Soviet Atomic Energy Vol. 57, No. 6

Body:  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 ~, ~ ISSN 0038-531 Y. Russian Original Vol. 57, No. 6, December, 1984 June, 1985 5ATEAZ 56(6) 803-880 (1984) SOVIET ATOMIC ENERGY ATOMHAI:I 3HEP~VIF1 (ATOMNAYA ENERGIYA) TRANSLATED FROM RUSSIAN CONSULTANTS BUREAU,NEW YORK Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 SOVIET ATOMIC ENERGY Soviet Atomic Energy is abstracted or in- dexed in Chemica/ Abstracts, Chemical Titles, Pollution Abstracts, Science Re- search Abstracts, Parts A and B, Safety Science Abstracts Journal, Current Con- tents, Energy Research Abstracts, and Engineering Index. Mailed in the USA by Publications Expediting, Inc., 200 Meacham Ave- nue, Elmont, NY 11003. POSTMASTER: Send address changes to Soviet Atomic Energy, Plenum Publish- ing Corporation, 233 Spring Street, New York, NY 10013. Soviet Atomic E,aergy~-fs a translation of Atomnaya Energiya, a publication of the Academy of Sciences of the USSR. An agreement with the Copyright Agency of the USSR (VAAP) makes available both advance copies of the Russian journal and original glossy photographs`and artwork. This serves to decrease the necessary time lag between publication of the original and publication of the translation and helps to improve the quality of the latter. The translation began with the first issue of the Russian journal. Editorial Board of Atomnaya ~nergiye: Editor: 0. D. Kazachkovskii ' Associate Editors: A. I. Artemov, N. N. Ponomarev-Stepnoi, and N. A. Vlasov I.A. Arkhangel'skii A. M. Petras'yants ' I. V. Chuviio E. P. Ryazantsev I. Ya. Emel'yanov A. S. Shtan I. N. Golovin B. A. Sidorenko V. I..II'ichev Yu. V. Sivintsev P. L. Kjril:lov M. F. Troyano Yu, I. Koryakin V. A. Tsykanov _ E. V. Kulov E. I. Vorob'ev B. N'. Laskorin V. F. Zelenskii V. V. Matveev Copyright ?1~5, Plenum Publishing Corporation. Soviet Atomic Energy partici- pates in the Copyright Clearance Center (CCC) Transactional Reporting Service. The appearance of a code line at the bottom of the first page of an article in this journal indicates the copyright owner's consent that copies of the article may be made for personal or internal use. however, this consent is given on the condition that the copier pay the flat fee of $ 850 per article (no additional per-page fees) directly to the Copyright Clearance Center, Inc., 27 Congress Street, Salem, Massachusetts 01970, for all copying not explicitly permitted by Sections 1O7 or 108 of the U.S. Copyright Law. The CCC is a nonprofit clearinghouse for the pay-. ment of photocopying fees by libraries and other users registered with the CCC. Therefore, this consent does not extend to other kinds of copying, such as copying for general distribution, for advertising or promotional purposes, for creating new collective works., or for resale; nor to the reprinting of figures, tables, and text ex= cerpts. 0038-531 X/ 84 $ 8:50 Conwltants Bureau journals appear about siz months after the ,publication of the original Russian issue. For bibliographic accuracy, the English issue published'by Consultants Bureau carries the same number and date. as the original Russian from which it was translated. For example, a Russian issue published in December will appear in a Consultants Bureau English translation about the following June, but the translation issue will carry the December date. When ordering any volume or particu- lar issue of a Consultants Bureau journal, please specify the date and, where appli- cable, the volume and issue numbers of the original Russian. The material you will receive will be a translation of that Russian volume or issue. Subscription (2 volumes per year') Vols. 56 & 57: $560 (domestic), $621 (foreign) Single Issue: S 100 Vols. 58 & 59: $645 (domestic), $715 (foreign) Single Article: $8.50 ' CONSULTANTS BUREAU, NEW YORK AND LONDON 6 lJ 233 Spring Street New York, New York 10013 Published monthly. 'Second-class postage paid at Jamaica, New York 11431. Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 SOVIET ATOMIC ENERGY A translation of Atomnaya Energiya Volume 57, Number 6 December, 1984 CONTENT8 Engl./Rues. ARTICLES ' Major Safety Provisions in Nuclear-Powered Ships - N. S. Khlopkin, O. B. Samoilov, V. M. Belyaev, A. M. Dubrovin, ~. M. Mel'nikov, and B. G. Pologikh ... . . .. .... .... . .... .... .. .. .. .... .... .. 803 379 Tests on Improved Steam Separators in the Third Unit at the Chernobyl Nuclear Power Station-O. Yu. Novosel'skii, V. B. Karasev, E. V..Sakovich, M. A.Lyutov,andV.I.An'kov ?. ~ 807 382 Peculiarities of the Distribution of Phases in the Updraft Section of a Housed Boiling Reactor - V . N. Fedulin, G. G. Bartolomei, V. A. Solodkii, and V. E. Shmelev ... .. .. .. .. .. .. .. .. .. .. .. .. .. ......... ....... 811 385 Effects of Steam Generator Sectioning on the Reliability of a Nuclear Power Station Containing a Fast Reactor - A. I. Klemin, O. B. Samoilov, and E. V . Frolov ..... .... .. ...... ...... ...... .. .... .......... .. 815 388 Statistical Analysis of Reactor Thermal Power by the Use of Thermal and Radiation Methods in the First Unit at the Armenian Nuclear Power Station - F. D. Barzali, L. N. Bogachek, V. V. Lysenko, A. M. Muradyan, A. I. Musorin, A.I. Rymarenko, I. V. Sokolova, and S. G. Tsypin .. .. .. .. . . .............. 821 393 Test Stand for Research on the Physics of High-Temperature Gas-Cooled Reactors - A. M. Bogomolov, V. A. Zavorokhin, A. S: Kaminskii, S.V. Loboda, A. D. Molodtsov, V. V. Paramonov, V. M. Talyzin, and A. V. Cherepanov . ? . ? ..... 825 397 Study of Model Coils Made of Suprerconductor Intended for the Winding of the T-15 Tokamak - I. O. Anashkin, E. Yu. Klimenko, S. A. Lelekhov, N. N. Martovetskii, S. I. Novikov, A. A. Pekhterev, and I. A. Posadskii ................ ....... 830 401 Oscillations in the Concentration of Artificial Radionuclides in the Waters of the Baltic and North Seas. in 197?-1982 - D. B. Styro, G.I. Kadzhene, I. V. Kleiza, and M. V . Lukinskene ? ............................... ....... 835 405 REVIEWS Enhancement of Heat Transfer in the RBMK and RBMKP - A. I. Emel'yanov, F. T. Kaman'shchikov, Yu. M. Nikitin, V. P. Smirnov, and V. N. Smolin ? ~ 839 408 LETTERS TO THE EDITOR Stresses in Spherical Fuel Elements of aHigh-Temperature Gas-Cooled Reactor (VTGR) as a Result of the Heat Load and Radiation Shrinkage of Graphite - V. S. Egorov, V. S. Eremeev, and E. A. Ivanova ...................... ... .... ?'? 846 415 Development of an Oil-Free Forevacuum Unit for an Operating Pressure of 150-300 Pa for the Tokamak-15 - I. A. Raizman, V. A. Pirogav, L. G. Reitsman, E. A.Maslennikov,andV.V.Martynenko 850 417 Surface Tension of Molten Mixtures of Fluorides of Lithium, Beryllium, and Thorium - A. A. Klimenkov, M. N. Kurbatov, S. P: Raspopin, and Yu. F. Chervinskii ? 853 419 Experimental Study of the Interaction of Pulsations of the Neutron Flux and the Coolant Flow in a Boiling-Water Reactor - P. A. Leppik ....... ? . ? . ? ? 855 420 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 CONTENTS (continued) Engl./Buss. Control Experiment on Critical Heat Transfer during Water Flow in Pipes -P.L.Kirillov,O.L.Peskov,andN.P.Serdun'?? 858 422 INDEX Author Index, Volumes 56-57, 1984 ....................................... 863 Tables of Contents, Volumes 56-57, 1984 ................................... 869 The Russian press date (podpisano k pechati) of this issue was 11/28/1984. Publication therefore did sot occur prior to this date, but must be assumed to have taken place reasonably soon thereafter. Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 MAJOR SAFETY PROVISIONS IN NUCLEAR- POW.ERED SHIPS N. S. Khlopkin, O.B. Samoilov, V. M. Belyaev, A. M Dubrovin, ~. M Mel'nikov, and B. G. Pologikh Considerable experience has been accumulated in this country on the design, construction, and operation of nuclear-powered civilian ships: the icebreakers Lenin, Leonid Brezhnev, and Sibir'. The nuclear steam plants (NSP) used on these as the main energy source have been found to be highly reliable and safe, and it is desirable to use them in the future not only in icebreakers but also in transport ships for use in ice fields. The Soviet program for building and developing nuclear-powered ships has involved careful attention to safety in ships containing NSP. The experience with the design and operation of nuclear icebreakers in recent years has led to the revision of safety standards for the nuclear ships and correspondingly ship NSP [1, 2] and international guidelines have been developed. If one meets the requirements of these documents, one has a safe basis for future Soviet nuclear-powered ships. Basic Safety Principles. Basic safety problems implied by the features of NSP are related to screening the radiation sources and preventing the escape of radioactive substances, in addition to reliable core cooling preventing hazardous situations by transferring the reactor to a subcritical state, and maintaining it in that state for an appropriate time. The execution of these functions is delegated to independent safety systems. The composition, structure, and mode of operation in these systems are chosen in accordance with the following basic principles: 1) high fault-free operation parameters for all units; 2) the safety system should perform its functions when there is a single failure in any component; 3) a reliable and fast emergency reactor protection system; 4) the conditions should rule out damage to the fuel pins in all modes of operation; 5) prevention of coolant loss from the core in emergencies related to loss of coolant from the first loop; 6) fission products should not be allowed to pass beyond set barriers; and 7) preventing the pressure in the first loop from rising above a permissible level set by the strength in any design emergency. To provide reliable NSP safety, one combines the deterministic method of meeting the standard safety re- quirements with a probability method, in which one considers not only the consequences of emergencies but also the possibilities of their occurrence; quantitative criteria from this method are currently being formulated. Two methods are to be considered as complementary in researching and providing safety. We note that a prob- ability analysis is obligatory in designing nuclear, power stations [3] . Basic Technical Specifications in Safety Provision in Future Nuclear-Powered Ships. -The NSP in nu- clear ships under design or construction usually employ well-tested designs and units, which have been thorough- ly checked out on nuclear icebreakers and which have shown excellent safety and working life: During the manu- facture, installation, and testing of NSP equipment, one uses well- evaluated technology and quality- control methods. Changes and improvements are made in the design of the equipment, particularly to increase the re- liability and the monitoring performance and completeness. Provision is made for ongoing automatic monitor- ing of all technological parameters during operation, including the radiation environment in spaces within the ship, as well as periodic examination and checking of equipment, pipelines, and systems as a whole. These measures together ensure reliable and safe operation. Translated from Atomnaya Energiya, Vol. 57, No. 6, pp. 379-382, December, 1984. Original article sub- mitted August 27, 1984. 