. 0.73+ 19.8± 14.8 1.38 + 0.03 0.08 0.07 0.014 0.19 0.21 0.1 2.8 0.09 51.3 0.08 3.8 0.9 0.08 0.11 4.1 0.09 c a SEK-5 0.348:- 1.292 2.133t 3.05± 0.94t 0.72 4.595+ 4.59i 1.01 + 578.6± 0.73* 0.19a 17.84+ 6.3+ 0.48- 0.43 17.6+ 12.3 1.01 + 0.01 0.04 0.09 0.4 0.31 0.09 0.9 1,9 0.07 24 0.23 2.5 1.1 0.09 3.8 0.05 1 Tab 1 en-5 Trace And Essential Elements in Milk Powders Determined by INAA ampl^ xx xxx XXX xx X xx xx VKK xx xxx xxx xx xx x Nax KX Brx Mn Rb Sb As Al C1 Mg Cu Zn Co Sc Cr Fe *u*** Ca .-ine.. S-l 1.776+ 0.464+ 1.97+ 5.97± 6.3i 1.25± T.E 10.87+ 0.78+ 997.9+ T.E. 1.6t 51± 8.9+ 0.36* 0.46+ 16.8± T.E 1.002 + 0.62 0.11 0.2 0.2 1.1 0.08 1.2 0.09 32 0.21 7 2 0.03 0.11 5.7 0.07 a S-2 1.811± 0.461t 1.196+ 6.95 + 9.6 + 1.4 i 6.7-* 9.53 + 0.757+ 820.13+ 1.7 + 0,98+ 79 i 5.2 + 0.54+ 0.38+ T.E 7.9 0.868 + 0.16 0.08 0.09 0.66 2.0 0.05 2.1 1.0 0.08 47 0,3 0.19 8.5 1.6 0.11 0.08 0.06 S-3 1.82 + 0.449i 1.80*. 5.05* 8.7± 0.91 + 7.91+ 4.59i 0.533* 57876+7 1.52+ 0,81! 36.7« 6.4 £ 0.26+ 0.29t 11.4x 9.6-' Û.742+- 0.21 0.02 0.04 0.09 1.2 0.04 3 2,8 0.1 49 0.2 0.08 8.3 2,3 0.08 0.01 1.8 2.5 0.09 a 1.769' 0.47- 1.74 J 4.67± '8.It 1.4 + 4.3+ 14,35 i 0.14k 330.45+ 0,96 1.3' 49 J. 6.1+ 0.39+ 0.35+ 20.2 + 8.1± 0.963r 0.2 0.07 0.02 0.2 |2.2 0.03 I 0,03 28 0.3 9.3 2.5 0.03 0,03 6.3 2.7 0.08 Table-6 •Mean Value of Some Investigated'Elements Sample XX K Name Na* K* Br Cl* As*** Sb Fe** Cr XX AI Mnx x Mg** Zn * Rb** Sc*** Co*** P 0.45 1.67 2.03 0.905 5.94 1.11 15.49 0.46 8.45 2.03 726.7 43.33 2.83 0.328 4.11 G 0.365 1.37 1.826 0.925 2.11 2.83 14.41 0.377 5.79 3.16 105.1 42.06 2.8 0.779 0.564 i tn SEK 0.338 1.3 1.768 0.098 3.12 1.54 19.6 0.704 5.27 3.657 798,9 28.45 1.547 0.517 0.57 S 1.79 1.46 1.677 0.557 6.31 1.24 16.14 0,37 9.81 5.66 681.8 53.9 8.2 0.388 6.65 THE USE OF THE TRIGA MARK II REACTOR, LJUBLJANA, YUGOSLAVIA IN THE kQ METHOD OF NEUTRON ACTIVATION ANALYSIS B.Smodis"1, S. Jovanovic2, P. Vukotic , R. Jadimovic , P. Stegnar ^'Jozef Stefan" Institute, "E. Kardelj" University, Jamova 39, 61000 Ljubljana, Yugoslavia institute for Mathematics and Physics, "V. Vlahovic" University, Cetinjski put bb., 81000 Titograd, Yugoslavia ABSTRACT The kQ standardisation method is suitable for routine multi-element determinations in various materials by reactor neutron activation. A 21 - element synthetic standard for biological material was analysed in this work by neutron activation analysis and the results obtained, as well as our experience with the method itself, are discussed. INTRODUCTION The kg method of (n,Y) activation analysis with reactor neutrons (NAA) / 1-3 / was recently introduced at the "Jo2ef Stefan" Institute in Ljubljana. When using this method, the concentration of an element Y in a sample, c(Y), determined by counting an i-th gamma line emitted by the induced isotope X, is calculated from: 6-47 VV l f+Q0,Au^ Ep(AU) c(Y) - (1) Asp(Au) kQ(X.) f+Q0)X(a) Zp(X.) where A is the specific count rate, "A Asp= (2) WSDC and kQ is a composite nuclear constant, related to the i-th gamma line of isotope X, M 9 Au X °0,XYXi ko(Xi) (3) MX6AU°0,AUYAU In eqs. (1), (2) and (3) : N - peak area corrected for pulse losses; tmm - measuring3 time; W - mass; S = l-exp(-Xt. } saturation factor; t. „ - irradiation time; i-\ irr' irr X - decay constant; D = exp(- X t .) decay factor; t. - decay time; C = (l-exp(- Xt ))/( Xt ) factor for decay during measurement; M - atomic mass; 0 - isotopic abundance; 3n - 2200 ms (n,y ) cross-section; 6-48 y - absolute gamma-intensity; f - thermal to epithermal neutron flux ratio; QQ(a) » IQ(CX)/ « ; IQ(CX) - resonance integral corrected for a non-ideal ( assumed 1/E ) epithermal neutron flux distribution; £ - absolute peak detection efficiency; For an actual analysis, the following parameters should be known: 1. All relevant nuclear data (i.e. kQ, QQ, Î , decay scheme data, etc.). 2. £ , absolute peak detection efficiency curve for the detector and the actual geometrical conditions. 3. f, thermal to epithermal neutron flux ratio and a , a parameter in the 1 + rr 1/E epithermal spectrum representation. Since many calculations are involved, the use of computer programs are necessary. EXPERIMENTAL All relevant nuclear data were taken from a computer library /I,2/. The full-energy peak efficiency (e ) was determined by using a semi-empirical method /4/, based on the concept of the effective solid angle . When applying this method, a "reference" efficiency curve (log e versus log Ey) obtained by using a set of calibrated point sources at a large distance from the detector (15-20 cm) is constructed. The e for an actual geometrical configuration is then calculated, taking into account detector response and gamma-attenuation effects. The Q- calculation was performed with the SOLANG program / 4 / on a VAX 11 / 750 computer. True-coincidence effects were taken into account as 6-49 well. The method applied for this correction / 1 / requires only the experimental determination of the peak-to-total curves for various sample - detector distances using some mono - gamma emitters like 203 Hg, 137Cs , and Zn. Typical efficiency and peak-to-total curves are presented in Figures 1 and 2. The "bare dual monitor" method /5/ and the "bare triple monitor" method were used for instantaneous f and a determination. Therefore , the samples were coirradiated with Au and Zr standards, thus keeping the experimental work as simple as possible. Irradiation conditions and counting. The synthetic standard No. 0 ( 21 elements ) prepared by Mosulishvili et al. /6/ was analysed in this work. The samples were 50 mg resin pellets, 1.85 mm thick with a diameter of 5.8 mm. The samples were irradiated for 20 hours in the rotating rack of the TRIGA Mark II reactor of the "Jo2ef Stefan" Institute, Ljubljana. At this position, 12 -1 -2 the thermal flux is 1.0-10 n s cm , the thermal to epithenmal ratio is 20 and a is - 0.02. The comparators were Au-Al wire (0.1 % Au) and 0.125 mm thick Zr foil. An Ortec intrinsic Ge detector connected to a Canberra series 80 analyser was used for counting. The resolution of the detector was 1.8 keV at 1332.5 keV and the relative efficiency was 17 %. The relatively short-lived isotopes 198Au, 1318a, 82Br, 140La, 24Na and 122Sb were measured approximately one week after the irradiation and other isotopes after two months. 6-5C Evaluation of data The spectra were evaluated by the SEQAL program /!/ on a VAX 11 / 750 computer. For the overall calculations the SINGCOM program /8/ was applied. This program, starting from peak areas, calculates f, a , absolute peak efficiences and true coincidence factors, and gives the individual element concentrations together with the averages and standard deviations as a final output. RESULTS AND DISCUSSION The synthetic standard No. 0 prepared on phenol - formaldehyde resin (PFR) is 12 13 -2 meant for use in nuclear reactors with a neutron flux of 10 - 10 n cm -1 s . The standard is in the form of a pellet and due to its clear - cut geometrical shape and high mechanical strength is very convenient for instrumental neutron activation analysis (INAA), and particularly for its absolute versions. The standard is specified for 21 elements. We were able to analyse 19 of 65 them. Due to its unfavourable half - live ( Ni, T,/«3 2.52 h) nickel was not analysed, nor was neodymium, whose peaks were not identified in the spectra. The relevant nuclear data for the isotopes used for the analysis are summarised in Table 1. The results are presented in Table 2 and graphically in Figure 3. The precision of the results was better than 10 %, mostly even better than 5 %. The exceptions were gold and mercury with standard deviations of 18 % and 25%, respectively. The accuracy of the results was not so good if we assume that 6-5i the real concentrations are the specified ones. Accepting this assumption, the mean error in average concentrations found for all elements, excluding mercury, was 11 % with the maximum error for gold (18 %). Some authors found for similar standards that batch-to-batch variations limit the accuracy to about 10 %, and precision is limited by heterogeneity to 4 to 6 %/6,9,10/. Since we found about three times lower concentration (3.0 ppm) of mercury than the specified one (9.73 ppm), we repeated the analysis by radiochemical neutron activation analysis / 11 /. The result obtained was the same as that determined by the kQ method, and it is evident that some systematic error has ben introduced during the preparation of the resin standard. Considering all the above outlined facts we can conclude that the kQ method recently introduced in our laboratory proved reliable and promising enough for further studies and routine analyses after additional checking of the method with standard reference materials. Acknowledgements The financial support of the Communities for Science of Slovenia and Montenegro (Yugoslavia) and of the National Bureau of Standards (USA) is highly appreciated. 6-52 REFERENCES 1. A. SIMONITS, L. MOENS, F. De CORTE, A. De WISPELAERE, A. ELEK, J. HOSTE, J. Radioanal.Chem., 60 (1980) 461. 2. L. MOENS, F.De CORTE, A. De WISPELAERE, J. HOSTE, A. SIMONITS, A. ELEK, E. SZABO, J. Radioanal. Nucl. Chenu, 82 (1984) 385. 3. F. De CORTE, Habil. Thesis, Univ. Gent (1987). 4. L. MOENS , J. De DONDER , L. XILEI , F. De CORTE , A. De WISPELAERE, A. SIMONITS, J. HOSTE, Nucl. Instr. Methods 187 (1981) 451. 5. F. De CORTE, K. S0RD0 - EL HAMMAMI, L. MOENS, A. SIMONITS, A. De WISPELAERE, J. HOSTE, J. Radioanal. Chem. 62 (1981) 209. 6. L. M. MOSULISHVILI, M. A. KOLOMITSEV, V. Yu. DUNDUA, N. I. SHONIA, 0. A. DANILOVA, J. Radioanal. Chem. 26 (1975) 175. 7. J .OP De BEECK, Internai Report INW-Gent (1976). 8. LIN XILEI, Ph. D. Thesis, University Gent, 1981. 9. K. HEYDORN, Neutron Activation Analysis for Clinical Trace Element Research Vol. I, p. 79, CRC Press, 1984 10. J. C. ROUCHAUD, L. DEBOVE, M. FEDOROFF, L. M. MOSULISHVILI, V. Yu. DUNDUA, N. E. KHARABADZE, N. I. SHONIA, E. Yu. EFREMOVA, N. V. CHIKHLADZE, J. Radioanal. Nucl. Chem. 113 No.l (1987) 209. 11. A. R. BYRNE, L. KOSTA, Talanta 21 (1974) 1083. 6-53 DETECTOR OR 1 16.845 cm a % 10-3 (50-3001 (300 - 550) (>650| 2 log£pS-il.4S52»49.7!97(!agcf|-2a.S73UlogerJ 'Z-S597(log=iP logSp *-0.SS2S-0.39*C(leg=j-t log£p .- 0.3432-0.3730 og£r) 10" . [ } r } i > > ) ; ' ' LiJ L_i 10 100 1000 energy (keV) Fig.l. Experimental reference efficiency curve for the detector 0R1. Er iksv Fig.2. Experimental P/T = f (E^) curves for the detector 0R1. 6-54 Table 1. Nuclear data used for the analysed isotopes /3/ Energy,keY k Ê ,eV Element Isotope T Q0 I/2 o r 110m, 2 Ag Ag 249.76 d 557. a 3.44-lû" 384.7 2.65-10" 17.5 6.08 937.5 1.25-lu" 198, Au Au 2.595 d 411.S 1.00 15.71 5.65 3a 8a 11.8 d 373.2 2.03-lo'j 24.8 69.9 496.3 6.84-10" 8 3r Z„8 r 35.30 h 554.3 2.38 -lu"2 776.5 2.76-10" 19.3 152 1044.0 9.14-10" 141. Ce Ce 32.501 d 145.4 3.66-10"3 0.33 7200 Co 50Co 5.271 y 1173.2 1.32 1.993 136 1332.5 1.32 Cr SiCr 27.69 d 320.1 2.62-10* 0.S3 7530 Cs Cs 2.082 y 569.3 7.34-10* 11.8 9.27 804.7 4.76-10" Fe 59Fe 44.53 d 1099.2 7.77-10"g 0.97 637 1291.5 5.93-lfl" Hg 203Hg 46.612 d 279.2 î.io-io" 0.88 1960 140 La La 40.22 h 815.8 3.32-lû"2 1.24 76.0 1596.S 1.34-10" 24 Na Na 14.959 h 1368.6 4.68-10" 0.59 3380 Rb 36Rb 18.66 d 1076.6 7.55-10"4 14.8 839 Sb U2Sb 2.70 d 564.1 4.38-lfl" 33.0 13.1 692.8 2.38-10" Sc 48Sc 33.82 d 889.3 1.22 0.43 S130 1120.S 1.22 Se 75Se 119.77 d 264.7 7.2S-lo"^ 10.0 29.4 400.7 1.45-lû" U3m, Sn In 1.658 h 391.7 5.99-lu"5 48.4 107 16 Tb °Tb 72.1 d 298.8 8.25-lû"2 17.9 13.1 879.4 9.24-10" In Zn 244.0 d 1115.5 5.72 *lo" 1.908 2S50 6-55 Table 2. Concentrations of elements in synthetic standard No. 0. Results are calculated from 5 independent measurements. Element Concentration (ppm) This work Specified /6/ Silver 3.39 ± 0.07 2.92 Gold (8.82 ± 1.60).10"3 9.73-10"3 Barium 462 ± 12 486 Bromine 21.9 + 0.2 19.6 Cerium 0.437 ± 0.030 0.490 Cobalt 3.37 + 0.04 2.92 Chromium 11.0 ± 0.2 9.73 Cesium 0.942 ± 0.020 0.973 Iron 345 ± 12 292 Mercury 3.0 + 0.8 9.73 Lanthanum 0.322 ± 0.050 0.292 Sodium 3360 ± 10 2920 Rubidium 29.1 + 0.9 29.2 Antimony 0.292 ± 0.030 0.293 Scandium 0.114 ± 0.001 0.0977 Selenium 5.58 ± 0.04 4.87 Tin 56.0 ± 3.0 48.4 Terbium 0.509 ± 0.009 0.487 Zinc 109 ± 1 100 6-56 TT 1187. T T S I 1007. 827. 31*8 a Ag Au Ba 8r Ca Co Cr Cs Fe Hg La Na Rb Sb Sc Se Sn Tb Zn Fig.3. Comparison of results for the synthetic standard No. 0, showing the ratios of results from this work to the specified ones taken as 100 ». The bars indicate the standard deviation of the present measurements; the uncertainty of the specified values is estimated to be 18% /6,9,10/ 6-57 NEUTRON FLUX CHARACTERISATION OF THE TRIGA MARK II REACTOR, LJUBLJANA, YUGOSLAVIA, FOR USE IN NAA * ** ** * ** S? Jovanovid , B. Smodis" , R. JaCimoviC , P. Vukotie , P. Stegnar * Institute for Mathematics and Physics, "V. Vlahovic" University, Cetinjski put bb, Yu-81000 Titograd, Yugoslavia ** "JoZef Stefan" Institute, "E. Kardelj" University, Jamova 39, Yu-61000 Ljubljana, Yugoslavia Abstract The following neutron flux characteristics are relevant for use in single comparator methods (e.g. kQ) of activation analysis with reactor neutrons (NAA) : (1) the thermal-to-epithermal flux ratio ( f = 0.,/ 0 ), ll/v (2) the parameter a in the 1/E epithermal spectrum representation and, (3) flux variability during irradiation. These characteristics were determined in a few typical irradiation positions of the TRIGA Mark II reactor, Ljubljana, Yugoslavia, as part of a programme to introduce the kQ-method of NAA at the "J. Stefan" Institute. 