0038-531X/84/5706- 0803$08.50 ?1985 Plenum Publishing Corporation 803 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 One of the major means of attaining high reliability parameters in NSP is the provision of equipment backup. The reactor core has a negative temperature coefficient of reactivity in the working temperature range, which provides self-regulation and dynamic stability. This self-regulation is a factor that restricts the in- crease in reactor power and thus tends to alleviate the consequences of emergencies associated with the power production rate exceeding the rate of core cooling. The control and protection system provides for multichannel parameter monitoring under working condi- tions. The emergency-protection circuits are galvanically decoupled from the measurement channels. The discrete signals that operate the protective equipment are produced independently in each channel and pass to the effectors along duplicated circuits . The links between systems viathe protection signals are also duplicated in the control system. Built-in monitoring devices at various levels enable one to detect and localize faults in individual units, which can be promptly replaced. The response rates of the electrical and mechanical emergency-protection components provide for shutting down the reactor in any emergency while maintaining permissible temperatures in the fuel pins . Under normal working conditions, with the circulation pumps running, the heat from the core is carried off by the coolant in the first loop passing through the steam generators and onward to contact with the coolant in the second loop. When the reactor is shut down, the residual heat is removed from the core by two indepen- dent cooling channels, which are activated automatically or remotely from the emergency protection signal. The first is the cooling channel for the water in the second loop passing through the steam generators when the circulation pumps in the first loop are working together with the pumps in the second loop. The second is the cooling channel with water in the third loop, which operates when it is impossible to supply feedwater to the steam generator. The emergency feed pumps in the second loop with their pipes and equipment form an independent cooling system, which is not dependent on the main condensate-feed system, and this provides ways of cooling the re- actor in any design emergency situation. The users of the NSP have reliable uninterrupted power supplies under all normal and emergency condi- tions, in accordance with the requirements of [1]. Experience with the power plants in nuclear icebreakers in- dicates that complete current failure in the NSP needs to be envisaged only for the time required for the emer- gency power supplies to be activated. On the other hand, calculations show that the systems remain viable with much longer current failures, in spite of the severe working conditions assumed for emergencies. It is to be considered only as a purely hypothetical situation that the NSP fails to provide power for a prolonged period (i.e., loss of the main power supply and failure to obtain a supply from the backup and emergency sources), as there is multiple backup in the electrical power system. Nevertheless, this is considered in designing NSP. The main task in that case is to provide ongoing reactor cooling by means of facilities independent of the state of the ship's electrical power supply. Foreign standardization documents dealing with safety in nuclear ships envisage installing safety valves in the first loop as a passive means of preventing overpressure. However, this is not an obvious solution, be- cause of features of the first loop: It is a powerful source of radioactivity, and there is the potential hazard that the core will be deprived of coolant and thereby damaged, with its serious radiation consequences, while. in addition there is the low reliability of such safety devices. The available experience with them confirms this viewpoint. One possible reason for pressure rise in the first loop is that the coolant acquires excess heat, and this makes it the most effective approach to provide high reliability in the facilities for shutting down the reac- tor and cooling the NSP, which should radically prevent impermissible pressure increase and eliminate the need to fit safety valves [4] . This is shown to be correct by prolonged experience in the accident-free operation of NSP in nuclear icebreakers. New designs envisage automatic connection of an independent system to the steam generator, which con- sists of tanks containing water, a gas cylinder, pipelines, and other equipment designed to supply feedwater to the steam generator and to discharge the steam to the atmosphere without resort to external power supplies, in order to provide emergency cooling in the first loop when there is complete power failure (or failure of all the cooling plant). The heat is removed from the core by the natural circulation of the coolant in the first loop. These cooling provisions restrict the pressure increase in the first loop under all normal and emergency conditions and ensure that it retains its integrity. Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 Fig. 1. Scheme for supplying the first loop and for providing emergency flushing to the core in a container transport ship: 1) reactor; 2) flushing pumps; 3) reserve tank;. 4, 5) feedwater tank and pumps, correspondingly; 6) emergency flush- ing valves. u e~ F 5 200 - 4 n, 10 I'Is .\ ~ - Core .~ P i i ~ S00 1000 ~ sec Fig. 2. Change in coolant parameters when the first loop fails; solid line and dot-dash line) pipe- line failure in the feed and compensation systems, correspondingly; P) pressure in reactor; Vr) coolant volume in reactor; Tsh) maximum tem- perature of fuel-.pin sheaths. One of the most serious emergencies is a leak in the first loop. To reduce the probability of this occur- ringandtorestrict the consequences, there are special design measures such as minimal length in the pipeline in the first loop and the pipelines connected to the reactor above the core. In new reactor systems such as for transport ships, there is provision for constriction nozzles in the reactor pipes. The systems adjoining the first loop are also fitted with a double pressurizing system. All the equipment and pipes in the first loop are located in a hermetic container. In accordance with the standardization documentation, it is conservatively as- sumed in considering emergencies involving leakage in the first loop that there is instantaneous failure in some pipe connecting the body of the reactor to the first-loop equipment. It should be noted that the estimate of the probability is less than 10-s event/reactor ? yr for an emergency involving failure in the structures within the containment .(reactor, steam generator, and power pipes in the reactor), which represents a practical impossi- bility, and the consequences are not examined. The following are the major operational measures to be taken when the coolant escapes. from the first loop: Shutting down the reactor, identifying and localizing the leak, feeding the loop and flushing the core, and` emergency cooling. The signal indicating pressure drop in the first loop below the permissible value leads to the reactorbe- ing shut down automatically, and at the same time the purification system is shut off along with the working Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 groups of cylinders in the pressure-compensating system, and the backup pumps are started up along with the backup and emergency feed pumps, and the equipment in the flushing system is activated. When the pressure in the first loop falls to the pressure provided by the flushing pumps, controlled nonreturn valves on the flush- ing pipes open, and water begins to enter the reactor through two channels (Fig. 1). All these automatically executed operations can be duplicated remotely from the control panel. The flushing water is supplied via two independent branches, while the backup in the pumps and other equipment enables one to meet the above failure principles in the safety system. If the pipe fails in a part that cannot be shut off, one cannot localize the leak, and the main task in that case is to ensure that the core is supplied with coolant in order to provide acceptable temperature conditions in the fuel pins. The output of the flushing pumps is chosen accordingly. As one provides temperatures such that the fuel pins remain viable, this prevents failure in the sheaths, which represent the first barrier to the escape of fission products, while the cooled core geometry is main- tained, and thus any change in the state of the fuel-pin cores is ruled out such that the radionuclides contained in them are rapidly released. If there is a leak in the first loop, which is the second barrier to the fission products, the protective con- tainment comes into action (third barrier). To prevent the pressure in the containment rising to the permissi- ble limit, provision is made for passing the steam-air mixture into a special ballast volume through a bubble tank or into a tank within the shielding. A study of the consequences of leaks shows that the core remains flushed with water if there'is a failure in any pipe in the first loop if the above protection facilities are fitted, and the fuel pins do not attain a critical state, while the pressure in the containment remains below the design value (Fig. 2). This conception of nuclear ship safety is based on experience accumulated in the construction of power plants for nuclear icebreakers and the operating experience with these of total extent about 80 reactor ears. This experience has been utilized in drawing up standardization documents and in designing new nuclear ships. 1. Rules for Classifying and Constructing Nuclear Ships [in Russian], Leningrad (1982). 2. Nuclear Safety Rules for Ship-Borne Nuclear Power Systems (PBYa-08-81) [in Russian], Atomizdat, Moscow (]981). 3. General Safety Provisions for Nuclear Power Stations daring Design, Constr_ uction, and Operation (OPB- 82) [in Russian], Moscow (1982). 4. F. M. Mitenkov, L. A. Zvereva, B. I. Motorov, ~. M. Mel'nikov, and O. B. Samoilov, "The use of asafety valve in the first loop of a nuclear power system containing a VVER reactor," At. Energ., 50, No. 5, 308- 310 (1981). Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 TESTS ON IMPROVED STEAM SEPARATORS IN THE THIRD UNIT AT THE CHERNOBYL NUCLEAR POWER STATION O. Yu. Novosel'skii, V . B. Karasev, E. V. Sakovich, M. A. Lyutov, and V . I. An'kov Improved separator drums (Fig. 1) have been tested at the Chernobyl nuclear power station with its RBMK-1000 reactors, these differing from those used at Leningrad power station and the first units at Kursk and Chernobyl power stations [1] in having a larger internal diameter (2600 instead of 2300 mm) and a different design for the internal devices (IIID). Features of the new ID design include the drainage diffusors 1 and the base of-the compartments 2, the inclined perforated sheets 3, and the breather tubes 4 for taking off the steam from the space 5 between the compartments into the steam volume 6. Also, the separator characteristics have been improved by increasing the distance between the immersed perforated sheet 7 and the upper perforated shield 8. The increased diameter has reduced the mean speed at the evaporation surface by about 7.5%? The drainage diffusors enable one to use the water in the segments in emergency and transitional states. The in- clined perforated sheet of section 12.5% equalizes the velocity pattern at the tapoff from the segments into the spaces between them, which tends to increase the water retained in the separator. The design of the drainage channel in the immersed perforated sheet (unit A in Fig. 1) incorporates the advantage from modifyingthe ID in the 2300-mm-diameter separator [2]: The side piece 9 has a height of 170 mm, while the height of the drainage slot is not more than 55 mm. At the surface of the sheet there are stif- fening ribs 10 with holes of height 100 mm and working section 18%. The. immersed .perforated sheet in the improved separator is fitted with the end pieces 1 ] of height 485 mm. The brackets 12 in the segments have windows allowing the steam-water mixture to flow along the drum and which.thus~ equalize the load along the length. These modifications were incorporated into the design when the separator drum of diameter 2300 mm constituted the bottleneck in the forced circulation loop in the RBMK- 1.000. The improved drum has four rows of steam~vater tubes 13, whose diameter has been increased to 100 mm . Fig. 1. Cross section of the separator drum in the third unit at Chernobyl nuclear power station. Translated from Atomnaya Energiya, .Vol. 57, No. 6, pp: 382- 385, December, 1984. Original article sub- mitted March 26, 1984. 0038-531X/84/5706- 0807$08.50 ?1985 Plenum Publishing Corporation Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 Fig. 2. Location of separator drums, steam pipes, balancing vessles, and steam sampling points: 1) pipes supplying steam to the machine hall; 2) steam collectors; 3) steam tapoff pipes in separator drum; 4) level gauges with balancing-vessel baseline of 630 mm to measure water level above immersed perforated sheet; 5) end level gauges with 1600-mm baseline of balancing vessels; 6) steam sampling lines in steam-collecting tubes; 7)' steam sampling from steam pipe of diameter 600 mm; and a, c,and b) end and central steam-collecting tubes, corre- spondingly. This third unit also employs a new assembly for the main equipment. In particular, the .separator drums are turned through 90? and are set parallel to the machine bay (Fig. 2). The steam-pipe -layout has also been altered. In the new layout, two collectors of diameter 400 mm from adjacent separator drums are combined in one steam pipe of diameter 600 mm, which goes to the turbine. The steam from the collectors is taken from the middle of the separator drum, which produces a more uniform distribution of the steam flow over the steam tubes. Anew system is also used to measure the water level. All the balancing vessels are set up in the staffed (cold) location, and use is made of vessels with partially heated central chambers as devised by the Dzerzhinskii All-Union Thermal Engineering Institute. In each separator drum, one measures the water level above the immersed perforated sheet in three sections along the length in vessels with a baseline of 630 mm and the general level at the end in vessels with a baseline of 1600 mm. During the installation and the runup to the nominal power, we commissioned and checked the level-mea- suring system, which involved checking that the balancing vessels had been correctly manufactured and in- stalled, that the connecting and pulse lines had been correctly installed, checking the blowers, and checking the readings of the level gauges with a pressure difference OP = 0. The readings of the gauges were checked be- fore commissioning the unitwith a water temperature of 20?C with the level in the separator drum varying in the range from -350 to +400 mm. In the period of runup to the nominal power, the scales of the secondary in- struments ranging from -100 to +315 mm inthe630-mm-baseline gauges were replaced by scales of from -200 to +315 mm. The 0 level on a secondary instrument in that case corresponds to a mass level on the im- mersed perforated sheet of 150 mm. To test the separators, steam samples were taken from the steam pipes and tapoffs in accordance with the scheme shown in Fig. 2. The water samples were taken from the body of the drum, while the steam. was taken from the central tubes in each separator drum, and also from the twotubes in the BS 11 separator drum. There were also steam sampling points in the machine hall before the turbines, from which one could deter- mine the average water content. However, the cooling system envisaged in the design did not provide iso- kinetic sampling ahead of the turbine with a thermal power of more than 80% of nominal. To determine the average water content of the steam at the exit from all the separator drums at power levels of over 80%, the samples were taken after the turbine condenser. The water content was determined from the ratio of the 24Na concentrations in the condensate and in the circuit water by the method given in [3] . The tests were performed at a pressure of 7 MPa and thermal power levels of 65, 83, 93, and 100% 'of nominal. The purpose of these was to determine the dependence of the water content at the exit from the sepa- Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 Fig. 3 Fig. 4 Fig. 3. Comparison of readings for gauges measuring the mass water level above the immersed perforated sheet at various reactor thermal powers in MW: O) 2100; ~) 2660; D) 2960; Q) 3200; ~) 3350; ?) 3200. Fig. 4. Dependence of the steam wetness 1 - x in the central steam-tapoff tubes in the separator drums on the mass water level above the immersed :perforated sheet (from the middle gauge) at the nominal thermal power: 1) O BS 11; 2) ~ BS 21; 3) O BS 22; 4) Q BS 12; -- central level gauge off scale; ~ water content less than U .019'0 U1 O,OB .o, o~ 0, 04 -zoo 0 200 HgSn, mm Fig. 5. Dependence of the average steam wetness after the separator drums on the mass water level above the immersed perforated sheet (from readings of middle gauge in BS 11) as indicated by taking samples in the turbine condensers at the nominal reactor thermal power; ? and O) condensers K 5 andK6, correspondingly. rator on the mass level at a given thermal power (steam production). The tests began at the nominal level, and then the level was raised in steps of 50 mm until the water content of the steam at the exit exceeded 0.19'0 or the upper limit was reached for the level gauges with scales of -200 to +315 mm. The limiting height of the mass level of water above the zero mark at which the upper sampling point was reached for the gauges with scales of -200 to +315 mm was determined by calculation from the formulas of [3], and it was 285 mm for the nominal unit power. We also made tests with a level of -100 mm indicated by the instruments with scales of -200 to +315 mm, i.e., +50 mm above the immersed perforated sheet. When the steady state had been reached (constant power, level in the separator drum, pressure, and flow rate), we established isokinetic flow rates for the steam samples, which were calculated from the measured steam flow rates in: the pipes. These conditions were maintained for 30 min, and during this time we recorded the basic working parameters. The levels in both pairs of separator drums were raised simultaneously in order to measure the average water content after all four. In the pair of drums 11 and 12, the level was main- tained from the gauge installed at the middle of drum 11, while in the other pair 21 and 22 it was maintained from the gauge installed at the end of drum 21 remote from the steam pipes. During the commissioning period, the readings of these gauges were found to be the most reliable. Comparison of the readings (Fig. 3) showed that they agreed well (remained within the accuracy class) before and after commissioning the sets of level gauges in August 1982. Straight lines 1 and 2 in Fig. 3 define the boundaries within which the set of gauges with baseline 630 mm remained within the class. The deviations in the readings of the gauges with 630- and- 1600- mm balancing-vessel baselines did not exceed the absolute error indicated by the accuracy class of various reactor power levels. Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 Fig. 6. Steam wetness distribution in steam tapoff tubes along the length of separator 11 for various values of the thermal power Nt in MW and various mass water levels Hm in mm above the immersed perforated sheet (as indicated by the middle gauge); OO) HM = 0, Nt = 2960; O) 10, 3350; O) 140, 3200; ~) 200, 3200; O) 250, 3200; O) 260, 3200; i wet- ness less than 0.01?Jo; 1-3) sections of separator 11 at steam tapoff tubes a, b, and c, correspondingly. Figure 4 shows that the wetness of the steam in the central tubes in all the separator drums. did not exceed 0.02% up to levels of 200 mm above the nominal, but there was a sharp increase at 250-270 mm. An exception was represented by the separator 12 (curve 4 in Fig. 4), which was evidently due to overestimated readings from the middle gauge in this separator. Similar results were obtained at power levels of 83 and 939'0 of nominal. However, as the power level decreased, the sharp increase in the water content began at higher levels in the separators. With a thermal power of about 65% of nominal, we measured the steam wetness in front of the tur- bines. The measurements agreed with those on the central tubes in the separators. The wetness data obtained at the nominal power by measuring the 24Na activity in samples from the tur- bine condensers (K 5 and K 6) are given in Fig. 5 by reference to the readings of the middle gauge in separator 11. With mass levels above the immersed perforated sheet of from -100 to +200 mm, the wetness was less than 0.02%. When the mass level rose above +200 mm, the average wetness began to increase, and it attained 0.1% at +255 mm. Comparison of Figs. 4 and 5 indicates that the limiting level corresponding to a wetness of 0.1% on mea- surements in the central tubes is higher than that from the determination of the average wetness, which is ex- plained by the measurements on the three tubes in separator 11, which illustrates