6-59 INTRODUCTION Since its introduction in the mid sixties /1,2/, the use of single comparator methods in activation analysis with reactor neutrons (NAA) has been steadily increasing. Among many variations, generally restricted to a particular laboratory (i.e. irradiation and counting facilities/conditions), the kQ-method /3,4/ turned out to be the most flexible and is nowadays applied or being introduced by some twenty institutes in various countries. Our experiences with the kQ-method at the "J. Stefan" Institute are presented at this Conference in a separate paper /5/. So as to apply a single comparator method, the characteristics of the counting equipment should be known (e.g. detector efficiency curve for the given source composition and source-detector geometry), as well as the neutron flux at the irradiation site. As to the latter, the main concerns are: -thermal-to-epithermal flux ratio - epithermal spectrum shape and - flux variability during irradiation. These characteristics were determined in three typical irradiation sites of our TRIGA Mark II reactor, namely the central channel (CC), the pneumatic transfer tube channel (PT) and carousel-position 25 (CR-25), which are indicated in Fig. 1. 6-60 EPITHERMAL NEUTRON FLUX CHARACTERISATION It can be shown that the generally accepted " idealised " epithermal neutron flux distribution ^'e(E)~l/E (1) (E = neutron energy) is often not satisfactory from the standpoint of accuracy in single comparator methods of NAA /6/. Instead, a semi-empirical representation /7,8/ ^(E)-l/E1+a (2) as a better approximation, proved to be satisfactory for NAA needs /9/. Parameter a characterises a given position in a reactor and represents a measure of the epithermal flux deviation from the ideal (1/E) distribution. In spite of some attempts to describe a theoretically as a function of the reactor, experimental determination remains the only reliable way of obtaining an a-value. In this work we used the " cadmium-ratio for multimonitor" method, which yields the most accurate results / 10 /. The following a-values were found: in CC: a = -0.051 ± 0.008 (monitor set: 197Au-59Co-68Zn-64Zn-94Zr) in PT: a = -0.048 ± 0.005 (monitor set: 197Au-124Sn-112Sn-59Co-122Sn-55Mn-64Zn-94Zr) in CR-25: a- -0.009 + 0.004 (monitor set: 197Au-124Sn-112Sn-59Co-98Mo-122Sn-68Zn-100Mo-64Zn-94Zr) 6-61 More details about a-determination at our TRIGA reactor can be found in Ref. /11/. It should be noted that for an actual analysis, the simpler "bare" method, with only a Au-Zr monitor set, can also be used with satisfactory results. THERMAL-TO-EPITHERMAL FLUX RATIO Determination of the thermal-to-epithermal flux ratio (f = $tn/ 4>e) is, in principle, a part of an actual analysis, by means of the " bare 197 94 96 triple monitor " method ( monitors Au- Zr- Zr ) /12/. We determined 197 f using the "cadmium ratio" method /12/, with Au monitor ( in the form of 0.1% Au-Al wire): in CC : f - 20.4 + 0.8 in PT : f = 19.4 ± 0.7 in CR-25 : f = 19.6 ± 0.8 The above shows that f can be assumed to be nearly constant in various irradiation sites of our reactor. Such a situation is understandable for a poorly thermalised reactor with rather highly enriched ( 20% and 70% 235U, Fig. 1 ) fuel elements. 6-62 NEUTRON FLUX VARIABILITY Neutron flux behaviour during irradiation should be known as well when applying the kQ-method /13/. From the operators' logbook record, the variations of reactor power during operation can be followed, as is illustrated for a 100-hour operating period in Fig. 2. Our measurements of the variations of thermal ( $+u)> epithermal ( Let us remark that a constant neutron flux during irradiation, which is preferable for the application of the kQ-method, can be obtained at our reactor over periods of up to 10-20 hours, on request to the operators. Such periods are indicated in Fig. 2. Acknowledgements The financial support of the National Bureau of Standards (Washington, USA) and of the Research Communities of Slovenia and of Montenegro (Yugoslavia) is highly appreciated. 6-63 REFERENCES /l/ F. GIRARDI, G. GUZZI, J. PAULY, Anal. Chem., 36/8 (1964)1588 IU F. GIRARDI, G. GUZZI, J. PAULY, Anal. Chem., 37/9 (1965)1085 /3/ A. SIMONITS, L. MOENS, F.DE CORTE, A. DE WISPELAERE, A. ELEK, J. HOSTE, J. Radioanal. Chem., 60(1980)461 /4/ L. MOENS, F.DE CORTE, A. DE WISPELAERE, J. HOSTE, A. SIMONITS, A. ELEK, E. SZA80, J. Radioanal. Nucl. Chem., Articles, 82(1984)385 /5/ B. SMODIS, S. JOVANOVIC, R. JACIMOVIC, P. VUKOTIC, P. STEGNAR, "The Use of the TRIGA Mark II Reactor, Ljubljana/Yugoslavia, in the kg-method of NAA", paper presented at the 10 Europen TRIGA Users Conference, ' Vienna/Austria, Sept. 14-16, 1988 /6/ F. DE CORTE, L. MOENS, A. SIMONITS, K. SORDO - EL HAMMAMI, A. DE WISPELAERE, J. HOSTE, J. Radioanal. Chem., 72(1982)275 /If P. SCHUMAN, D. ALBERT, Kernenangie, 2(1965)88 /8/ T. B. RYVES, Metrologia, 5(1969)119 /9/ F. DE CORTE, L. MOENS, S. JOVANOVIC, A. SIMONITS, A. DE WISPEAERE, J. Radioanal. Chem., 102(1986)37 /10/ F. DE CORTE, K. SORDO - EL HAMMAMI, L. MOENS, A. DE WISPELAERE, A. SIMONITS, J. HOSTE, J. Radioanal. Chem., 62(1981)209 /ll/ S. JOVANOVIC, P. VUKOTIC, 8. SMODIS, R. JACIMOVIC, N. MIHALJEVIC, P. STEGNAR, J. Radioanal. Nucl. Chem., Articles, in press /12/ L. MOENS, Ph. D. Thesis, Univ. Gent/Belgium, 1981 /13/ F. DE CORTE, Habil. Thesis, Univ. Gent/Belgium, 1987. 6-64 TRIGA MARK II 250 kW, LIGHT-WATER MODERATED REACTOR CAROUSEL ORIVE AXIS "J.Stefan" Institute Ljubljana .Yugoslavia CAROUSEL (CC) CENTRAL CHANNEL IC ) IRRADIATION CHANNELS © SAFETY ROD FT) PNEUMATIC TRANSPORTEES CH. © SHIM ROD (~J FUEL ELEMENTS 20«/. 23S.J © REGULATION ROD FUEL ELEMENTS 70V. 235|J Fig.l. Ground plan of the TRIGA MARK II reactor 6-65 5.=r ü 7.2- 7.5r, 1 ! ö II < I [ i 41 i/r, 4.3.- L i< ->=?- I f. . , 22 r r I ; -, • - " ï • • 1 r i h• • , ; 1 lb .. . t t ' .. ,. Ï, „, ' , 1 ( 1 t lOSr 5 101 PC? I PCF 5 q-s SSh iO 20 30 CO 50 60 70 80 90 100 no lime (hours) Fig. 2. Neutron flux variability at the TRIGA Mark II reactor, Ljubljana, during a 100 - hour operating period; $ tot * th and *epi scales arbitrary units; PCF - period of constant flux 6-66 SESSION VII Reactor Physics NONLINEAR DYNAMICS OP ITU TRIGA REACTOR^ H.A. Hizal^2\ g. Gençay^3\ 0. Ciftgiaglu 4', B. Can^5' E. Güngördü^, M. Geçkinlj/3) ' ^6) ABSTRACT Complete dynamics af a reactor could be developed starting from the very basic principles. However such a detailed approach is often not worth the effort for a rather simple pool type reactor which may be subjected to various power excursion maneuvers without challenging its safety system. Therefore a coupled point kinetics-lumped thermal hydraulics model is talcen up as the basis of the system model. Response of the reactor to ramp'insertion of reactiviy"is " observed by sampling the power channel, water, and fuel temperatures with the help of a PC. One of the important model parameters, fuel temperature feedback effect is studied during power excursions and the results are compared with those of static tests. INTRODUCTION Dynamics of systems are conveniently modelled in terms of ordinary differential equations of the form: '.York supported by Electronics and Control Systems Research Center, Istanbul Technical University. 2) Mechanical Engineering Department, Division of Machinery and Dynamics, ITU. institute for Nuclear Energy, ITU, Maslak, Istanbul. 4vFaculty of Electrical and Electronics Engineering, Division of Nuclear Power - ITU. 5) .«acuity of Technical Education, division of Electronics, Marmara University, Istanbul. °'TQ whom all carresnondcnee should be addressed. 7-1 g^» y. yi p) = o (i) where t is the independent variable; y_ represents a vector of state variables, some components of which may not be directly observable, and vector p is a set of unknown system parameters. This problem can be tackled in various ways depending on the nature of the system of equations, /l/. If all of the state variables are readily observable, then one may calculate jr by finite differencing y and the parameter estimation problem becomes one of algebraic equaTion solution. Even in a case as simple as this, one runs into a set of redundant equations from which an optimal solution must be extracted. Another approach of the nest higher complexity is to define an objective function which is a certain, weighted measure of the discrepancy or mismatch between what was observed and what is calculated by the numerical integration of (1) with some judicious initial estimation of p_; then a search is conducted to find the minimum of the objective function by varying p, according to a scheme such as the simplex method, /2, 3/. This presentation is a short account of the experimental part of the preliminary work.related to a project. Steady state parameters of the TRIJA core operating at rated power are well documanted, and at other power levels they can also be determined easily, yet a question arises as to the use of those parameters in a fully fledged nonlinear model. Therefore our task is to determine the margins of error involved in using steady state parameters in place of dynamic parameters. This point is crucial because some of the unknown parameters may be eliminated at the outset, and secondly a wise initial guess saves both computer time and avoids the minimization scheme from running astray into the other local minima. 7-2 GOVSRITIICg 3ÇUATI0IIS In the following table the model equations are summarized. Table 1 Model Description Typo of the Dependent Description of the Unknown model phenomenon variables rate equations parameters 1. Heutronic - Power - Point kinetics - Temperature equations feedback of reactivity - Delayed neutron precursor concentrations 2. Thermal - ITuel temperature Lumped energy Heat balance for a transfer fuel element coefficients coupling fuel elements to the coolant Coolarit Lumped energy temperature at balance equation the channel for convected outlet or energy for the average coolant entire volume of a temperature coolant channel vvhich constitutes an integral control volume 3. Hydraulic Coolant Lumped momentum Srictional velocity balance for the loss buoyancy driven factors for coolant the entire channe1 7-3 NUI.3RICAL SOLUTION The presence of point kinetics equations makes the above set of coupled equations extremely stiff. The ratio of the largest tine constant to the smallest is a measure of the stiffness, and this ratio, in the case of point kinetics is around (A . £/ß) ~ 10^ For the integration of such stiff system of equations we need numerical schemes which are accurate for short time steps and stable for larger steps. We are currently comparing Gear's popular algorithm /4/ to Hansen's prescription /5/ where the latter seems to be more promising on account of computation time. Reactor power acts as a source term driving the thermal hydraulics equations. In order to allow for large time steps, h, for integration, the solution of point kinetics equations both accurate for time steps much smaller than h and stable (or asymptotic) for time steps of size h is taken up. Therefore the said prescription constitutes a typical example of splitting methods utilized in the integration of certain kinds of stiff problems, /6/. Experiments are conducted to see how the steady state loss of reactivity with an increase in fuel temperature compares with the case when the system is undergoing a change of state through external reactivity disturbances. Various dynamic states have been attained by near-ramp reactivity insertion lovelling off at rod positions corresponding to rod worths of 30, 50 and 70 0 (Run 1, 2,and 3) respectively through the manipulation of the regulating rod. The observable state variables are reactor power, fuel temperature and bulk temperature, and a relevant observable parameter is the regulating rod position. Signals from the reactor console have been sampled and stored in the memory through a run of tho mill 12-bit ADDA card mounted on a PC board, ',/ith the QBASIC-driver the sampling and memory access time per channel is loss than 10 ms. An assembler-driver allows for a sampling rate as high as 10 .kHz for a 3ingle channel operation, in which case the real time use of the system is realizable. As for the output signals, wide-range log channel (LxlO~^to 150 fa power) preamplifier DC return is amplified with an HP 410 G ammeter before being sampled. Output of thermocouple amplifier, whose junction is located in a B-ring instrumented fuel element, i3 fed into the AD converter. 7-4 Regulating rod position is monitored from the output of the helical potentiometer already attached to the drive mechanism of the rod. EVALUATION OJ? THE RESULTS An important parameter of the reactor core is the net reactivity, which ia measurable instantly with the help of inverse kinetic equation. This equation is exact as it stands regardless of the feedback effects. In calculating the reactivity of the core from the inverse kinetic equation, a Gaussian quadrature type Laguerre integration formula has been used for the convolution integral. The usual practice is to use a recursive formula which updates the convolution integral at every time step, for example see Ref.. /7/. A shortcoming of the latter is that a temporary noise surge vail contaminate the calculation from than on. On. the other hand Laguerre integration formula requires the power history for a certain duration back in time, which may be found disadvantageous. ?or this purpose sampled values of .power are stored in a revolving stack memory,namely old data is discarded as the new data fl.;ws in. In the inverse kinetics equation the logarithmic derivate term corresponding to the inverse of the dynamic period is neglected; since, when reactor period, is around 3 sees, which is the minimum allowed value, the contribution of this term is of the order of 1 0. The not reactivity is the sum of external and feedback reactivities, feedback reactivity calculated this way is attributed solely to the rise in the temperature of the fuel elements. In i,lig. 1. data from one of the runs (Run 3.) is displayed. Ho corrections have been made for the time constant of the thermocouple. In lrig. 2 feedback reactivity - fuel temperature data from 3 different runs is superimposed. The small deviation between these two curves with a certain trond suggests that feedback reactivity ( Pfb ) also be a function of the rate of change of fuel temperature ( ÂTp ), and with the least squares we have estimated the following relation. pfb (ATF, ÂTF) = -(1.72 + 0.673 • ATp * 0.249-ATf), ( (2) The first term is systematic ejaror which stems from tho calibration. The last term is the net contribution of the deviation from the steady st."to. 7-5 DISCUSSIOn AID CONCLUSIONS A joint project has been launched to extensivly study the nonlinear dynamics of ITU TMGA Reactor. Our short term objectives include nonlinear parameter estimation, whereas , in the long run, similar to öET experience /B/f a closed loop digital controller will be developed and tested on the nonlinear model of the reactor. Furthermore a linear model for small power fluctuations around a set point can be directly deduced from the developed model. In_ this work the reactivity feedback due to fuel temperature changes is observed to deviate by a small amount from what was reported earlier as static test results, /9/. ACIC:0'7Ii:DG..-ÜSI"i!S V/e are grateful to the director of the Institute, Professor 3. Yasgan for his constant support and encouragement. ïïe also express our thanks to the x>eactor staff, especially to our colleagues Assoc. Professors i-I. Yavuz and A. Bayülken, and Assist. Pcof. S. Sdgü for their help in conducting our experiments. 7-6 I13EEÏ2::C33 /!/. Y. Bard, Honlinear Parameter Estimation Academic Press (1974) 12/. F.I. Ibitoye, et al. "Modelling and Parametric Analysis of an Advanced Gas-Cooled Reactor" Ann.Hue10Energy 14, 581 (1987). /3/. \7.H. Press, et al., numerical Recipes Cambridge U.Press (1936). /4/. C.7. Gear, Numerical Initial Value Problems in ODE" Prentice-Hall (1971). /5/. R.Jo One2a, and K.3. Karcher "Nonlinear Dynamics of a Pressurized •Vater Reactor Core". Hucl.Sci.2ng. 61,, 276 (1976). /6/. J.H. Persiger "Numerical Methods for Engineering Application" John Wiley (1981). 111. A.7. Grachev, et al., "A Digital Reactimeter for Nuclear Reactors", Atomnaya Energiya j5l, 110 (1986) . (Translation in Soviet Atomic Energy) . /3/. J.A. Bernard, et al., "The Design and Experimental Evaluation of Closed-Loop Digital Controller Based on an Alternate Formulation of the Dynamic Period Equation" Proceedings of the Topical Meeting on Reactor Physics and Safety, p.610, Saratoga Springs, K.Y. (1986). /9/. M. Ok!:a, Nonlinear Dynamic Llodel of Triga Marie II Reactor, US Thesis ITU (Jun. 1985). 7-7 POWER, kW REACTIVITY, Ç TEMPERATURE ATp, °C 200 100. 180. 16C 140. . T20. > ICO _ 40 . 20. o Power * Rise in fuel temperature v External reactivity inset • Feedback reactivity J I i 0.G 20 30. 40. 50. 60. TIME,s .''i-'-;. 1. Results from Run 3. 7-8 FEEDBACK REACTIVITY, Ç • • Run 3 • • 60.0 O Run 2 • o 40.0 o • o O AA Run 1 • O A O A 20.0 O A O • A 0.0 A 0.0 20.0 40.0 60.0 80.0 100.0 ATp.C Pig. 2 Feedback reactivity vs. fuel temperature rise for three different runs. REACTOR SURVEILLANCE BY NOISE ANALYSIS Ozer Ciftcioglu Istanbul Technical University, Electrical Engineering Faculty 80191 Teknik Universite-Gumussuyu, Istanbul, Turkey ABSTRACT A real-time noise analysis system is designed for the TRIGA reactor at Istanbul Technical University. By means of the noise techniques, reactor surveillance is performed together with failure diagnosis. The fast data processing is carried out by FFT in real-time so that malfunction or non-stationary operation of the reactor in long term can be identified by comparing the noise power spectra with the cor responding reference patterns while the decision making procedure is accomplished by the method of hypothesis testing. The system being computer based safety instrumentation involves CAMAC in conjunction with the RT-11 (PDP-11) single user dedicated environment. 1. INTRODUCTION As a research reactor of relatively low power, TRIGA provides cer tain advantages over other type of nuclear reactors such as convenience for basic studies of the application of noise techniques for reactor analysis, surveillance and mulfunction detection together with diagnosis. Hence, it is the aim of this work, as an elaborated application of noise techniques in reactor surveillance, to describe a real-time noise analysis system designed for the TRIGA reactor at Nuclear Energy Institute of Istanbul Technical University (ITU). The system makes use of the existing modular computer based instrumentation partly used for reactor analysis and safe reactor operation studies as reported before |l-3|. The work, therefore, at the same time indicates the flexible configuration and diverse utilization properties of modular instrumentation in instrumental analysis. 7-11 2. SURVEILLANCE METHOD One of the outstanding feature of noise analysis is the possibility of assessing the correct operation of the plant under surveillance and providing early warning of incipient malfunctions or expected and unexpected anomalies. This task is carried out by studying the fluctuations of signals coming from different sensors which might be neutronic, temperature, mechanical, acoustic etc. The process noise signals respond sensitively to small changes in plant characteris tics and they can be used effectively to detect an incipient abnormality with diagnostic information. For this aim the methods can roughly be classified into two categories, namely, model based approach and data-based approach. In model based approach a plant model is concerned. In data-based approach a pattern recognition technique is used. Concerning the identification of the abnormality cause, pattern recognition is to be preferred rather than the model- based approach which is favorite method for quantitative evaluations. The approach involved in this work is based on the noise pattern recognition and the power spectral density is monitored to determine if the plant state has changed significantly from normal state. As a basic concept of the surveillance, the relation between the noise patterns and the plant operating state being known, the present plant state is identified by comparing the newly obtained simple noise patterns in real-time. In the comparison procedure, the reference noise power spectrum and the relevant confidence limits corresponding to normal operation are determined. For the determina tion of the confidence limits, we define a logarithmic probability density function (pdf) of the form Above, y is logarithmic gaussian random variable given by Y = In | (2) r where x is the computed spectrum; x the reference spectrum counterpart. The variance a is identified during the learning period and it is found to be independent of the frequency. Concerning the detection of anomaly in a stochastic signal, some times noise is mistaken for anomaly when anomaly is absent and sometimes anomaly is not identified when it is in noise present. 7-12 These two separate cases can be distinguished by naming false alarm (FA) and alarm failure (AF). The associated error probabilities are denoted as p„. i.e. false alarm probability (FAP) and p._ i.e. alarm failure probability (AFP). Any anomaly detection process would be one which maximizes the probability of detecting the anomaly when it appears and minimizes the probability of mistaking noise for anomaly when it is not present. Hence, the process essentially implies the minimization of the penalty function defined by the sum of FAP and AFP. In a nuclear reactor operation FA and AF are two fundamental con cepts related to deviations from normal operation. The deviations are referred to as anomaly which can occur in variety of ways. In this work, the emphasis is put on the anomaly detection rather than the minimization of false alarm from the view point of safe reactor operation that it consequently implies a minimization for alarm failure. In a nuclear reactor there are a number of inherent noise sources which can be used in variety of ways for malfunction detec tion since these noise sources contain essential information about the operating state of the reactor. In particular, use can be made of the change of noise statistics which are termed as 'noise signa tures'. To this end, a gaussian anomaly detection processor is devised where anomaly acts on the reactor power power spectral den sity (psd) fluctuations which are verified to be gaussian the result being in accord with the well-known central limit theorem in statistics. The anomaly is assumed to be gaussian acting on the gaussian psd fluctuations for the sake of simplicity as the validity of the results 'remain essentially the same in practical applications. In the detector processor, a detection level is defined so that any psd amplitude exceeding this level is considered to be anomaly. For the establishment of the appropriate detection level £ method of hypothesis testing, in the statistical terminology, is used where the null hypothesis H refers to anomaly is not present, against the alternative hypothesis H. indicating the presence of anomaly. The corresponding variances Being a and c. respectively where o*> a the corresponding probability densxty functions (pdf) are given by ° y /2 a f<*il°> = Tab1 " e ~ i o (3) v o 2/o 2 1 ~yi /2 °l f«y±l1> = 72^ e <*> 7-13 With respect to the hypotheses stated, in the statistical terminol ogy, type I and II error probabilities are 2 type I error (FAP) • 1 - / f(y|0) dx (5) -2 2 type II error {AFP) = / f(y|l) dx (6) -2 For the optimum detection level 2 , we can define a penalty function as the sum of the type I and II error probabilities and look for the minimum of this function with respect to £. Accordingly, 2 is found to be \k\ ° In (a1 /o ) ,2 2 • i - 'o" = *l a2 (7) ^2 -1 a o 2 2 For weak anomalies, i.e. o\ a ,2 a is obtained. l — o o - o Because of discrete form of data used and from the view point of practical implementation of the processor in real-time, the process ing of N samples by the gaussian processor is carried out in a special way as follows. We define a function f(y.) and an associated random variable z as i f(y±) =0 if y± £ 2 (8) f(y£) =1 if y± > 2 and N z = I f(y,) (9) i=l where N is the number of samples used in each processing period in the gaussian processor. Any count exceeding the critical level 2 implies anomaly. Since the probability of exactly z samples exceed ing 2 in absence of anomaly, is given by binomial distribution, the type I error probability (FAP) for N samples is given by N Z N 2 N (FAP)N = I (I) po (1-P0) " = 1- (1~P0) (10) 2 = 1 7-14 where p is given by 0» p = 2 / f(y|0) (11) £ For 2 = £ -cr i.e. 2 is optimum o o FAP = p = 0.3174 (12) for one sample and N (FAP)N = 1- (1 - 0.317^) (13) for N samples. For N = 100, (FAP).. - 100# which indicates large FAP unacceptable in practical applications. To reduce FAP for the price of increase in AFP, we can choose p smaller. However even for p = 0.001, FAP is still too high for applications in nuclear reactor. The inconvenience described above can be improved not reporting every excessive amplitude as anomaly and certain number of count of z can be prescribed in advance before a decision for anomaly takes place. This number is termed as 'decision level'. The null hypotheses is based on the non-existence of anomaly and the 'critical region' is defined as the region where the number of sampled signal amplitudes exceeding the critical level £ exceeds the decision level denoted by K. The FAP for this case is computed by FAP = I (2) p * (1-PJN"Z (14) z=K+l ° or K FAP = 1- 1 a) po* (l-p r (i5) z=0 so that the FAP can now be designed in any range desired. Implementation of the processor can be carried out as follows. For a given AFP, the number of data samples (N) and the critical level £, the decision level K is computed from Eq.15 where p is related to 2 by Eq.ll. ° 7-15 If the processor determines that the random variable z in Eq.9 satisfies the inequality z>K (16) then the anomaly is declared. Since the statistics of z is described by binomial distribution, for N>>1 and p <<1 and Np =u = z where z indicates the average, the binomial distribution can be replaced by poisson distribution so that Eq.l4 is replaced by k v k » u u K u u FAP = I e-"0-^ =1-1 e^°-£ k=K+l K- k=0 The relative value of K is defined as 'level of significance' (LOS) and it is given by LOS =| (17) According to the hypotheses adopted in the decision making process, the associated type I and type II error probabilities are K type I error (FAP) = 1- I e-» °o -ry C- (18) k=0 v k type II error (AFP) = £ e -rf- (19) k=o R- As in the previous case, the sum of type I and type II error prob abilities is adopted as penalty function subject to minimization so that the optimal value for the decision level K can be searched |4|. In the present surveillance system, p is selected to be 0.0075. which gives, from Eq.ll, the critical level as 2 = 2.43 ay (20) Hence, the corresponding confidence limits on the power spectra are computed from x In - = * 2.43 a (21) x r y where x is the noise power spectrum corresponding to the confidence limits computed. In this implementation LOS is considered to be 0.04. The relevant FAP is computed from Eq.15 where N indicates the number of the frequency points involved in the frequency range under consideration that it yields FAP = 0.014 for N = 128. 