the wetness distribution over the length (Fig. 6). For a level less than 200 mm, the wetness in the central or end tubes does not exceed 0.02- 0.030J0a although there is a tendency for the wetness to be higher in tube a. With levels over +200 mm, the wetness in the end tubes is higher than that in the central one, and the difference is by two orders of magnitude with levels between +250 and +260 mm. The reason for the elevated wetness at the ends of the separator when the water level exceeds +200 mm is not due to the nonuniformity in the steam and water loads along the length, since the maximum occurs in the central part of the drum. With the working mass level in the separator, the wetness near the ends does not exceed 0.02%, and consequently this does not make an appreciable contribution to the average wetness ahead of the turbine. An average exit wetness of 0.1% is attained with the level lower than the wetness in. the central tube on account of the elevated wetness at the ends of the separator drum when the level is more than +200 mm as indicated by the instruments with scales from -200 to +315 The test results led us to increase the nominal mass level in the separator by 100 mm (to the +100 mm mark as indicated by the gauges from -200 to +315 mm), which increased the water stock by 7 m3 while keep- ing the steam wetness below 0.19'0. Therefore, the separators of diameter 2600 mm have a working margin as regards limiting permissible wetness. By maintaining the level over the immersed perforated sheet 100 mm higher than the design value, Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 one can increase the water stock in the forced circulation loop by 28 m3 without interfering with increasing the steam output. LITERATURE CITED 1. V .. B. Karasev, Yu. M. Nikitin, O. Yu. Novosel'skii, and E . V .Sakovich, "The performance of steam sepa- rators at power units containing RBMK reactors," At. Energ., 53, No. 2, 70-74 (1982). 2. O. Yu. Novosel'skii, V. B. Karasev, E. V. Sakovich, et al., "E~tperience with operating and modifying separator drums in the first unit at Kursk nuclear power station," in: Nuclear Power Stations [in Rus- sian], Issue 3, ~nergiya, Moscow (1980), pp. 101 105. 3. A. G. Ageev, V. B. Karasev, I. T. Serov, and V. F. Titov, Separation Devices at Nuclear Power Stations [in Russian], Energoizdat, Moscow (1982). PECULIARITIES OF THE DISTRIBUTION OF PHASES IN THE UPDRAFT SECTION OF A HOUSED BOILING REACTOR V . N. Fedulin, G.. G. Bartolomei, UDC 621.039.536.2 V. A. Solodkii, and V. E. Shmelev In order to increase the driving pressure head of the coolant natural circulation loop, one can mount a large-diameter updraft section above the active zone of a boiling reactor. The motion of a steam~.vater mixture in such a section is characterized by a complicated distribution of phases, and consequently the steam content. . Nonuniformity of the energy liberation in the active zone and the velocity field of the phases at the entrance to the updraft section exerts a decisive influence on the structure of the two-phase flow. At present there are no reliable recommendations on the calculation of the true volume steam content in alarge=diameter channel, which is what the updraft section of a boiling reactor is. The value of the average steam content in the updraft section of a VK-50 reactor calculated by the VTI procedure [1] is 4090 higher than the experimental values obtained on the basis of the measurement of the hydrostatic pressure differential. The main cause for this discrepancy consists, in our opinion, of the fact that the existing procedures for calculation of the steam content are based on the data of experimental investigations obtained in channels of small cross section and. at a rather large distance from the entrance to the channel (l/d > 20), i.e., upon the motion of a hy- drodynamically stabilized two-phase flow in which no significant transformation of the phase velocities and the steam content occurs from cross section to cross section in the channel. The updraft section of boiling reactors is characterized by commensurable height and diameter dimen- sions (H/D ~ 1). -This specifies incompleteness of hydrodynamic stabilization of the flow and. makes inadmissi- ble the use for calculation of the true volume steam content in them of the same relationships as for hydrody- namically stabilized flows. It is necessary for the improvement of the methods for calculation of the volume steam content in large-diameter channels to use data on the structure of hydrodynamically unstabilized two- phase .flows. This is explained by the fact that one-dimensional computational models cannot take account of all the peculiarities of the spatial distribution of the phases. An investigation of the structure of a two-phase flow in alarge-diameter updraft section (D = 2 m, H = 3 m) using the electroprobing method was carried out on a VK-50 boiling reactor. The structure of the reactor, its operating regimes, and the layout of the placement of local steam content sensors have been described in [2, 3], and the procedure for analysis of the experimental results i.s discussed in [4]. The following peculiarities in the structure of a two-phase flow and in the hydrodynamics of the updraft section have been revealed in the analysis of the results: 1. The radial distribution of the steam content cp at the entrance to the updraft section is determined by the complex influence of the energy liberation q and the coolant flow rate W across the .radius of the active zone Translated from Atomnaya Energiya, Vol. 57, No. 6, pp. 385-388, December, 1984. Original article sub- mitted March 11, 1984. 0038-531X/84/5706-0811$08.50 ?1985 Plenum Publishing Corporation Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 zo t~ ~ 0,2 0,6 4~ Mw 2s ~-o.zs~ Fig. 1. Distribution of the energy liberation (O), steam content (O), and coolant velocity (^) across the radius of the active zone for a pressure of 2.0 MPa in the reactor and a thermal power of 80 MW. i ~ i 7,0 74 7B Fig. 2. Peculiarities of the s-team content distribution with height of the updraft sec- tion for (a) p = 6.0 MPa and N = 170 MW and (b) p = 3.2 MPa and N = 150 MW: O, x, ^, and O) R =0.92, 0.74, 0.56, and 0.37 m, respectively. R and, in connection withthis,is of a sharply expressed nonuniform nature (Fig. 1). In particular, the steam content at the exit of the heat-generating assemblies (HGA) depends significantly on the position of the com- pensating rods (CR) in the active zone. Upon a rise in the reactor power, when the peripheral CR are essential- ly extracted, an increased steam content is: noted above the HGA of the peripheral part of the active zone. The maximum of the steam content is shifted towards the center upon the subsequent extraction of the central CR. 2. The structure of the two-phase flow in the lower part of the updraft section, which is adjacent to the active zone, is determined to an appreciable extent by the jet nature of the coolant outflow from the HGA heads. To a large extent the jet nature of the outflow appears above the central HGA, and to a lesser extent above the peripheral HGA. The height of the jets depends on the velocity of the exiting two-phase flow, the working pres- sure p, and thehydrodyna.micsof the updraft section. For the VK-50 reactor the height of the jets does not ex- ceed 0.4 m. The jet nature of the outflow of the two-phase flow is exhibited in: a corresponding variation of the steam content above the HGA heads. A maximum steam content is noted on axial lines of the jets. In power regimes (p = 6.0 MPa) with a supply of feed water to the "cold"-drop section (i.e., under the central part of the active zone) the following effect of jet outflow of coolant from the central HGA has been noted: The steam content in the jet is higher at a distance of 280 mm from the HGA heads than at adistance of 80 mm (Fig. 2a). This ef- fect is evidently explained by a "pressing down" of the steam-water flow, an increase in the rate of its move- ment, and a migration.of the steam of its periphery to the jet axis. The jet effect disappears at a large height, which indicates a local equalization of the steam content. At a low reactor power (< 100 MW), a lowered working pressure, and a change in the scheme of the feed- water supply the steam content in the jet at a distance of 280 mm from the HGA does not increase-(Fig. 2b). This can be explained by a decrease in the velocity of the steam-water flow at the exit from the HGA and a breakdown of the jets at a lower level. This effect has not been detected in all regimes in the peripheral part of the updraft section, which may be caused by a deflection of the steam-water jets toward the center due to recirculation flows. Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 ~i ~o ~s n, ~s g ~s > ~s z, 2s y, Fig. 3. Distribution of the steam content with height in the updraft section: O, O, ^) p = 6.5 MPaandN=180 MW; ~)p=6.4MPaandN= 181 MW; ~) p = 6.4 MPa and N = 180 MW; ~, ?, ^)p= 6.0 MPa and N= 170 MW; O,~)R= 0.74 ~m; ^,~)R= 0.37 m; and O, ?,~,~)R=0.92 m. m. 3. On the whole, the steam content distribution over the volume of the updraft section appears as follows. Near the axial line of the updraft section a .coolant flow saturated with steam is formed with the maximum steam content at a distance of about 1-1.5 m from the active zone. The steam content decreases towards the output cross section of the updraft section. The steam content on the periphery of the updraft section first decreases sharply with height, reaching a minimum value at a height of 0.8 m, and then it gradually increases (Fig. 3). Complete equalization of the steam content in the output cross section of the updraft section is not attained [4] . 4. It has been established on the basis of hydrodynamic measurements that the pressure differential on the periphery of the updraft section is higher than in its central region. In particular, the following values of the pressure differentials are obtained at a power of 168 MW with a pressure of 6.0 MPa: ' on the periphery, .12.8 kPa, and in `the center; 11.22 kPa. Consequently, a definite radial pressure gradient exists in the updraft section:~which acts on the corresponding radial flows of the coolant.- Taking into account that in the output cross section of the updraft section equalization of the steam con- , tent (although incomplete) has been noted, one can assume in the first. approximation that the static pressure on this cross section is constant. Then one can conclude that the maximum radial pressure gradient is observed in the lower part of,the updraft section. An intensive migration of steam from the- periphery of the updraft section to its central part has been noted precisely in this zone. 5. The 'values of the steam content averaged over the height of the updraft section which are calculated on the basis of the measured pressure differentials are related to each other and the calculated value as fol- lows: where ~Pc and Wp are the experimental values of the steam content in the central and peripheral regions of the updraft section and ~calc is the average value of the steam content calculated by the procedure of [1]. This re- sult shows that the volume steam content in the updraft section above the central HGA is close to the calculated value and decreases significantly towards the periphery. 