7-16 3. COMPUTER BASED INSTRUMENTATION The computer based noise analysis system developed comprises two essential parts. The first part includes both hardware and software components to obtain the analog signals from the reactor and trans fer the data to the computer for processing. The second part is the software component of the system for the computation of fast auto/cross power spectral density (psd) spectra in real-time together with the verification of inequality 16 and display of the computed spectrum. The hardware part of the system includes a PDP-11/05 computer which is used for both controlling the CAMAC system and executing the real-time computations. On the other hand, CAMAC is used for both data acquisition and visual display. To eliminate the detector and electronics background noise effects, two detectors were considered in the noise analysis system. After the dc components of the analog signals obtained from the detectors have been removed and noise signals have been amplified, the analog data are transferred to CAMAC system in digital form. The data transfers to CAMAC are accomplished by means of a two-channel CAMAC ADC (analog-to digital converter) module driven by the standard |5| CAMAC subroutines developed in our laboratory. The detailed descrip tion of the subroutines were presented before |6,7|. In order to be able to perform the auto/cross correlation computa tions in real-time the software part of the system is divided into two distinct parts since FORTRAN is not fast enough for this purpose. The first part of the program is coded in FORTRAN and used to introduce the operational parameters involved, into the program in the beginning of the run. The parameters include the number of points (N) used in FFT computations, switch for auto/cross psd com putation, sampling period (At), number of power spectra (n) used for averaging. The second part performs the fast data processing by means of assembly-written FFT algorithm developed in our laboratory and combined with the assembly-written CAMAC data acquisition and visual display driver routines. The average noise power spectrum is computed from the spectra obtained from last n psd computations and the spectrum thus obtained is displayed for continuous visual inspection. Decision making process for anomaly detection is ex ecuted in accordance with Eq.9- The spectrum having been displayed, the execution of the program returns back to the beginning of the second part of the program to carry out the same computations once again and hence constitutes a loop. During the continuous run of the program, parameter change such as N, At, n, lin/log display switch can be performed by means of CAMAC interrupt routine included in the software. 7-17 4. SIMULATION STUDIES FOR REACTOR SURVEILLANCE To test the noise analysis system described and verify its perfor mance, use was made of the data obtained from ITU (Istanbul Technical University) TRIGA reactor at Institute for Nuclear Energy. For anomaly detection studies, data were obtained in particular, from reactor simulator including the modelling of the reactor stochastics. The main feature of the present noise analysis system is its flexibility for determining operational conditions of the noise analysis system and processing the data, in wide range by software while the hardware involved imposes the ultimate limitations. The detectors used are two identical Reuter stokes made Cobalt type self-powered neutron detector (SPND)s connected to ECN (Energieonderzoek Centrum Nederland, the Netherlands) made two- channel low-noise current amplifier (601A-C) and deviation amplifier (601B) both being expecially designed for reactor noise experiments. The CAMAC ADC involved is two channel Kinetic Systems 3512 module. The analog signals obtained from the detectors are recorded on a 4- channel HP-3960 FM low-noise and wide frequency range tape recorder which allows to carry out the experiments in various playback speeds so that the process under investigation can effectively be speeded up or slowed down in case the speed of playback is different than that used for recording. This possibility is especially useful for investigations in low frequency range and in testing the anomaly detection processor in different frequency regions under the identi cal operating state of the reactor. The CAMAC driven display is a plasma dot-matrix with resolution of 512 x 512 on a standard visual display screen. The driver comprises 4 ORTEC made CAMAC modules, namely, point plotter PP009; character generator CG009; lin/log gen erator LL009 and matrix driver MD009 modules which provide efficient display of the computed spectra in real-time operation. For the investigations, reactor power was selected as process vari able from which the noise signal is obtained. For the sampling of the analog signals, use was made of the line-clock of the operating system which provides sampling period (At) of 20 ms or its integer multiples, At=20 ms being the minimum. In order to maintain the operation in real-time, no data window is applied to raw data. Every power spectrum displayed is obtained by taking the average of the present spectrum computed and n - 1 previously computed spectra, n being an input parameter for computations. Since the analog signals obtained from SPND s are rather weak in low-power operating state, the zero-power reactor analysis for test ing the noise analysis system was carried out by means of fission detector placed next to the core and coupled to a HP-410G low-noise current amplifier. The test results indicated the presence of some additional background noise contributions other than power noise as this was reported before |8|. These investigations are also of in terest for determining the reactor parameters which can be estimated by the study of the power spectrum obtained. 7-18 The experiments using SPND s were carried out at maximum reactor power level (250 KW) of the TRIGA reactor. For this case the noise analysis system was operated in cross-correlation mode so that the unwanted background noise contributions are kept to a minimum |91• Since the noise analysis system is essentially designed for real time reactor surveillance, extensive anomaly detection studies were carried out by means of reactor simulator with the simulated reactor stochastics. The simulator is used to obtain noise data comprising purposely introduced anomalies. A typical psd spectrum of the reac tor power fluctuations is shown in Fig.l where sampling period (At) is 20 ms, number of spectra (n) used for averaging is 80 and the number of frequency points (N/2) is 256. Note that 0.7 x (N/2) points are considered in the plotting. The corresponding pdf relevant to Fig.l is shown in Fig.2 where the number of points form ing the histogram is 128. NAPSO (Ht! PDF rl-OxlO3 •0*5 f •oA / • '^ . -4-3-2-10123 4xlCT Fig. 1 Fig. 2 psd spectrum of reactor power pdf of reactor power fluctuations of TRIGA reactor fluctuations of TRIGA reactor n=80, p=250 KW, At=20 ms, N/2=256 At=20 ms, sfed.dev.=6.8lE , time of data acquis!tion=6.83 min mean=1.07E~ , numb, of pdf (normal operation) points=128 (normal operation) One of the anomalies designed for testing the reactor surveillance system is vibrating control rod the dynamic of which is represented by a second-order system with damping. The natural frequency of this system is assumed to be 3 Hz and the damping factor (g) is 0.4 that it yields underdamped oscillations causing instability in an operat ing reactor. 7-19 The corresponding psd spectrum and pdf are illustrated in Figs.3 and 4 respectively where N/2 = 256, n = 80 and At = 20 ms. The duration of the anomaly (t ) is computed from cL ta = (N/2) . it . n = 256 x 0.02 x 80 = 6.827 min (15) The resonance at around f = 3 Hz can be identified as result of comparison of Fig.l and Fig.3? -\ . 10 \ POP r 1.0 x I03 10 ^\ - • 0.8 tf.e •V 10 V. / • •°\ •0 2 \ 10 1 1 -*^ L. . \ " 4 »10 10 8 10 12 14 16 18 20 HZ Fig. 3 Fig. 4 psd spectrum of reactor power pdf of reactor power fluctuations of TRIGA reactor fluctuations of TRIGA reactor n=80, p=250 KW, At=20 ms, N/2=256, At=20 ms, std.dev.=9.13E" , time of data acquisitions6.83 min, mean«l.OOE" ,numb. of pdf duration of the anomaly=6.83 min. pr,ints=128, duration of the anomaly=6.83 min (see Fig. 3) The same illustrations as Fig.3 and Fig.4, for the same anomaly, durations of which is 2.73 min is shown in Fig.5 and Fig.6 where duration of the data acquisition is still the same as that used to obtain Figs.2-5, i.e. 6.82 min. From the coparison of the relevant psd and pdf figures, the development of the anomaly is clearly ob served . For this particular type of anomaly, the malfunction can be detected by the suboptimum processor devised, after the lapse of 1 min with the anomaly. The relevant false anomaly detection probability is given by Eq.l4 since the parameter N is equal to 50, covering the low frequency range of the noise power spectra up to 5 Hz. 7,-20 NAPSO (Ht1) PDF , T I Ox 10 0 2 6 a 10 12 14 16 18 20 Hz Fig. 5 Fig. 6 psd spectrum of reactor power pdf of reactor power fluctuations of TRIGA reactor fluctuations of TRIGA reactor n=80, p=250 KW, At=20 ms, N/2=256, it=20 ms, std.dev.=7.91E~ , time of data acquisition=6.83 min, mean=9.84E~', numb, of pdf duration of the anomaly=2.73 min points=128, duration of the anomaly=2.73 min The related psd spectrum and pdf are illustrated in Fig.7 and Fig.8 where duration of the data acquisition is 3-^1 min and the duration of the anomaly is 1.19 min. The sampling period (At) is 20 ms as mentioned before. PDF 500 -4 -3 -2 2 3 4x10 0 2 4 6 8 10 12 14 16 18 20 Hz Fig. 7 Fig. 8 psd spectrum of reactor power pdf of reactor power fluctuations of TRIGA reactor fluctuations of TRIGA reactor n=40, p=250 KW, At=20 ms, N/2=256, At=20 ms, std.dev.=7.66E~ , time of data acquisition=3-4l min, mean=1.05E , numb, of pdf duration of anomaly=1.19 min ptints-128, duration of anomaly=1.19 min (see Fig. 7) 7-21 5. CONCLUDING REMARKS AND CONCLUSIONS The real-time operation of the noise analysis system is accomplished by assembly-written FFT algorithms and the anomaly detection is carried out by means of a particular algorithm developed. The exact time needed for the detection of a certain anomaly is dependent on the nature of the anomaly. As result of the performance tests, the system is found to be satisfactory for real-time reactor surveil lance and anomaly detecton applications as the system provides also continuous display for the inspection of the psd of the fluctuations of the process variable under investigations, in real-time. Presently the system is designed to be redundant monitoring device to be added to the existing reactor safety instrumentation. Since the system operates under program control, it can easily be improved for more elaborate form by means of due programming observ ing the hardware limitations as well as the software provisions required. References : u 1. 0. Ciftcioglu and M. Geçkinli, Nucl. Inst, and Meth., Vol.177, 321-326 (1980) 2. 0. Ciftcioglu, DECUS Europe Symposium, September 1-4 hamburg, FRG £1981) 3. 0. Ciftcioglu et. al. 7th European Conference of TRIGA Users, September 15-17, 1982, Istanbul, Turkey 4. 0. Ciftcioglu, Interfacultair Reactor Institute report IRI-131- 87-08, October 1987, TU Delft, The Netherlands 5. ES0NE SR 101, Subroutines for CAMAC (1978) 6. 0. Ciftcioglu, Nucl. Inst, and Meth., Nos. 1-2, 21 (1980) 7. 0. Ciftcioglu, DECUS Europe Symposium, Amsterdam, 16-19 September 1980 8. 0. Ciftcioglu, Natinonal Electronical Symposium, TU Istanbul, 16- 19 October 1984, Istanbul, Turkey (in Turkish) 9- R.E. Uhrig, "Random Noise Techniques in Nuclear Reactor Systems", The Ronald Press Company, New York (1970) 7-22 i ! I i MEASUREMENT OF NEW CORE CHARACTERITICS AT THE MUSASHI REACTOR N. Horiuchi, T. Matsumoto, T. Nozaki, 0. Aizawa and T. Sato Atomic Energy Research Laboratory, Musashi Institute of Technology Abstract Important characteristics of the new core have been measured, after we have all replaced the old fuel elements of Al-cladding with new ones of stainless-steel cladding. The old fuel elements are stored in a storage pool. An Am-Be neutron source and three neutron detectors (two fission chambers and one BFq-counter) have been arranged in the core tank with proper separations from each other before we began to load new fuel elements. The fuel loading was performed step by step from the center region of the core. The core reached critical when 66 units of fuel elements were loaded. The critical mass was increased from 58 units of old fuel elements to this value by the replacement work. We must consider an additional reactivity for the actual reactor operation because the reactivity will be reduced by the core . temperature rising. We will report our experience in operating the Musashi reactor including the following points: (1) critical approaching experiments, (2) adjustment of excess reactivity, (3) measurement of temperature coefficient, and (4) substitution reactivity of dummy-fuel and void. 7-23 MEASUREMENT OF NEW CORE CHARACTERISTICS AT THE MUSASHI REACTOR N. Horiuchi, T. Matsumoto, T. Nozaki, 0. Aizawa and T. Sato Atomic Energy Research Laboratory, Musashi Institute of Technology I Introduction Important characteristics of the new core have been measured, after we have all replaced the old fuel elements of Al-cladding with new ones of stainless-steel cladding. The old fuel elements were stored in a storage pool. In this paper we describe the results obtained from the following experiments and measurements : (1) critical approaching experiments, (2) adjustment of excess reactivity, (3) calibration of control rods, (4) estimation of substitution reactivity of graphite-dummy and void, and (5) measurement of temperature coefficient in reactivity. The exchange work of fuel elements, the storage facility of spent fuel elements and cooling water activity will be reported in other papers. II Results and Discussion 1. Critical Approaching Experiments A couple of neutron detectors were arranged in the core tank with proper separations from each other, before the new fuel elements were loaded. Fig.l shows the arrangement of the neutron detectors. An Am-Be neutron source of 2 Ci was set at the F7 position of the circular-ring of the core. A BF counter and a UIC2 chamber were installed in the central thimble, and a fission chamber (FC2) in the circular well of reflector. Four detectors (CICl, CIC2, UIC1 and FCl) surrounding the core are ordinary detectors used for operation of the reactor. The fuel loading was performed step by step from the center 7-24 region of the core. Fig.2 shows the geometrical configurations of the fuel elements and control rods loaded into the core. The step number denotes the order of procedures. The first 24 fuel elements were loaded by keeping three control rods (Safety, Shim and Regulating) at the top position of the core. Further elements were loaded by holding only the Shim rod at the bottom position to prevent the critical accident. Fig.3 shows inverse multiplication factors as a function of the number of fuel elements loaded. The values are expressed in the reciprocal of neutron counts. Note the fact that the results were normalized to the value for the case of 50 fuel elements. The critical mass of the fuel can be acquired where the curve intersects the abscissa. As shown in the figure, the core reached the critical value when 66 fuel elements were loaded. The critical mass was increased from 58 units of old fuel elements to this value by the replacement work. This increase was mainly caused by the difference of the neutron cross section between aluminum and stainless-steel of cladding materials. The total weight of U-235 loaded is 2.5 kg and the core possesses the excess reactivity of 0.18 % Ak/k at the step of 66 units. C1.C2.C3: Control Rods N.S.: Neutron Source O : Neutron Detectors Fig.l The arrangement of neutron detecto rs in critical approaching experiments. 7-25 Step 0 fuel 0 Step 1 fuel 6 control rod -fc neutron source fuel Step 2 fuel 24 Step 28 fuel 50 control rod -K neutron 3ource fuel Step 39 fuel 8 1 Step 44 fuel 86 I control rod -fc neutron source • fuel Fig.2 The position and the total number of the fuel elements loaded into the core. 7-26 1.0 - • ü • FC1 - • A BF3 c o - 6 • O FC2 • — A • — • • E 0.5 • CD O OJ L. A • > _ o c O • • O *• l i | I I I I ?«l I I 50 60 70 Number of fuel elements Fig.3 Inverse multiplication factors measured at three different positions of neutron detectors. 2. Adjustment of Excess Reactivity We need an additional reactivity for the actual reactor operation. There are some requirements for the fuel addition and reactivity adjustment. The first is the compensation for reactivity losses, i.e. (1) neutron capture by some irradiated samples, and (2) negative effect of the core temperature. The second limitation is storage capability of the fuel elements. We don't have any storage facility except a pair of the fuel storage racks set in the core tank. The racks are capable to store only 12 fuel elements. The third requirement is to increase neutron intensity at the irradiation field for the medical use. The thermal column of the 7-27 Musashi reactor was remodeled as a Boron Neutron Capture Therapy- facility. The last restriction is due to Japanese regulations; (1) the allowed excess reactivity of the Musashi reactor is less than 1.6 % Ak/k, and (2) rod worth values of Shim and Safety are more than 2 % Ak/k that of Regulating more than 0.5 % Ak/k. The fuel and the graphite dummy elements were added to initially critical core with 66 units. Table 1 gives all procedures consisting of 25 steps of fuel operations. Fig.4 shows 7 typical core configurations. The measurement of excess reactivity was executed by using the period method. A desirable value of excess reactivity, which was as close as possible to 1.6 % Ak/k (= 200 cents), was achieved at Step 10. However, other problems of fuel element storage, neutron intensity and rod worth were still unsolved. To extend the core volume, a new trial was attempted in Step 14. Both fuel elements of B2 and B6 were exchange with two graphite dummy elements. As a result, the value of excess reactivity was drastically forced down. Step 21 was carried out to raise the neutron intensity at irradiation facilities such as a terminus assembly of pneumatic transfer system and an irradiation field for treatment of brain tumor. Note the fact that a fan-shaped graphite block was inserted in the circular well of reflector, in order to replace water with graphite. The replacement of graphite block increase not only the neutron intensity , but also the reactivity by 16.7 cents. The graphite dummy elements surrounding the Regulating rod were exchanged with a fuel element at Step 24. Then, the problem of the rod worth was solved. The graphite dummy element of F27 was further replaced with a fuel element to increase the excess reactivity by 27.3 cents at the final Step 25. The excess reactivity of 181 cents was finally achieved with 73 fuel elements. 7-28 Table 1 History of excess reactivity adjustments consisting of 25 steps. Fuel elements Excess Reactivity Step Subject matter (units) (cents) 0 (Initially critical ) 66 22.5 1 Fuel(F14) 67 77.2 2 Fuel(F15) 68 118.3 3 Graphite(F4) 68 122.9 4 Graphite(F3) 68 130.8 5 Graphite(F2) 68 140.3 6 Graphite(Fl) 68 148.3 7 Graphite(F30) 68 155.5 8 Graphite(F29) 68 161.0 9 Graphite(F28) 68 165.4 10 Graphite(F26, F27) 68 175.9 11 Unload fuel(B6) 67 41.4 12 Graphite(B6) 67 80.0 13 Fuel(F17) 68 117.7 14 Unload fuel(B2)s GraiDhite(B2 ) 67 7.7 15 Fuel(F18, F19) 69 68.7 16 Unload fuel(B4), Graphite(B4), Fuel(F20) 69 Sub critical 17 Fuel(F21) 70 20.3 18 Fuel(F22) 71 52.3 19 Fuel(F23, F25) 73 119.0 20 Unload fuel(F13, F14, F23), Graphite(F13, F14, F16), Fuel(F24) 71 45.4 21 Fan-shaped graphite 1Dloc k 71 62.1 22 Unload graphite(B4), Fuel(B4) 72 165.1 23 Unload fuel(F25), Graphite(F25), Unload graphite(Fl), Fuel(Fl) 72 174.2 24 Unload fuel(F5, F12) 5 Graphite(F5, F12), Unload graphite(F25, F26), Fuel(F25, F26) 72 153.7 25 Unload graphite(F27) , Fuel(F27) 73 181.0 7-29 Step 1 Step 6 62.1 $ 153.7 $ fuel A graphite control rod Fig.4-a Typical core configurations used for adjustment of excess reactivity. 7-30 Step 25 New Core Arrangement 181$ fuel • graphite • control rod Fig.4-b Final core configuration obtained with new fuel elements. 3. Substitution Reactivity of Graphite Dummy Fuel and Void. We have proceeded with the rod calibration of the Regulating rod by means of the period method, after the values of control rod worth for three rods were measured by the rod-drop method. The results were applied to evaluate a substitution reactivity of graphite dummy fuel and void. Table 2 denotes the measured values of control rod worth. The neutron counts of FC1 chamber were used to evaluate the rod worth. The worth p, was calculated with the equation NO - Nb (1) /N(t)dt -/Nb(t)dt 7-31 where NÖ : count rate of neutrons at 100 mw critical power /N(t)dt : accumulated counts after the rod was dropped down Nb : count rate of background /Nb(t)dt : Accumulated counts of background Fig.5 shows a calibration curve of integrated worth for the Regulating rod where the values were plotted as a function of the rod positions. The doubling time Td was measured with a LIN-N output. The worth p was calculated with the equations Y. 8S .. a \ efetf + 7, • kT 1 + X.T T = T-d (3) In 2 where the constant value used are given in table 3. We have further measured the substitution reactivity for graphite dummy fuel and void. The experiments were performed at five different rings by substituting (1) water with fuel, (2) water with graphite, and (3) water with void. Void was modelled by inserting a hollow Al- tube of 3 cm I.D. through the hole of the grid plate. Table 4 shows substitution reactivity measured for fuel, graphite and void. The measurements were at five different positions. The results are plotted as a function of the position of elements in fig.6. The substitution reactivities decreased gradually, as the element position was changed from the center to the rim for three cases. It is noticeable that the void effect of reactivity becomes negative outside the D-ring. The tendency is understood as follows; (1) the bulk of water of the center region works as a neutron absorber, and (2) the bulk of water close to the rim acts as a neutron reflector. 7-32 Table 2 Rod worth values measured with rod-drop method. NO /N(t)dt ' Nb /Nb(t)dt Worth Rod (cpm) (counts) (cpm) (counts) (% Ak/k) Safety 89408 4875 3.8 23 3.2 Shim 82107 6019 4.3 26 2.4 Reg. 89752 22233 17.3 104 0.7 X = 0.0767 sec , BQff = 0.008 * accumulation time = 6 min 100r Reg.Rod Worth 84 censt3 25 50 75 100 Rod Position (%) Fig.5 A calibration curve for the Regulating rod. 7-33 Table 3 Constant values used in the calculation of reactivity. - L Y = 0.03300 xl 0.01240 sec l L Y x2 = 0.03050 sec "' 2 = 0.21900 L Y X = 0.11100 sec "' = 0.19600 3 3 L Y = 0.39500 X4 = 0.30100 sec 4 = JL Y = 0.11500 X5 1.14000 sec " 5 1 Y X sec "' = 0.04200 6 = 3.01000 6 I = 0.80E-04 sec ßeff = 0.00800 - Fuel-water •^150 - • - Graphite-water +•> c O- - Void-water 0) •I 100 o CO CD cc C o 50 CO .a CO +20 + 10 • 0 -10 o -20 B c D E F 4 7 11 15 19 Element Position Fig.6 The substitution reactivity for fuel, graphite and void. 7-34 Table 4 Substitution reactivities. (Unit: cents) Measured points B4 C7 Dil E15 F19 Fuel - Water +150.3 +113.6 +90.1 +60.3 +34.5 Graphite - Water +44.7 +25.4 +18.5 +8.4 +3.8 Void - Water +16.2 +0.3 -3.6 -8.5 -12.3 4. Measurement of Temperature coefficient in Reactivity We prepared 80 units of the fuel elements. Three of them are instrumented fuel ones. These instrumented elements are equipped with three chromel- alumel thermocouples and have the same dimensions as standard elements. The temperature distribution in the core and the temperature coefficient of reactivity were measured by using the instrumented fuel elements. Fig.7 shows the fuel temperature versus the controlled power measured at the position of B4 and F9. The numerical values are given in table 5. The highest temperature of 154 °G was observed at the position B4 when the reactor was operated at 100 kW. The value of 120 °C was measured at the position F9 of the most peripheral ring at the different day with the same power. Two curves are intersecting in the figure, because the initial temperatures of water in the core tank were different between two measurements. Fig.8 gives time dependences of LOG-N, LIN-N, two control rods position, fuel temperature and water temperature. The data were collected at every 0.5 sec. The rod position is deffined as 100 % when the rod is completely pulled out until UP-light turns on. At a low power operation, the control rods return to the previous position after the desired power is achieved. When the operation power is raised to a high level, the movement of control rods follows the change of the fuel temperature. A typical movement of the Regulating rod in fig.9. The fuel temperature reaches to a stable state with a time constant of 6 - 9 sec. 7-35 Fig.10 shows two curves of the fuel temperature and the reactivity loss. The reactivity decreases, as the core temperature rises. In order to keep the reactor power constant at a desired level, the reactivity loss is compensated by drawing the control rods. The reactivity losses were evaluated from the displacement of the Regulating rod fay using the calibration curve (see fig.5). The value of reactivity was lost by 83 cents due to the core temperature rise during the 100 kW power operation. This loss is approximately a half of total excess reactivity of 181 cents. Fig.11 shows the reactivity losses as a function of the fuel temperature measured at B-ring and F-ring. The temperature coefficient of reactivity is related to the slope of the curve. The temperature coefficient of reactivity dp/dT is evaluated by using the equation E k. N. dp i % % P_ = _! (4) dT EN. ^ v where k. is a slope of the reactivity loss to the fuel temperature if at i-th ring, and N. the number of fuel elements loaded in i-th ring. The temperature coefficient of reactivity was obtained as - 6.5 x 10 ~5 Ak/k/°C. A theoritical calculation with the code of Wims and Citation predicted the coefficients of - 8.0 to - 9.4 x 10 "5 Ak/k/°C in the temperature range of 50 to 100 °C. 7-36 200 r 150 Ü o 100 « a. E 0 25 50 75 100 Reactor Power (kW) Fig.7 Fuel temperature as a function of reactor power. Table 5 Fuel and water temperatures for two different runs measured as a function of reactor power. 