6. The nonuniformity of the steam content distribution in the updraft section determines the effectiveness of the individual updraft tubes mounted on the HGA heads. The isolated individual updraft tubes on the VK-50 reactor (height, 1.5 m, diameter, 150 mm) are mounted above different rows of HGA of the active zone. The investigations have shown that such mounting increases the coolant circulation velocity in the fifth (the peri- phery) row of HGA by 0.2 m/sec and in the fourth row by 0.1 m/sec. No increase has been noted in the coolant circulation velocity in the third row of HGA, which is determined by the rather high steam content above the HGA in question without installation of an individual updraft tube. The experimental data obtained permit suggesting the following qualitative physical model of two-phase flow in the updraft section (Fig. 4). A jet outflow of a steam-water mixture exists at the exit from the HGA of the active zone. Due to the enhanced output .of steam from the central HGA and the existence of a radial pres- sure gradient the steam-water mixture from the periphery of the updraft section is directed into its central region. In this connection an accelerated steam-water flow with an enhanced steam content is formed in this Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 Fig. 4. Scheme of the flow of the steam- water mix- ture in the updraft section of the VK-50 reactor. region. Upon the attainment of a specified height a gradual widening of the steam-water flow begins, which leads to equalization of the steam content over the cross section of the updraft section. A recirculation flow of coolant which deflects the steam-water jets from the peripheral HGA to the center is formed on the periphery of the updraft section in the zone of low steam contents. The proposed physical model of the flow of a steam-water mixture and the experimental data obtained on the distribution of the steam content permit calculating more accuratelythe hydrodynamics of the updraft section of boiling reactors. LITERATURE CITED 1. A.Ya.KramerovandYa.V.Shevelev,EngineeringCalculationsofNuclear Reactors [in Russian], Atomizdat, Moscow (1964), pp. 320-322. 2. G. G. Bartolomei, V. A. Solodkii, V. N. Fedulin, et al., "Determination of the steam content in boiling reactors by the electroprobe method," Teploenergetika, No. 10, 15 (1980). 3. V. A. Solodkii, G. G. Bartolomei, V. N. Fedulin, et al., "An investigation of the hydrodynamics of an adia- batic two-phase flow in the updraft section of a VK-50 reactor," Teploenergetika, No. 6, 73 (1981). 4. V. N. Fedulin, V. E. Shmelev, V. A. Solodkii, and G. G. Bartolomei, "The steam content of a two-phase flow in the updraft section of a VK-50 boiling reactor," Preprint NIIAR-33 (598), Dmitrovgrad (1983). Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 EFFECTS OF STEAM GENERATOR SECTIONING ON THE RELIABILITY OF A NUCLEAR POWER STATION CONTAINING A FAST REACTOR A. I. Klemin, O. B. Samoilov, and E . V . Frolov One must provide adequate reliability in the individual units and in the station as a whole in order to meet current specifications for the economic performance and safety in nuclear power stations. In the case of sodi- um-cooled fast reactors, a current problem is to improve the reliability in the steam generators (SG) of sodi- um-watertype. Various Soviet and foreign designs for such reactors use sectional SG, which raise the reli- ability and safety of the station because they localize possible faults caused by leaks in the SG heat-exchanger tubes to a single steam-generating section [1] .The SG sectioning makes it necessary to incorporate factors in- to the reliability model such as the SG repair strategy, the number of steam- generating sections in one cooling loop, and the working constraints on the available* unit power. We give below a model that has been used to ex- amine the effects of these factors on the reliability of afast-reactor system. Model for the Structural Reliability of a Nuclear Power Station Unit Containing a Sectional SG. We con- sider asimplified structural scheme for the unit with one turbine and a fast reactor having m cooling loops; each of which includes pumps in the first and second loops, an intermediate heat exchanger, and a sectional SG (Fig: 1). The latter-consists of a large number of components, including tube bundles, module bodies, steam- water chambers, sodium, steam, and water pipes, and cutoff equipment for the sodium and water (steam). The reliabilitycalculationsarebased on two types of enlarged element in the sectional SG (Fig. 1): the element 5, which includes components whose failure makes it necessary to disconnect the entire SG and run down the cor- responding cooling loop (for example, the sodium-handling gear), and element 6 (steam-generating section), which combines components whose failure leads to the disconnection of a separate section and corresponding reduction in the power of the given SG (tube bundle, module body, and seals on the covers on the steam-water chambers). Clearly, each of the elements 5 and 6 consists of corresponding SG components in series connection in the sense of reliability. The reliability of the unit is most fully characterized by the power availability factor gip, which is calcu- lated from the following formula for the basic mode of operation: ~W ~= 1 ~ Na (i) dT, (1) t ~? 0 where Na(T) is the mean available power from the unit (as a fraction of the nominal power) at times TEty? while tv, is the relevant working period. A basic stage in constructing a model for power station reliability is to determine the dependence of Na . on the state y, which in turn is dependent on the states of various items. The basic principles have been con- sidered in [2]; according to which Na(y) is defined by means of special characteristics that are functions of the states of the individual units or sets of equipment .(for example, cooling loop), which are subsequently called component parts. The state function for a component part characterizes the contribution to the station power (as a fraction of the nominal power) in a given state of the component part when the rest of the power station unit is completely free from faults. The state of element l (equipment unit) is defined by means of the binary variable x~, which takes the value 0 in fault free operation and 1 when the element fails. The state functions gl(xl), l = 1, 7 for the equipment in this station are defined as follows: *The maximum power that the unit can provide with all items fully viable. Translated from Atomnaya Energiya, Vol. 57, No. 6, pp? 388-393, December, 1984. Original article sub- mitted August 9, 1982; revision submitted November 15, ] 983. 0038-531X/84/5706- 0815$08.50 ?1985 Plenum Publishing Corporation 815 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 Fig. 1. Simplified structural scheme for a unit in a nuclear power station containing a fast reactor; 1) core; 2) first-circuit pump; 3) intermediate heat exchanger; 4) second- circuit pump; 5) en- larged element including the SG sodium equip- ment; 6) SG section; 7) turbine generator. 1-xl ~It(xl)= m (1-xi) l m \ 1 nl / R where ns is the number of steam-generator sections in a cooling loop. The values of the subscript l of 1 and 7 correspond to the core and the turbine generator, while the values of l of Z-6 correspond to the pump in the first circuit, the intermediate heat exchanger, the pump in the second circuit, and the enlarged elements of the SG, correspondingly. As the sections are indistinguishable from the viewpoint of reliability (the elements 6), the state function qs(j) for a sectional SG and a separate cooling loop are defined by plc (1) _ ~ m l ns l O+ if 1=7cr+ ne+ where j is the number of failed sections in the SG for a cooling loop and jcr is the critical number of SG sec- tions whose failure causes the operation of the corresponding loop to be halted. The state function for cooling loop j (i = 1, m) takes the following form because of the serial structure: II1t = Illlil lI]9 ('1'^)+ ~13 (~'a)+ Q4 (xa)+ Q5 (xr,)+ qR ()i)}. We use the state functions for the equipment in the cooling loop to represent the state vector for a unit in enlarged form: If the reactor is operated with SG sections switched out (for overhaul), there may be differences in sodi- um temperature in the first circuit at the entry to the core resulting from the loops with different numbers of fault-free (or failed) SG sections. When certain limiting values of the temperature differences are reached, one usually has to impose constraints on the available power. To characterize such working constraints, we introduce the parameter o(y) for the state vector y, which is equal to the maximum difference between the power values (in relative units) for any two cooling loops: ~(!1)=nl{ nias ~V',l- nli^ [il"z]}, ~;t>U. (6) We note that in (6) we envisage loops for which the values of ~i are different from zero, since otherwise the corresponding loop is taken out of service. The constraints on the available power are determined by the limiting permissible value [~ of o(y), which canbe estimated as follows: aR OR= n ~ where do is the maximum permissible difference in the numbers of fault-free (failed) SG sections in any two operating loops. The value of do is defined from heat-engineering and strength calculations. Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 In accordance with (5), the desired Na(y) for the unit is where Q is the state function for the system of m cooling loops: a~ =min {~~, ~b-I- e? 1 ~; ~,~ b=min [~t1; i=1, na, 11>z>0. Expression (9) -has been derived on the assumption that if there is a difference in sodium temperatures in the first circuit between any two cooling loops that exceeds the limiting permissible value, that difference is reduced to the permissible value by switching out the necessary number of sections in the loop with the higher power in order to provide the maximum unit power. We now calculate the probabilities of realizing various states in the unit. We assume that planned pre- ventative measures applied to the equipment are combined with reloading the reactor and form a periodic sequence, where the instants at which they are performed are points at which the equipment viability is com- pletely restored. Also, we restrict consideration to the integral (tank) style for the equipment in the first cir- cuit, which is very characteristic of fast reactors [1], where unplanned (emergency) repairs to the first-cir- cuit pump and the intermediate heat exchanger make it essential to shut down the reactor. We subsequently as- sume that the equipment in the first circuit (pumps and intermediate heat exchanger) is not repairable in the in- terval [0, T] between planned preventative. repairs, while unplanned repairs are combined with the next planned ones: Parts of 'the equipment that are. repairable in the interval between.planned" preventative repairs include the ?core, the pumps in the second circuit, the SG elements, and the turbine generator. The probability that ele- ment l? (Fig: 1) is fault=free at time T is given by the standard formula in the case of exponential distributions for the time to failure and the recovery time: Pr (ti) - Ni~-~i {?