0r>. (Unit: UCJ ** Power (W) B4 F9 Fuel Water Fuel Water 100 16.3 15.5 27.3 31.7 1 k 19.1 15.5 31.4 30.6 5 k 28.5 15.5 35.8 30.7 10 k 38.5 15.8 43.5 30.7 25 k 63.8 16.2 59.0 30.8 50 k 99.2 16.7 82.5 31.2 75 k 128.0 17.9 102.2 31.8 100 k 154.2 19.9 120.1 32.6 * 1985.12.05 ** 1985.11 12 7-37 100 LIN-N ae 50 01 hr ii it i» 'JI ii » '\M \\t \a \n in 100 S3 50 SHIM ROD 0>- IT il il il \> hi ' 'lui in -W> 100r REG ROD £ 50 r ~\ /I. A/v /V^vJ^. . ij ijl JT— 'ill ill i—' " II ' '• Ji ^ 'ii h n H 'il— 200 60 go 120 150 TIME (min) Fig.8 Time dependences of LOG-N, LIN-N, two control rods position, fuel temperature and water temperature. The fuel temperature was measured at B-ring. 7-38 c o = 65 o(0 a. 100 •o 55 o (X oc o 5 75 150 t, CD o O as w a. CD a S 50 Ca 100 I O "© Œ 3 25 50 10 20 30 40 0 10 20 30 10 20 30 40 Time (sec) Fi g. 9 Responses of fuel temperature, reactor power and Regulating rod position. The temperature was measured at B-ring. 10-3 10-1 1 10 102 Reactor Power (kW) Fig.10 Fuel temperature and reactivity loss versus reactor power. 7-39 100 40 80 120 40 80 120 Fuel Temperature (°C) FueJ Temperature (°C) Fig.11 Reactivity losses as a function of fuel temperature measured at B-ring and F-ring. 7-40 Ill Conclusion Some fundamental experiments were performed to obtain several characteristics of the new core. We summarize the characteristics of the new core as well as the old core in table 6. The present work of the fuel, replacement will be help a 250 kW power-up plan. Table 6 Characteristics of the new and old cores, Item New core Old core Fuel Cladding Stainless steel Aluminum Critical mass 66 units 58 units (1985.07.25) (1963.01.30) Actually loaded mass 73 units 65 units Excess reactivity 181 cents 140 cents * Fuel temperature 154.2 °C (B4) Rod worth Safety rod 3.2 % Ak/k 4.1 % Ak/k * Shim rod 2.4 % Ak/k 3.3 % Ak/k * Regulating rod 0.7 % Ak/k 0.9 % Ak/k * Neutron flux 2 Central thimble 4E+12 n/cm sec 3.8E+12 n/cm sec 2 2 Pneumatic tube 1E+12 n/cm sec 1.5E+12 n/cm sec 2 2 Ci re. well of reflect. 4E+11 n/cm sec 8.0E+11 n/cm sec 2 2 Irradiation room 1.3E+09 n/cm sec 1.0E+09 n/cm sec * 1984.12.05 7-41 i \ CRITICALITY CALCULATIONS FOR THE TRIGA MARK-11 REACTOR OF ITü BYOTHE FINITE ELEMENT AND FINITE DIFFERENCE METHODS t \ i H. Atilla özgener Istanbul Technical University Institute -for Nuclear Energy I_s_ INTRODUCTION Finite di-f-ference method Although the finite element method (FEM) has been used extensively in the field of structural mechanics since mid-1950's, the application of FEM to the neutron diffusion/trasport theories dates back only to mid—1970's. FEM has two major advantages over FDM. For one thing, irregular geometries could easily be modeled due to the large number of elements of different shape available. For another, by increasing the order of polynomial approximation, the order of convergence could easily be increased. Our recent study C13 indicates that the asymptotic order of convergence for multiplication eigenvalue is 2xN, whre N denotes the degree of polynomial approximation, if the souce term is consistently treated. If we recall that the asymptotic order of convergence is only 2 for the FDM, it is obvious that there is a potential to decrease the computer time requirements dramatically by prefering high degree FEM to FDM. The question, whether this potential is realized in practical calculations or not, is one of the questions which we would try to answer in this work. In this study, TRIGA Mark-II reactor of the Istanbul Technical University (ÎTu) is treated in cylindrical geometry. Using two-region and ten-region models C23 of this reactor, both FDM and FEM have been utilized to solve multiplication eigenvalue problems. Polynomial approximations up to degree ten have been used in the FEM solutions. Such high degree polynomial approximations are not reported in the literature, perhaps due to the difficulty of assembling the 7-43 coefficient matrix- By the use of the computer also in the •formulation o-f the problem CID, such high degree approximations are made possible. The relative computer execution times of FDM and various degree FEM solutions are compared and their relative merits in TRIGA calculations are assessed. Both consistent and lumped source variety FEM solutions are obtained. II. THEORY II.1 MITHIN-GROUP DIFFUSION EQUATION The within-group diffusion equation could be written as: -?.D s (r) = If* Z (r) QT (r) + r-%* TL v 2 It can be shown C3D that the minimum of the functional: 7-44 FCg ldV+d/2) §^ (P)dS V <5> is the solution o-f (1) satisfying also the boundary conditions of (3) and- (4). Hence, the minimization of the functional is equivalent to the solution of the within-group diffusion equation. In FEM, the minimum of (5) is sought in the space of piecewise polynomials of certain degree. That is, V is divided into a finite number of subregions, which are called elements. In each element, the flux is assumed to be a polynomial of a certain degree. In our study, the flux is assumed to be continuous across interelement boundaries, but no continuity condition on the derivatives are placed. That is, Lagrangian type C4J finite elements are utilized. In this study, we resticted ourselves to one-dimensional cylindrical geometry in our models. Hence the brief FDM and FEM developments of the next sections are simplified. II.2 FINITE DIFFERENCE FORMULATION We consider a cylindrical system of outer radius R and divide it into I concentric homogeneous cylindrical regions, as shown in Figure 1. ri=0 rz r,-i rt ri*t r, rx*i=R Fig.1 Finite Difference Mesh Material interfaces are allowed only at the boundaries rt for i=2, . . . , I. We take the concentric region extending from r4-K = lr±-x •+• rt )/2 to r^, = (r,. + r1^-1)/2 as a typical control region. Di.-x,2t-x D*,I* -i •t-i r^-'-i r4 rx^ rt»i Fig.2 Control Region Then we integrate the within-group diffusion equation in cylindrical geometry over the two homogeneous parts of the typical control region as: r» rt rx f 1 d d0 i-i f f 1 1 - D " J_a_CraF3rdr+Zr J 0(r) r dr = J s(r) r dr r^—ut rt—kt r x—t (6) 7-45 r~ i •*•>< T t •»•'•* d d0 - DJ -^Cr-r— Drdr + E 0(r) r dr s (7) Adding (&) and (7) and noting the continuity o-f current at r± we obtain: d0 d0 i — i - D1 ri* — + Zr- | 0 r i -•..* r t r i «i< + I,- [ 0 I-f we make the 0(h2) approximations: d0 „ 0i - 0*_ x (9) dr r=r±_^ rt - rt_x d0 I „ 0i-i - 0i (10) dr Ir^r-i.*.^: ri*! - rt ( r2± - r2i-^ ) 0(r) r dr ~ 0 Ul> J X J < r t*4 - r2t ) J 0(r) r dr s; 0± (12) rt we obtain the result: ilZ) ai.i-i 0i-i + <3i.i 0i + at.1*1 0i*, = fi where äi.1—1 — (14) ri - rt-i at .1*1 (15) ri-».i - ri i-i (r2,-r^-H) i (r^LH-r^) = — 3» ,l (â| ,i-i + 3i ,i»i)+ïr "*" Er (16) Using (2) and the approximations (11) and (12) we can write: 7-46 0_i - - (r2t-r2t-^) * ir^^-r^) fl = S C r..^-, : + 2..^Q 3 01 »-1—1 2 2 1 1 A %*, 0 " (r»t-r2t^) (rî.^-rM + S C v £*,„ + V 2*.r, 3 0i * 2 * 2 (17) The central point i=l and the outer point i=I+l are special points since the inner part o-f the control volume is nonexistent -for the -former, and the outer part o-f the control volume is nonexistent for the latter. Also the general boundary conditions (3) and (4) which can be written simpler as: d0 I 1 dr = ot 0X^X <18) r=ri-Hj. where oc = 0 -for re-flecting boundary (19) -for vacuum boundary. 2D1 are to be applied -for i = I + l. The equations -for these special points srez ai.i 0i + ax.2 03 = -fi (20) ax»!.x 0i + aIM,, + 1 0,*! = fx*i (21) The definitions o-f ai,2 and ai-^x,i are as in (15) and (14) respectively, whereas ai,i and aj«!,!*! are de-fined by: I I ( r 2 — r- 2 \ ax-^xmI — l — ai*!,; OC ri-t-i Lf f 2,,. — lijj The saure terms are de-fined by: a-i * (r23/2-r2t) H -fi = S 2..^-,, 0X H—1 -> + 2: v 2*.»n 0i (24) 7-47 f 1*1 = S Em.n-i,, 2l*t t-,— 1 n The resulting linear system -From the -finite difference discretization is of the form: A 0 = f <26> The matrix A is a (1+1) dimensional, symmetric, positive de-finite, tridiagonal matrix. 0 and f are (I + i) dimensional vectors and (26) is to be solved for 0. The right hand side vector f can also be written in terms o-f the scattering and fission matrices as: f = S S 0 + --— 2 F 0 (27) k-, f»— » S1-*^» and F^1 are diagonal matrices in the finite difference formulation. II.5 FINITE ELEMENT FORMULATION In one dimensional cylindrical geometry, the functional (5) simplifies to: R P d5 FC5(r) 31 == ]CD(r)(— CD In FEM, the system extending from r=0 to r=R is divided into J elements and 5(D is assumed to be a polynomial of degree N within each element. Each element is assumed to be homogeneous and <5(r) is restricted to be continuous across interelement boundaries. A typical element j extends from ri,j to rN—x.j and the unifomly spaced points within the element r^.j for n=l, . . ., N+l are called the nodes. Under these circumstances, we can write: §(r) = S hrl.J(r) 5„.j , for rx,A< r < r„+XmJ (30) 7-48 where h„.j (32) This behaviour could be represented simpler as: 5 , i-f we de-fine the (N+l) dimensional vectors as and (§j)n = 5r, . J (35) The -functional (28) could be written as the sum of the element functional s as : FC5(r)D = E FjC5(r)3 (36) J F_,E§(r)3 f d$- ) 2 + I,_<3 (r) ) 2- 25(r)s(r)3rdr CD < ; Jri. j dr + ± rN»».j(5 ,if JAJ where Ss = { (38) 1 ,if J=J When (33) is inserted into (37), we obtain: T FJCSJII = ïj ftj îj - 2 ÏJT Sj (39) where i»»-» j . t Aj = D (h,) (hjT) r dr + hj hjT r dr dr dr i . -i 7-49 r x J ^T ' Sv Sx S (40) hj s Aj is a (N+l) dimensional square matrix called the element matrix and fj is a CN+1) dimensional vector called the element source vector, ê is a (N+l) dimensional square matrix whose all elements arB zero except the (N+l,N+l)'th element which equals one. The system is divided into J elements and thus contains NxJ + 1 nodes. The n ' th node of the j'th element is actually Nx ( j — l)+n'th node o-f the system. I-f we define the NxJ + 1 dimensional vector § as the vector which has the function values at the nodes o-f the system as its elements: (£) (J-l)KN«n = (|j)n (42) , then the equation: fj = Jj £ (43) could be written where J^ is a (N+l) by (NxJ+1) dimensional rectangular matrix, called the Boolean matrix, de-fined by: Then all element -functional s could be summed up to obtain the total -functional as: FE53 = Sp" A | - 2 ^7 s (44) where A = X JjT Aj Jj (45) and j -f = S JjT 5j (46) j —i — I-f we recall that, we are trying to -find the minimum of the functional in FEM, we look for the minimum of (44); but the vector 0 which gives the minimum of (44) is just the solution of the linear system: 7-50 A 0 = f (47) Thus, as in FDM, also in FEM, we end up with a linear system to solve. The coe-f -f ici ent matrix A in (47) is symmetric and positive de-finite. But this matrix, called the system matrix, has a width equal to II.4 CONSISTENT AND LUMPED SOURCE TREATMENTS IN FEM I-f we use (2) in (41) , we see that the element source vector is o-f the -form: fj = 2 I-.h^«, hj(r) 0 (r) r dr t-i—i J r i•. -» IM-»-l » J v 3 j p n + _JZ_ 2: v Z+ H hjîr) 0 (r) r dr (48) in FEM. Thus, the evaluation o-f the integrals in (48) requires some treatment o-f 0*Mr). There are two di-f-ferent treatments, reported in the literature CI, 33, -for this purpose. One method is using the -finite element expression, (33) in (48), that is: hj(r> 0 (r) r dr = ( hj (r) hj (r) r dr) 0j (49) This treatment is called the consistent source treatment since it is consistent with the previous -finite element -formulation. Another treatment is based on using the flux value at a particular node in place o-f the position dependent -flux in evaluating the element o-f the source vector corresponding to that node. That is: (fj)n = S I..M-4, ( h„.j(r) r dr) <0„)„ n— l + S v E*.„ ( h„,j(r) r dr ) (0j)n (50) k.