~ -~ die-~~+?~~s}, ti E I0, T ], (l = 1, 7) (10) where ~l and ?l are the rates of failure and recovery for element 1, correspondingly. We note that (10) applies also when the equipment is not repairable, when ?l = 0. On the basis of the above assumptions about the repairs, (1) becomes" T ~ ' L T , Na(ti)dti,(1-b), (11) n where T is the interval between planned preventative repairs and 8 is the mean proportion of station shutdowns related to planned preventative repairs and' unplanned repairs combined with them for equipment in the first circuit not repairable in the interval [0, T] (pumps and intermediate heat exchanger). We consider two types of repair strategy for a sectional SG in an individual loop: 1) strategy Cl, in-which an SG section that' fails is disconnected and repaired while the other sections operate; and 2) strategy C2, where when a certain (critical) number of sections fail one switches out the correspond- ing cooling loop in order to repair the SG. .Within the framework of strategies C1 and Cz there is also the particular case where the repair of failed SG sections is combined with. the next planned preventative repair to the unit. Let gj(T) be the probability that at time TE(0, T] the sectional SG in a loop has j failed sections. For re= pair strategy C1, the probability gj(r) is ~; (i) = C;,l.~s.4_; (i)11 - Pa (i)1', TE [0, Tl, (12) where C~ is a binomial coefficient and Ps(T) is-the probability of a viable state in the steam-generator section ns .at time ,z, which is calculated from (10) . . Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 For strategy C2, the operation of the sectional SG is described by a Markov random determined as the solution of a system of differential equations: dQ~ (T) dei (i)_ dT -ai-i6'i-i(i)-a;g;("r), 1=i+1cr-i; icr,- t ~ r=n where ? is the rate of SG recovery as a whole (with allowance for the time consumed in preparatory operations in disconnecting and connecting the loop), ?j = (ns - j)7~, and 7~ is the failure rate in SG sections. According to (2)-(4), the state function ~i for loop i takes jcr+1 discrete values 0,...,gs(j), ...,1/m. The probability that ~i is equal to one of these discrete values at an instant TE [0, T] is defined as follows: Pt{Y~t=40(1)}=~ icr-t l 1 -z. ~--o 5 gi (i) l~ Pt (i), if 'fi't = 9s (1) > 0; i=^_ where gj(T) is calculated from (12) and (13) correspondingly for strategies C1 and C2. One way of raising the reliability of a unit having sectional SG is to employ backup steam-generating .sec- tions. Let the SG in each loop have r backup sections, i.e., ns= ns+ r, where ns is the minimum necessary number of working sections in the SG for an individual loop. We assume that the backup sections work in hot backup, and the loop is shut down for repairing the SG when jo = r + 1 sections fail (we envisage only strategy ,CZ). Clearly, the SG state function qs(j) and that for the cooling loop ry take two discrete values: 1/m and 0. The probability of realizing these states for loop i is 6 r 1 =2 7==0 5 r 1=2 7=0 where gj (T) is calculated as the solution to (13) for 1cr=10=r~-1, Formulas (10) and (12)-(15) enable one to calculate the probability of realizing state process, and gj(T) is P: (y) _ ~'i (ti) 1'~ (T) i11i P,, {,pi _ 9s (li)}+ where ji is the number of failed SG sections in loop i, and the probability P~{~i = ?0 (Ii)}is defined by (14). The mean available unit power is !SEE where E is the set of unit states for which Na(y) > 0. Dependence of cp on Various Factors. The .effect of using a sectional SG to improve unit reliability is dependent on various factors: the number of the sections and the reliability characteristics in the SG elements, the type of repair strategy, and the mode constraints on the available unit power. In what follows we consider the effects of these factors on the power use factor rP for a unit in a nuclear power station containing a fast re- actor, having four cooling loops with sectional SG. The- calculations are based on the above model, which has been implemented as a computer program.. Table 1 gives the equipment reliability parameters used in the cal- culations. The values have been taken on the basis of operating- experience with the corresponding equipment or data on the reliability of similar equipment (these may be considered as the authors' approximate estimates). The failure rates are taken as constant, which is usually correct for the period of normal operation, when the Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 TABLE 1. Equipment Fault-Free Parameters and Repairability Fault rate, h-t Recoveryt rate, h' First-circuit pump 3,7.1(:- Second-circuit pump 2 2.10-5 Intermediate heat ex- , 1 4.10-~ changer . Steam generator(r~=16): element 5 , 1,8.10-'~ fi 7.10-g element 6 (SG section) 4,5.10-~ , running-in period has ended but the period of rapid aging has not begun. To simplify the calculations, the reliability in the core and turbine generator has been taken as substantially higher than that in the other equip- ment (Table 1) . We estimated the failure rate for element 5 (Fig. 1), which combines elements in the sectional SG whose failure leads to the entire SG and the corresponding loop being shut down, on the assumption that the main con- tribution comes from the sodium-handling equipment (failures of leak type). The failure rate for element 5 was estimated from where y is the number of sodium-handling items per SG section and ~a is the rate of failure of leak type for the sodium-handling- equipment (~a~ 10-7 1/h). We also considered types of SG with different degrees of sectioning (ns = 16 and ns = 4), where we incor- porated the dimensional effect, which is that the fault-free level of a section decreases as the power increases, which isdueto'variations in various factors (heat-transfer surface, number of tubes, and number of welded joints), all of which influence the fault experience in an SG section [3]. We assumed that this scale effect may be approximated by a power function of the form ~r ?~a=fir,(','-.< 1 ~ (19) ns 1 where 7~ and ~ are the- failure rates in an individual section for an SG with the numbers of sections (per loop) ns and ns, correspondingly, while R = const is the parameter of the scale effect. The calculations were per- formed, for R= 1 and R = 0.5. The recovery rate ? for the SG with repair strategy CZ was determined from data on the repairability of SG sections (Tabled) and the mean times involved in the preparatory operations, which was taken as about 70 h. Table 2 gives calculations on cp for a system containing a fast reactor in the interval between reloading operations (planned preventative repair) t~ 3000 h with the proportion of planned shutdown time b ~ 0.127. The values of ~P in Table 2 have been derived for do = ns and ~ = 1, i.e., on the assumption that there are no mode constraints on the available power due to possible temperature differences. It is evident that strategy C1 pro- vides higher values of rp than does C2, which occurs because it most fully uses the advantages of the sectional SG structure. The advantage thereby gained is largely determined by the degree of SG sectioning, as well as the scale effect, and the maximum values occur for an SG with a small number of sections (ns = 4) when reliability is low (R = 1). The calculations also showed that when the section reliability is high (1/7~ > 7? 104 h), the type of SG repair strategy has virtually no effect on cp. It should be noted that implementing strategy C1 involves technical difficulties of the shutoff equipment, which has to enable one to repair a failed section while the other sections of the SG in the loop are operating. Therefore, one usually employs strategy CZ in practice. We considered two particular cases: jcr = 1 and jcr = ns. In the first case, failure in any section leads to the corresponding loop being shut down for SG repair, while in the second case the loop is not shut down when sections fail (one merely shuts down the failed sections) and the repair is combined with the next reactor reloading. Table 2 shows that the latter strategy (jcr = ns) is preferable for an SG with a large number of sections (ns = 16) since it provides a substantial increase (about 109'x) in the equipment reliability parameter. We note that this repair strategy, with failed sections shut down, is possible if a certain specification for the level of repairability in the SG sections is met: .The mean.. . section repair (replacement) time should not exceed the reactor reloading time. Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 TABLE 2. Values of cp in Relation to Repair Strategy and Degree of SG Sectioning Steam gen- erator form R=0,5 tta=4 R=1 0,710 I n,71`6 11,644 _ U,62Ei TABLE 3. Values of Win Relation to Mode Constraints Steam ggen- ~ I do-,~F erator {orm I `I0 - I 1 `'u = ~ I ~ ~o = s cno - 1) ns = 1('i I (1,73 4 ( 11, 74fi I 1),7511 I (1,751 =4 N=11 n ' 5 11,(iA 1 I 11, 7114 I (1,7116 I 11,7116 y R=1 U,:~f3 5 0, 620 1,,626 11,626 Two factors determine the effects of SG sectioning on cp: the scale effect for the reliability parameters of a section, which is due to the change in unit power, and the scheme effect, which is related to the change in the number of SG sections in the loop. Also, according to (18) any change in the number of SG sections leads to a change in the amount of shutdown equipment in the loop and to a corresponding change in the reliability characteristics of the enlarged element 5 in the SG (Fig. 1). The calculations showed (Table 2) that enlarging the SG sections reduces cp for both repair. strategies. As would be expected, the reduction in cp is more sub- stantial for strategy CZ when there is a substantial scale effect (R = 1). Also, when the number of SG sections is small (ns = 4) and the section reliability is low (R = 1), there isapronounced optimum for gyp, which is at- tained if one switches out a loop for SG repair when two sections fail (jcr = 2), Clearly, this optimum is de- pendent on the degree of SG sectioning, the fault experience level,. and the repairability, as well as on the value of the interval [0, t] between reactor reloads (planned preventative repair), and the value in each particular case can be determined from a more detailed analysis. We now consider the dependence of cp on the available-power constraints due to the possibility of