-f- f i->—1 7-51 This kind of treatment is called the lumped source treatment. Both treatments result in an equation like (27): f = S S 0 + -2- 2: F 0 <51) But when the lumped source treatment is used, the scattering and -fission matrices are diagonal matrices like in the FDM- If the consistent source treatment is pre-f erred, the scattering and -fission matrices become symmetric matrices with a hal-fbandwidth equal to III. APPLICATION OF THE FDM AND FEM TO THE SOLUTION QF MULTIPLICATION EIGENVALUE PROBLEMS OF THE TRIGA REACTOR III.1 TWO REGION MODEL OF TRIGA MARK II REACTOR OF 1TÜ For the test of relative performance of FDM and FEM, a two, region two group model C2D of TRIGA Mark-II reactor of ITÜ is considered. In this model, the central region is the core with a radius of IS.51 cm; the core is surrounded by a reflector of outer radius 54.61 cm. The transverse neutron 2 leakage is modeled by applying a classical DB a correction to the relevant cross sections given in Ref.2. Core Reflector 18.51cm. 54.61cm. Figure 3 Two Region Model The reference multiplication eigenvalue 7-52 ten with each of the source treatments and -For the -finite difference solution, the coarsest mesh, which still gives a percent error in k«.+ ^ less than 7.0.01, is reported in column five. For the FEM solutions, the used source appproximation is given in column two. The calculated k.** is given in column three and the percent error in k«.** is reported beneath, in paranthesis. The normalized execution time is defined as the ratio of the execution time of a run to the execution time of the run, which has the shortest execution time. Then, by definition, the fastest solution has a normalized execution time of one. The normalized execution times are given in column four of Table 1. The last column of Table 1 gives the number of unknowns per group, which is equal to degreex2xn+l in this case. Table 1. TWO REGION TRIGA RUNS WHICH REDUCE k.++ ERROR TO LESS THAN 7.0.01 RELATIVE TO REFERENCE EIGENVALUE OF k.**=l.116223 Degree Source k_ * Normalized Mesh No. of + exec, time poi nts FD — 1.116112 3.26 A22 45 (7.0.0099) 1 con- 1.116116 3.24 A20 41 si stent (7.0.0096) 1 lumped 1.116334 9.86 A72 145 (7.0.0099) 2 con- 1.116205 1.29 A3 13 si stent (7.0.0016) 2 lumped 1.116292 1.87 A6 25 (7.0.0062) 3 con 1.116266 1.23 A2 13 sistent (7.0.0039) 3 lumped 1.116295 3. 18 A7 43 (7.0.0064) 4 con 1.116277 1.00 Al 9 sistent (7.0.0043) 4 lumped 1.116224 2. 14 A3 25 (7.0.00009) 5 con 1.116242 1. 14 Al 11 sistent (7.0.0017) 5 lumped 1.116275 3.69 A5 51 (7.0.0047) 6 con- 1.116227 1.34 Al 13 si stent (7.0.00035) 6 1 Limped 1.116258 2.70 A3 37 (7.O.0O31) 7 con- 1.116224 1.61 Al 15 si stent (7.0.00009) 7 lumped 1.116308 4.52 A4 57 (7.0.0076) S con- 1.116223 1.79 Al 17 si stent (7.0.0) 8 lumped 1.116257 3.94 A3 49 (7.0.0030) 9 con- 1.116223 2.02 Al 19 si stent ( 7.0. 0 ) 9 1umped 1.116321 5.90 A4 73 (7.0.0088) 7-53 r ä «Hfl led eiuij. uoijn)ax3 psznouwoN 7-54 The results presented in Table 1 is also depicted graphically in Fig. 4, where the normalized execution time is shown as a •function o-f element degree- One thing that could be noticed •from Fig. 4, is the -fact that consistent source FEM solutions are much faster than their lumped source counterparts. Also, even degree lumped source FEM solutions are -faster than odd degree lumped source FEM solutions. These observations &re consistent with our earlier investigations CI,51. The shortest execution time is obtained with consistent source •fourth degree -finite elements. The classical -finite di-f-ference solution is approximately Z'£ times slower than this solution with the shortest execution time. Also, the lumped source linear -finite element solution is very slow, compared to either consistent source linear FEM or FDM solutions. A very fine, A72, mesh is necessary to reduce the k_-f+ error to below y.0.01 for linear lumped source finite elements. In Fig. 5, the number of unknowns per group is shown graphically as a function of element degree. The general trend in Fig. 5 is similar to Fig. 4. The increase in the execution time or the number of unknowns per group, as the element degree increases, for degrees greater than three , with consistent source, is due to the increase in the number of nodes per element as the element degree increaseses. Actually, consistent source FEM solutions use all Al mesh for element degrees greater than three. III.2 TEN REGION MODEL OF TRISA MARK-11 REACTOR OF ITu The second problem we have treated is a ten region model C23 of TRIBA Mark-II reactor of ITü. In this model the first region is the central thimble. B,C,D and E rings constitute the second, fourth, sixth and eigth regions, Regions three, five and seven are left for the insertion of control rods. F ring (the dummy elements) constitutes th e ninth region. Radial graphite reflector is region ten The transverse leakage is again modeled by applying a classical DB*0 correction to the relevant cross sections giv en in Ref. 2. l.?27cm 6.268cm 9.395cm 13.86cm 24.69cm —1- -S- -10- -> 0. cm 5.968cm 9.708cm 13.73cm 18.51cm 54.ölcm Figure 6. Ten Region Model The reference multiplication eigenvalue 7-55 contains only one element. When each region is divided into n elements, the resulting mesh is called the Bn mesh. Again for all approximations, n is increased so that the percent error in k.-r-i« is less than 710.01 compared to the reference k«,-?* value of 1.089877. In Table 2, for each finite element solution and for the finite difference solution, the coarsest mesh, which still gives a percent error in k«,** less than X0.01 is reported in column five. The format of Table 2 is the same as the format of Table 1. The number of unknowns per group is reported again on the sixth column and is equal to degreexlOxn+1, this time. Table 2. TEN REGION TRIBA RUNS WHICH REDUCE k...,., ERROR TO LESS THAN 7.0.01 RELATIVE TO REFERENCE EIGENVALUE OF k_**=1.089877 Degree Source k.-n. Normali zed Mesh No. of e>i ec. t i me points FD — 1.089982 3.26 B12 121 (7.0.0096) 1 con 1.089969 1.67 B6 61 sistent (7.0.00844) 1 lumped 1.089984 5.35 B22 221 (7.0.0098) *"> con 1.089938 1.21 B2 41 sistent (7.0.0056) 2 lumped 1.089952 1. 11 B2 41 (7.0.0069) 3 con 1.089936 1.00 Bl 31 sistent (7.0.0054) 3 lumped 1.089947 1.56 B2 61 (7.0.0064) 4 con- 1.089879 1.33 Bl 41 si stent (7.0.00018) 4 1umped 1.089893 1.11 Bl 41 (7.0.0015) 5 con 1.089878 1.72 Bl 51 sistent (7.0.00009) 5 1umped 1.089886 2.65 B2 101 (7.O.0O083) 6 con 1.089877 2.26 Bl 61 sistent (7.0.0) 6 lumped 1.089880 1.72 Bl 61 (7.0.00027) 7 con- 1.089877 2.76 Bl 71 si stent (7.0.0) -7 j lumped 1.089972 Bl 71 (7.0.0087) 8 con- 1.089877 3. 30 Bl 81 si stent (7.0.0) S 1umped 1.089880 Bl 81 (7.0.00027) 9 con- 1.089877 3.79 Bl 91 si stent (7.0.0) 9 lumped 1.089986 2.68 Bl 91 (7.0.01) 7-56 dnojg jad SUMOU^UD JO jaqiwiN atutj. uoijnaaxg pazijowjoN 7-57 The results presented in Table 2 is given also graphically in Fig- 7, where the normalized execution time is shown as a -function of element degree. In contrast to the two region model, consistent source -finite element solutions Are not always faster than their lumped source counterparts. For element degrees greater than three, both consistent and lumped source finite element solutions require only the Bl mesh, with the exception of consistent source fifth degree elements. Since the diagonalization of the scattering and fission matrices reduces computing time, we observe that lumped source FEM solutions are faster than consistent source FEM solutions of the same degree, with the exception of fifth degree elements. The physical reason for this behaviour is the small thickness of most of the regions in this problem. Obviously, the cruder lumped source treatment is sufficient when the regions are thin and the source is reasonably flat. The shortest execution time is obtained with third degree consistent source finite elements, using the Bl mesh. Like in the previous model, lumped source linear finite element solution is the slowest solution. One thing, that souid be noticed in this problem, is the performance of the consistent source linear finite element solution compared to the finite diffrence solution. The execution time of the consistent source linear finite element solution is nearly one half of the finite difference solution, in contrast to the situation observed in the previous two region problem. This situation is due to the B6 mesh used by consistent source linear FEM in contrast to the B12 mesh used by the FDM. The execution time increases as the element degree increases for degrees greater than three when consistent source treatment is used. This is due to the increase in the number of unknowns per group as the element degree increases when the basic Bl mesh is retained. This situation is especially obvious in Fig. 8 where the number of unknowns per group is shown graphically as a function of element degree. IV.CONCLUSION In this study, we applied FDM and FEM for solution of multiplication eigenvalue problems of a TRIGA reactor and tried to make sin assesment of their relative numerical performance. This study indicates that the FDM solution is slower than most of the FEM solutions. Especially, finite element solutions of third and fourth degree with the consistent source treatment seem to be the most effective solutions. Finite element solutions of degree greater than four are not very efficient in the solution of model problems we have studied. This is due to the fact that each region could be modeled adequately only by one element. Thus, an increase in the element degree results in an increase in the number of unknowns per group and in the execution time. This study was limited to one dimensional cylindrical geometry applications. Hence, the gains obtained by using FEM 7-58 instead o-f FDM are not very dramatic. In multidimensional TRIGA applications much more dramatic reductions in execution time are possible by using high order FEM instead o-f FDM. REFERENCES CI 3. H.A.özgener, "An Order o-f Convergence Study -for Lagrangian Type Finite Elements o-f Arbitrary Degree Using the Consistent and the Lumped Source Approximations", Istanbul Technical University, Institute -for Nuclear Energy, Research Report No:37, NEE-37 (1988). 121. A. Anacan, A. Yücel, T. Yarman, "ITü TRIBA Mark-II Reaktörü Fizik Hesaplari", Istanbul Technical University, Institute for Nuclear Energy, unpublished research report. C3D„ E.E. Lewis, "Finite Element Approximation to the Even Parity Transport Equation", Adv. Nucl• Sei. and Tech. , 13 (1981). C43. G. Str An Analysis o-f the Finite Element Method. Prentice Hall (1973). C5D. H. A. özgener, "A Timing Comparison Study of Finite Difference and Arbitrary Degree Finite Element Methods", Istanbul Technical University, Institute for Nuclear Energy, Research Report No:38, NEE-38 (1988). 7-59 \ An Analytical Approach to the Positive Reactivity Void Coeffitient of TRIGA Mark-II Reactor Erding Edgü Istanbul Technical University, Institute for Nuclear Energy Tolga Yarman Anadolu Science and Technology Strategies Research Institute ABSTRACT Previously calculations of reactivity void coefficient of I.T.U.* TRIGA Mark-II Reactor was done by the second author et al.. The theoritical predictions were afterwards, checked in this reactor alright experimentally. In this work an analytical approach is developed to evaluate rather quickly the reactivity void coefficient of I.T.Ü. TRIGA Mark-II, versus the size of the void inserted into the reactor. It is thus assumed that the reactor is a cylindrical, bare nuclear system. Next a belt of water of 2TTrArH is introduced axially at a distance r from the center line of the system. r here, is the thickness of the belt, and H is the height of the reactor. The void is described by decreasing the water density in the belt region. A two group diffusion theory is adopted to determine the criticality of our configuration. The space dependency of the group fluxes are, thereby, assumed to be r J0 ( -2--M05 , ) cos(J[-l_)j -the same as that associated with the original bare reactor uniformly loaded prior to the change. A perturbation type of approach, thence, furnishes the effect of introducing a void in the belt region. The reactivity void coefficient can, rather surprisingly, be indeed positive. To our knowledge, this fact had not been established, by the supplier. The agreement of our predictions with the experimental results is good. * * I.T.Ü. : Istanbul Technical University 7-61 INTRODUCTION Many of us still find tiresome to go through conventional spectrum and. multigroup diffusion codes in order to establish the physics characteristics of a reactor, no matter what the sise of this may be. It is naturally desirable to predict various features of a given reactor, without having to deal with fancy and uneasy tools, if this were ever possible. This is the philosophy behind the work presented here. Of course, the supplier (G.A.) did the necessary reactor physics calculations of TRIGA Mark-II prior to, and surely, after the commisionning. Following the decision in regards to the erection of a TRIGA Mark-II at I.T.U., Institute for Nuclear Energy, the second author and al., neverthless, undertook, as a local exercise, the reactor physics and dynamics calculations of this reactor (1-10). By the time, we did not have access to the supplier's constants or computational tools. We thus aimed to determine few group cross-sections of I.T.