the above temperature differences. The calculations were performed for wide ranges in do and ~, which determine these constraints. The limiting value do = ns (moo = 1) corresponds to the case where there are no constraints of this type, while do = 1 corresponds to the case where the latter are considerable. The effects of the con- straints on cp are slight for strategy C1. Table 3 gives calculations for these forms of SG with strategy C2 (jcr = ns) for t = 3 ?103 hand a = 0.127. Table 3 shows that the constraints may 'reduce W if the SG has a small number of sections because these constraints begin to make themselves felt as shutdown states with small numbers of failed SG sections as the unit power of an SG section increases. Because of the scale effect, the probability of such states is fairly high, and therefore the effects of the constraints (or of do and ~) on cp are most pronounced. This model was also used in analyzing the scope for increasing the reliability of a system containing a fast reactor by using r backup sections in each cooling loop. It was assumed that the backup sections operate in hot backup, while a loop is shut down for repair of the corresponding SG when r + 1 sections fail. For the SG with ns = 16 working sections, we obtained the following values of cp (t = 3.103 h, S = 0.127): 0,659, if r = 0; W = 0.742, if r = 1; 0,773, if r=L, where the case r = 0 corresponds to strategy CZ (jcr = 1)? -These results showed that backup sections in the SG provide an effective means of raising the system reliability parameters. The largest cp is attained when the number of backup sections is small (r = 1 or 2).. A real SG design always has some spare power margin, which canbe considered as hot backup in relation to the Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 c1 trategy 0,789 0,774 n,72s f'I,fi60 0,724 Q 675  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 ~ ;~-__ nominal SG power. Therefore, the above advantages due to section backup can be realized in practice for ex- ample as follows: When one SG section fails, the loop continues to operate at its nominal power because the power levels in the fault-free sections are raised. ` ~~ LITERATURE CITED 1. O. L. Kazachkovskii et al., At. Energ., 43, No. 5, 343 (1977). 2. R. A. Peskov and E. V. Frolov, Aspects of Nuclear Science and Engineering: Series Nuclear Reactor Physics and Engineering [in Russian], Issue 1 (10) (1980), p. 29. 3. H. Procaccia, Reliability of Nuclear Power Plants, lAEA, Vienna (1975), p. 351. STATISTICAL ANALYSIS 'OF REACTOR THERMAL POWER BY THE USE OF THERMAL AND RADIATION METHODS IN THE FIRST UNIT AT THE ARMENIAN NUCLEAR POWER STATION F. D. Barzali, L. N. Bogachek, V. V. Lysenko, A. M. Muradyan, A. I. Musorin, A. I. Rymarenko, I.V. Sokolova, and S. G. Tsypin Avery important factor in providing safe operation in a nuclear power station unit is to determine the thermal power of the reactor W with an error of 1 2q [1-4]. Operational methods have been proposed [1, 2, 4] for;.measuring the thermal power by statistical computer calculation methods; in [3, 5], operational correlation dispersion methods were considered for determining the thermal power by means of radiation meters (with an. error of 1-2qo at the nominal power), these meters operating with the neutrons leaking from the reactor con- tainment [power meters (PM)] or on the y radiation from 16N in the coolant [loop detector (LD)] . The data en- tering these meters are used in determining discrete quantities of statistical nature, and this enables one to ob- tain given errors.in the readings of the PM (NpM) and LD (NLD). Also., a major feature of a radiation meter is that NpM and N)~ are strictly proportional to W, as has been confirmed by experiment [3]. In view of these advantages, it is of interest to perform a combined statistical analysis of the data ob- tained by these methods and those provided by the standard thermal detectors (TD) in order to define the reac- tor thermal power more accurately and to determine the errors. Table 1 gives data for one-factor variance analysis, which have been obtained with the first unit at the Armenian nuclear power station (the ratio of the thermal reactor power measuredby the'TD and expressed as a percentage of the nominal power to the readings of the PM in count/sec). This approach supplements traditional analysis of measurement methods in enabling one to analyze results corresponding to several levels of thermal power (which may differ substantially). The dots in ~V.j and -Wi, indicate averaging correspondingly with re- spect to the subscripts i and j . The thermal power was determined by the following thermal methods: 1) i = 1, from the sum of the products of the differences of the enthalpies for the hot and cold .parts of the loop by the coolant. flow -rates in the corresponding main circulation loops; 2) i = 2, from the product of -the mean difference in enthalpies. of the hot and cold parts of the loop by the total coolant flow through the reactor; 3) i = 3, from the thermophysical_ coolant parameters in the second circuit beyond the feed pumps; 4) i = 4, from the steam in the steam generators; 5) i = 5, from the feedwater in the steam generators; Translated from Atomnaya Energiya, Vol. 57, No. 6, pp. 393-39~, December, 1984. Original article sub- mitted September 5, 1983. 0038-531.X/84/5706-0821$08.50 ?1985 Plenum Publishing Corporation '821 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 TABLE 1. Initial Data for One-Factor Variance Analysis of the Ratio Wij/NpMj of the Ther- mal Detector Readings on the Reactor Power to the PM Readings Number of measure- ment method and statistical-character- istic estimator for ther- mal power level j W.i= 1 ~ Wii INpMj i_1 Esflmate of the vari- ance $~ ? 103 Weight of mean i 8i~ S~ Nph1 ?, count/sec; W ?, % 7 ~ Mean thermal power for method Varianceesttmator ofineasurementi f :1.,93 70,911 ri9,55 93,13 T w ~~ [ili W~,= or measurement method $i'10S 4s.1;7 cs,7s sc.7z 92,72 N J i=t PMj 0,823f_~ 0,9317 0,9485 0,9787 0,9205 4,6'111 0,8639 (1.9647 0,9749 1,0037 0,9518 3,7075 0,8951 1,0333 1,0172 1,0336 0,9948 4,4796 0,819(1 0,9234 0,9174 0,9685 0,9071 3,970(1 0,7849 0,9306 0,9358 0,9417 0,8982 5,8304 0,9561 0,9942 0,9903 1,0254 0,9915 8,0496.10-1 1,0137 1,0127 0,9949 1,0171 1,0096 9,8724.10-9 0,8794 0,9701 0,9685 0,9955 - - 6,6924 1,9381 1,3776 1,1306 - - 149,4232 515,9692 782,7176 884,4861 - - *Here i is the number of the method of measuring the thermal power, 1 ~i ~ I-7 while j is the thermal power level 1 < j ~ J = 4. 6) i = 6, from the coolant parameters in the second circuit in the turbine control-stage chambers; and 7) i = 7, from the coolant parameters in the second circuit after the high-pressure heaters. The variance analysis is performed with the method of measuriig the thermal power as the factor. The Bartlett test [6] and the Cochran test [Z] enable one to determine whether there is equality in the variances?in the data groups . The Bartlett test is based on comparing the calculated value of XZ with the Xtab distribution having. I - 1 degrees of freedom. In that case, with probability 1 -c? = 0.95, Xtab (I - 1 = 6) = 12.592 > ~: _ 9.338, so the variance series is homogeneous. The Cochran test is based bn considering the quantity u si t=t The value found G~ =max{Gi} =.0.245 is less than the tabulated one Gtab (I = 7, J - 1= 3) = 0.48 for 1 - ~ = 0.95, i.e., the series of variances Si is homogeneous. We test the hypothesis Ho : wi ? = W2. = Ws ? _ ... = W7?? Adopting the hypothesis corresponds to the method of measuring the thermal power not being significant as a factor. The quantities needed to .test the hypothesis are given in Table 2. - We also used the Fisher test [8] ,whose statistic is the ratio F = SB/S$. If F < Fib, the hypothesis is accepted with probability 1 - a, where Ftab is the value of the F distribution with vB and "R degrees of freedom. `Table 2 shows that F < Fts,b, i.e., hypothesis Ho is true with probability 1 - a = 0.95. As the method of measuring the thermal power is not significant as a factor, the overall mean, which is an estimator for the true value of a ratio, is [6] given by W4. i=1 I The estimator for the variance of the overall mean is given by [6] Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 TABLE 2 . Quantities Needed to Test the Hypothesis Ho: Wl? = Wz ? _ ? ? ? = W7 Variance Sum of the squares of Degree of 3 F ratio (Fisher I source the difference I freedom Me I an square, x 10 test) t _ _ J z I ~, (Wi.-iV)z to Sz I; Between J~ (Wi.-W)z=0,0503 vB=I-1=6 B S = =5,3855 F= =2,507 ga levels i=t r yp J R , Within -W 0702 )z=0 ~, ~~ (W v =I7- I=21 \\ ,, i=t , `` i=t =3 46 (6 2!)=2 57 F levels ii i. , i=t i=1 R Sit= , vR , , tab I J (W;;=w)z 'kFor simplicity, the symbol Wij/NpMj is replaced by Wij. 1 ) sw=lJ(iJ-l)L(.1-1) ~ Sep-T~' (W~-iF)2~~ =t i=t J where Si = J 1 1 ~ (W,i - W,.)~; Sw = 7.38.10-5. i=1 Further, we use the same data in aone-factor variance analysis now assuming that the PM reading is the factor, or, which is the same, the reactor thermal power level. We checked 'the variances for equality on the Bartlett and Cochran tests, which showed that with 1 -a = 0..95,_ Xz = 6.3 Gtab (J = 4, I - 1 = 6) = 0.6. In the case of the more searching Cochran test, we assume that the variance series is inhomogeneous. On the Fisher test, the hypothesis Ho: W.1= W+z = W.s = Wa is rejected, i.e., the thermal power level factor is ?ignificants SN = O.U1811, JH . U.U027(i, F=6.SG; Ftab (ti'i~=~3> ve?~~!')=3.01 andF>Ftab? One.-also has to consider whether there is a monotone trend (bias). The check was made via the Abby test [6], which showed that there was no systematic bias. Therefore, as the variance series is inhomogeneous here, the overall mean is calculated from the follow- ing formula [6]: st_. = i I J ~, s~ ~t ./ J -1 v~ nisi -~ LI k'i (W.i - Wa)3J = $.`~(i~)li ? 11)-''. I i =t i=1 Table 1 shows that the data group corresponding to the 45.790 thermal power level has a variance devia- ting from homogeneity. In fact, if we exclude this level, we find that the Bartlett and Cochran tests. are obeyed: Xz = 0.462 < Xt b (J -1 = 2) = 5-.991; Gm~ = 0.45 < Gtab (J = 3; I -1 = 6) = 0.68, and the hypothesis Ho ~ W.t = W?z = W?s is a~opted, since F= 0.12 < Ftab ("B ? 2; vR = 18) = 3.55. As the overall mean we take W = 0.97802 as calculated from (1) with the variance W = -6.9787 ?10-5 de- rived from (2). There is agreement between Wv (for four levels) and W (for three levels) within limits of 0.59'0, which means that any of the methods can be used to determine the overall mean to represent the data. Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 TABLE 3. Initial Data for One-Factor Variance Analysis of the Ratio (NLD j/NPMj) of the Readings of the LD and PM* LD number and statisticalcharac- teristics at ther- malpower level j 5 (i 7 8 Arithmetic mean r N.; _~ ~ nLDij PM i' i=- t ~ Variancezestima- tor On Sj ,s,c; I cs,;a ,r,~s I a~,~s 98,78513 '51,11217!) 