Ü. TRIGA Mark-II, quite separately from the supplier's methods. For this purpose, based on the actual configuration of the reactor, we dressed up a two dimensional (r,s) equivalent reactor configurations. Next, as an acceptable approximation, the Bondarenko set of cross-sections was used (11). Further refinements were achieved through the use of the spectrum and few group cross-section calculation codes GGC-4, ANISN and LEOPARD (4,5,6). The calculated few group homogenised cross-sections were input to a 2-D multigroup diffusion code, TWENTY-GRAND (12). The results of the calculations in question were presented to the previous TRIGA meeting held in 1982 in Istanbul (9,10). It is worthwile to note that the work we had done apart from the supplier, can not be considered as a duplication. This, for at least two reasons. The first one is that the prediction of the supplier had been a failure. Î.T.U. TRIGA Mark-II reactor was, in effect, guessed to become critical with 63 fuel elements. Yet our criticality was achieved with 69 fuel elements. The second reason in question is that our approach predicted features of I.T.Ü. TRIGA Mark-II that were rather unknown to the supplier as well as the TRIGA users. 7-62 Here, we should namely cite the likely positive reactivity void coefficient of the reactor(8,9). REACTIVITY VOID COEFFICIENT OF TRIGA MARK-II Let us summarise the previous findings about the reactivity void coefficient of I.T.Ü. TRIGA Mark-II reactor. 2-D, two-group diffusion calculations were performed based on the equivalent core configuration that we had originally drawn. The void is thus described by decreasing the water density in the central water region. Additional computations are performed where an outer fuel ring region is removed and the region is supposed to be filled with water to form a water belt. The void is, here again, simulated by decreasing the water density in the belt region. It was found that the reactivity void coefficient can essentially be positive in the central region. This is shown in Fig. 1 (8,9). Thus the reactivity void coefficient, first, gradually increases with respect to the void diameter, then abruptly decreases. -P(D) x io5 D (cm) Figure 1 Reactivity values computed in 2-D when voids with various diameters are inserted to central water region (3,9) 7-63 When void is inserted, absorption in water is decreased. This is a negative effect. But moderation is increased. This is a positive effect. Leakage is also increased, which is a negative effect too. We thence observe that the net effect of introducing an axial void in the central region of TRIGA Mark-II can be positive. After the positive reactivity void coefficient was established in I.T.Ü. TRIGA Mark-II, it was proposed to check the finding experimentally (13). The result was that, the prediction was quite correct. The experiment was undertaken with an empty Al tube of 6 cm in diameter which was introduced at different locations of the reactor, in different depths. The reactivity thus becomes as positive as 1x10" (about 15 cents). Note, first of all, that experimentally, the diameter of the void could not have been varied at a given location. Instead a void with the same diameter was axially introduced from different places to the core. Neverthless, by varying the depth in the reactor, to which the empty tube is introduced, one can hypothetically vary the diameter of the void volume. This way if we wish, we can hope (by extending the depth of the void to the height of the reactor core and still keeping constant the volume of the void introduced) to develop an idea of how the diameter affects the reactivity, for a void created from the top to the bottom of the reactor core. Secondly, what is measured is the actual reactivity void coefficient. What we predict, however, is a slightly different quantity. In effect, in our model, void was simulated by merely decreasing the water density in a uniform way. Another point is that the mentioned measured value is obtained by introducing the empty tube in the place of a fuel element in ring B. Whereas Fig.1 is (theorettically) obtained in the central region without the removal of any fuel elements. It should be for these reasons (mainly perhaps the last reason) that, 2-D prediction of the void effect is a little lower than the measured value. Note that creating a larger volume of water around the central water region should enhance the 7-64 subsequent void effect. In this case, indeed, the .introduction of the void leads a lesser negative effect of moderation, whereas the effects due to the decrease in absorption and increase in leakage do not pratically change when we switch from the initial central water region to the enlarged water region in question. * Now we aim to predict the reactivity void coefficient of I.T.U. TRIGA Mark-II in a much simpler manner. AN ANALYTICAL APPROACH TO THE TRIGA REACTIVITY VOID COEFFICIENT Here we propose to present I.T.U. TRIGA Mark-II as a bare cylindrical reactor having a central water region (Fig.2). Figure 2 (r,s), bare model for I.T.Ü. TRIGA Mark-II, R = 24.6 cm, H = 38. 1 cm, Ar = 5.968 cm Void is introduced in the central region. The introduction of the void is simulated by the uniform decrease of various macroscopic cross-sections associated with this region. 7-65 Cross-sections are obtained through ordinary homogenisation of the cross-sections, previously established, over the space (8,9). We thus obtain two-group macroscopic cross-sections for the configuration shown in Fig.2 (Table 1). Table 1 Cross-sections for the simple (r,a) core configuration "a.2. ^-"S — irl Core 6.98E-2 6.19E-2 3.51E-1 1.81 8.36E-2 Water 1.97E-2 7.75E-2 2.95E-1 2.09 To describe the behaviour of the space and energy dependent neutron flux, we now adopt a two-group diffusion scheme. We thus have with the familiar notation: V-D,(r) V(p,(r) -2_(r)(p,(r) + -L v2f2(r)(D,(r) = 0 (i) V. D,(r) V(D,(r) - 2aJ (r) (D2(r) + ST(r) 0,(r) = 0 . (2) Here for simplicity, we neglect absorptions in the first group. Thus 2(r), alsoZ(r) are taken to be zero. Furthermore: a(r) = —1 (3) 32., (r) We assume that,<£(r) can be expanded, again with the familiar notation, * 0,(r)= q), J.(2^)cos(Jîf)-, g = 1.2. (4) R H This expansion is of course an approximate one. <$Kr) is thus different from d)(r)V . 3 7-66 If now we go back to Eqs. (1) and (2) with the approximate expansion of Eq.(4) we obtain residuals. We can restore the broken equalities in a weighted residual sense., We thus weigh the residuals in question by the function, J-. (• 2.405 r •) cos(^-); integrate the weighted residuals over the reactor Tvolum e and set the products to aero. These.manipulations gives us an equation for \, the unknown (approximate) eigenvalue of the reactor. Let us define the following quantities for the central water region and the rest of the core. (We denote with "M" the water region and "F" the core region.) D, = D, (f ) r< A r0 (5) D, = D,(r) (6) 2ir ^(r) (7) (£) (8) a2 a2 D = D, (r) r> Ar, (9) D: = D, (D (10) (11) 21 21 = 2.,(r) (12) Using these quantities A is found as follows I,* (i.^;*!^;) A = (13) [I,(D:B^P + I2(D;B^:)] x 2 2 x [ I,(D;B +2:) + I2(D;B +2:2) ] 7-67 Here 1^ and lz are given as, (14) 2 2 I, = (ArJ [J0 UAr0) + J,' «Arj] 2 r2 = (0.519 R) - I, (15) with o<=2.405/R . It is thus question of modifying the water region macroscopic cross-sections in order to describe the introduction of the void into this region. Let than 21 be a macroscopic cross-section of interest. Suppose we introduce an axial void (of height H) having the diameter D, into the water region. The new macroscopic cross-section X'(D)'is thus supposed to be; (16) 2'(D) = 2(1-qD2) , ^= 4(ArJJ ' (17) Calculations are performed by varying D, and A each time, is determined through Eq.(13). Let us further define ^P(D) to be the reactivity due to the introduction of the void in question: ß[D) -- ^(0)-A • (18) A(D).A A is the eigenvalue of the reactor with no void in the water region, whereas A(D) is the eigenvalue of the reactor in the central region of which an axial void of diameter D is introduced. 7-68 The behaviour of J>(D) with respect to D is sketched in ! I I 55- 50 15 (0 35 30 IS 30 IS S Figure 3 Effect of the introduction of a void of various diameters into the central water region of the bare model This result seems quite satisfactory as compared, chiefly to the experimental result. The maximum of Fig.3 is about the same order of magnitude as the measured value for an empty tube of 6 crn in diameter. The maximum of Fig. 3 occurs at around 4 cm in diameter. In fact two opposite effects should really make our analyticaly approach work all right. The first one is that the analytical approach is applied to a core configuration where the original central region is enlarged after the removal of fuel elements in ring B. As discussed earlier this enhances the void effect, if now void is introduced in the central ring, A. Because of this fact the analytical approach should furnish a result higher than that of 2-D calculations (Fig.l). However there is another effect that should also be cosidered. In our analytical approach we overestimate the group fluxes around the central portion of the reactor. The result should be that we overestimate the negative effects due to the introduction of the void. The positive effect due to the decrease in neutron absorptions is also overestimated within the frame of our analytical approach. Perhaps the balance absorption-moderation is not altered much. But we conjecture 7-69 that an important extra leakage effect is created. Anyway the net effect due to the introduction of the void is softened. Thus, because of this reason, the perturbation type of approach achieved, somewhat underestimates the effect of the void. The result of the two contradictory effects we just described is that the analytical approach we performed, at last furnished a rather satisfactory output. CONCLUSIONS Throughout the exercise presented here, we have established a rather simple method to determine the reactivity void coefficient of TRIGA Mark-II. The method predicts quite correctly the reactivity void effect with respect to the diameter of the axial void inserted to the reactor. The prediction also turns to be quantitatively satisfactory. The method we presented is, on the other hand, in an acceptable aggreement with the results of a previous detailed computation achieved by the second author et al.. Note that the positive occurance of the reactivity void coefficient of TRIGA Mark-II had by then newly established. The reactivity due to the insertion of a void to TRIGA Mark-II can thus become as positive as lxlO"3. The reactivity void coefficient can become positive around only the central region of the reactor. The maximum reactivity shown in Fig.3 decreases as the void insertion location moves away from the center. This fact is obviously due to the increasing leakage effect. Such features too, can be analyticaly predicted through rather easy further elaborations. REFERENCES 1. A.Anacan, Physics Calculations I of I.T.U. TRIGA Mark-II, M.Sc. Thesis (Supervisor: Prof. T.Yarman), Institute for Nuclear Energy, Î.T.U., 1979. 7-70 2. A.Yücel, Physics Calculations II of I.T.U. TRIGA Mark-II, M.Sc. Thesis (Supervisor: Prof. T.Yarman), Institute for Nuclear Energy, I.T.Ü., 1979. 3. A.Akyatan, Physics Calculations III of I.T.U. TRIGA Mark-II, M. Sc. Thesis (Supervisor: Prof. T.Yarman), Institute for Nuclear Energy, I. T. Ü. , 1979. 4. N.Kiyak, Spectrum and Few Group Cross Sections Calculations via GGC-4, of I.T.U. TRIGA Mark-II, M.Sc. Thesis (Supervisor: Prof. T.Yarman), Institute for Nuclear Energy, I.T.U., 1982. 5. B.Çelik, Spectrum and Few Group Cross Sections Calculations via ANISN, of I.T.Ü. TRIGA Mark-II, M.Sc. Thesis (Supervisor: Prof. T.Yarman), Institute for Nuclear Energy, I.T.U., 1932. 6. R.Birol, Spectrum and Few Group Cross Sections Calculations via LEOPARD, of Î.T.U. TRIGA Mark-II, M.Sc. Thesis (Supervisor: Prof. T.Yarrnan), Institute for Nuclear Energy, I.T.U., 1982. 7. H.Arslan, The Inverse Kinetics Problem in I.T.Ü. TRIGA Mark-II , M.Sc. Thesis (Supervisor: Prof. T.Yarman), Institute for Nuclear Energy, I.T.Ü., 1982. 8. A.Anacan, A.Yücel, T.Yarman, Physics Calculations of I.T.Ü. TRIGA Mark-II, Bulletin of the Institute for Nuclear Energy, I. T.Ü. , 1981. 9. T.Yarman, C.Alp, T. Türker, S.Güngör, A.Anacan, A. Yücel, B.Çelik, R.Birol, Reactor Physics Calculations of I.T.U. TRIGA Mark-II, 7th European Conference of TRIGA Users, September 15-17, 1982, Istanbul. 10. Ö.Çiftçioglu, T.Yarman, E.Edgü, H.Arslan, Inverse Kinetics Analysis of I.T.Ü. TRIGA Mark-II, 7th European Conference of TRIGA Users, September 15-17, 1982, Istanbul. 11. I.I.Bondarenko Group Constants for Nuclear Reactor Calculations. 12. M.L.Tobias, T.B.Fowler, TWENTY GRAND Program for the Numerical Solution of Few-Group Neutron Diffusion Equations in Two Dimension, ORNL-3200. 13. M.A.Atalay, The Reactivity Void Coefficient Measurement in I.T.Ü. TRIGA Mark-II , M.Sc. Thesis (Supervisor: Prof. H.Yavus), Institute for Nuclear Energy, I.T.U., 1982. 7-71 TRIGA® Reactors GENERAL ATOMICS P.O. Box 85608 • San Diego, CA 92138-5608 Phone (619) 455-4255 • Telex 695065 GEN ATOM SDG • Fax (619) 455-4169