21,21145(1 21,i352U4 16,9I12u6 15,415n:S 15,76677 1!),3974 99,67.5165 'L1 ,51Lt79 21,911823 21,7:3/i:iL 18,1!18(17 19,3118!14 1!),32148 19,711fi1i!I 18,4738 1!),11873 18, 79!)37 1!1,4758 16,86981 97,13:117 17,6121114 17,1','5577 21,(144:58 3L,4:;;r7:{ _"2,7!1485 2L,~~7595 98,965112 19,8117(16 3(1,111x92 211,73:3.97 18,71121153 19, 8!d189fi 19,9571126 i 211,3'1548 *Here 1 < i s 8 is the LD number and 1 s j ~4 is the reactor thermal power level. The thermal power level factor is significant in analyzing four levels but is not significant for three, which indicates considerable error in measuring low thermal power either with TD or by radiation methods since the data set is represented by the ratio Wij/NPMj? To elucidate what influences the error in measuring the relative thermal power, we checked the readings of the PM and LD against one another when they were working simultaneously in the first unit at the Armenian nuclear power station, i.e., we performed aone-factor variance analysis of the data given in Table 3. On testing the hypothesis Ho: N,1 = N,2 = N,3 = N.4, we found that the thermal .power level factor is not significant. In fact SB = 3.9696, SR = 2.8841, while F = 1.38 < Ftab (vB = 3; vR = 28) = 2.95. Hypothesis Ho is adopted with probability 1 - a = 0.95, so the significance of this factor for the previous. set of data with four power levels is due to the larger error in measurement by means of the TD at power levels below the nominal value. This was confirmed by regression analysis in deriving the dependence of Wi on the readings NPM i [3]. According to this evaluation, the confidence limit with probability 1 -0!/2 = 0.975 is 5`~o for a thermal power of about 50oI~ nominal, while the limit is 2.59x', for the thermal power determined from the dependence of NLD i on NPM i Therefore, combined statistical analysis of the thermal power determined by the standard thermaldetec- tors and by radiation methods firstly enables one to calculate the means and the errors of these independently of the thermal power level and secondly enables one to check radiation meters against one another during operation. I 1. V. A. Voznesenskii, "Some results from operating nuclear power stations containing WER-440 reactors," At. Energ., 44, No. 4, 294 (1976). 2. F. Ya. Ovchinnikov, L. I. Golubev, V. D. Dobrynin, et al., Working Conditions in Pressurized-Water Power Reactors (in Russian], Atomizdat, Moscow (1979). ! 3. L. N. Bogachek, K. A. Gazaryan, and V. V. Lysenko, "A correlation-variance method of measuring VVER power," At. Energ., 50, No. 6, 420-422 (1981). 4. E. I. Igna.tenko, V. V: Zverkov, and V. N: Dement'ev, "Computer use in calculating reactor thermal- power," Elektricheskie Stantsii, No. 2, 9-12 (1982). 5. L. N. Bogachek, A . L. Egorov, V . V . Lysenko, et al., "Measuring coolant flow rate and power by radiation methods in the .first unit at the Armenian nuclear power station," At. Energ., 46, No. 6, 390-393 (1979).. 6. K. P. Shirokov (ed.), Methods of Processing Observational Results from Measurments, Trudy Metrologi- cheskikh Institutakh SSSR, Issue 134(194), Standartov, Moscow-Leningrad (1.972). Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 7. A. S. Zazhigaev, A. A. Kish'yan, and Yu. I. Romanikov, Methods of Planning Physics Experiments and of Processing the Results [in Russian], Atomizdat, Moscow (1978). 8. A. Afifi and S. Eisen, Statistical Analysis: a Computer Approach [Russian translation], Mir, Moscow (1982). TEST STAND FOR RESEARCH ON THE PHYSICS OF HIGH-TEMPERATURE GAS-COOLED REACTORS A. M. Bogomolov, V. A. Zavorokhin, A. S. Kaminskii, S. V. Loboda, A. D. Molodtsov, V. V. Paramonov, V. M. Talyzin, and A. V. Cherepanov During the development of new reactors their physical characteristics are investigated on critical assem- blies. High~emperature gas-cooled reactors (HTGR), under development in our country and abroad, have many new aspects. The I.V. Kurchatov Institute of Atomic Energy has built a critical stand GROG which is used to study the physics of HTGR, i.e., the physics that is common to all modifications and the specific physics tha takes into account the features of different versions of HTGR; primarily the VG-400 [1]. The breadth of the studies of the HTGR physics on the GROG stand as well as economic and time con- siderations have made it necessary to employ modeling principles, making it possible to create a simple and flexible design that provides a capability for forming critical assemblies with different parameters quickly and at low cost and performing a large number of experiments on them. Experiments on one-zone and two-zone as- semblies can be carried out on the stand. In the two-zone assemblies the zone under study is formed of full- scale reactor elements and the seed zone is formed of model elements and as a result the criticality of the as- sembly of the whole is produced. This ensures wide variation of the neutron-physical characteristics of the seed zone and their similarity to the system of full-scale elements as far as .properties are concerned. This. similarity eliminates boundary effects and this makes the test zone more representative. The capability for wide variation of the neutron-physical parameters in the critical assemblies using model elements and the similarity of such assemblies to the reactors under study make it possible to solve many problems on model systems. Therefore, in the initial stage, when the full-scale elements are absent, the physics of HTGR are studied on critical assemblies formed of model elements. Since products of the. reaction of the fuel with neu- trons are.difficult to use in assemblies, problems pertaining to the burn-up of fuel are also considered on sys- tems containing model elements. The necessary zones can be built from a combination of elements that are simple and convenient to use: fuel elements (FE) of two types -with enrichment to 109'0 2ssU (at a uranium dioxide density of 0.5 g/cm3 or with the dioxide of natural density at a density of 0.5 and 1 g/cm3), dummy moderating elements (DE) (graphite blocks and inserts into them), and absorbing elements (AE). The full-scale HTGR fuel elements use microfuel dispersed in a graphite matrix. The model fuel elements of the stand GROG, universal physical imitators of HTGR fuel elements, are made in the form of a homogeneous mixture of uranium dioxide and Teflon [2]. This composition of the fuel-element imitators allows them to be prepared by means of a substantially simplified technology and at comparatively low cost. The fuel-element imitators and graphite inserts are cylinders of diameter 50 mm and height 10, 25, or 50 mm. Graphite sleeves are also used to produce the necessary porosity. Continuous variation of the amount of neutron-absorbing material in the assemblies is effected with the aid of a set of absorbers of diameter 50 mm, made of boron-containing water or aluminum-boron alloy, as well as boron carbide rods of diameter 12 mm. The necessary combination of cylindrical elements is placed in the channels of graphite blocks with a 250 x 250 mm cross section. The graphite blocks have nine 55-mm holes, located at the lattice points of a square lattice with a pitch of 83.5 mm. Translated from Atomnaya Energiya, Vol. 57, No. 6, pp. 397-400, December, 1984. Original article sub- mitted February 21, 1984. Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 ~~- Fig. 1. Critical stand GROG for research on the physics of HTGR: 1) spherical fuel elements of the zone under study; 2) fuel-assem- blystorage; 3, 4) graphite blocks of the seed zone and reflector; respectively; 5) experimental.channel; 6) rod of control and safety ' system; 7) oscillator. The neutron-physical characteristics of fresh fuel elements are modeled on the critical assemblies of the stand under the following conditions: the model system has the same principal isotopes and the same ratios of their nuclear concentration as does the full-scale system; the model and full-scale systems approximate a homogeneous system for nonresonance neutrons (the blocking in the fuel is less than 100); the resonance effects in the model and full-scale systems are the same. The necessary ratio of the nu- clear concentrations of the isotopes and the required resonance effects are attained by means of an appropriate set and combination of fuel elements and graphite inserts. The considered set of elements makes it possible also to simulate burn-up on the assemblies of the stand. The products of the reaction of the fuel with neutrons are simulated by using zasU~ zsaU, and boron. The com- bination of fuel elements and absorbing elements (fuel=assembly module) required for this purpose is selected from the condition that the fission and absorption integrals be preserved for the.full-scale and model systems . and that the macroscopic cross sections coincide in the energy range of the neutrons. Computational studies showed that the necessary neutron-physical properties of an HTGR over the entire range of parameters considered are ensured in the critical assemblies of the stand. In particular, the stand provides a good simulation of the field of energy release and spectral characteristics, even in a highly inho- mogeneous system with the principle of OPAZ in the equilibrium state. Structurally; the critical assembly of the stand GROG has been built as follows: A set of graphite blocks forms a cubic stack with a 450-cm face. Different geometrical and physical parameters of the core and reflec- Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 Parameter 2 02 - 2 03 ~ t-O1 1 2-01 1 I A B I A 1 B I 3-OS I ~-O1 ' Size of core, mm z 75 75 75 25 75 25 75 75 y 92 75 75 25 75 25 75 75 Z SO 110 110 1!0 !!0 !10 170 120 Ratio of carbon and uranium nu- 250 500 500 500 51H) 500 250 250 clear concentrations Structure of fuel-assembly module FE-50 FE -50 FE-50 FE 30 FE -50 F)r-10 FE -25 FE-50 DE-50 DE-50 FE-30 DE-50 DE-10 AE-0.5/5 FE-50 FE -50 FE -50' FE-30 FE -50 FE;-10 FE -25 Composition of the elements of the module: FE-50 (30, 10, 25) - a universal physical imita- tor of height 50 (30, 10, 25) mm with logo uranium enrichment; DE-50 (30, 10) - a graphite insert of height 50 (30, 10) mm; AE-0.5/5 (0.5/10) -a disk of boron-containing paper of thickness 0.5 mm and boron content 5 and 10 mg. for are provided by placing different combinations of cylindrical elements into the channels of the graphite stack. Control rods can be placed in the central channels of the columns of the graphite stack. Cylindrical channels of diameter 15 mm in the corners of the graphite columns can hold boron carbide absorbing elements, of diameter 12 mm for simulating perturbations, as well as sensors of the measuring system. When some of the graphite blocks are removed their place can be taken by a fragment, or a mockup of a fragment, of the reac- tor under study. In particular, mockups of spherical fuel elements, absorbing elements, and dummy elements of the VG-400 reactor are used in the stand. Figure 1 shows the makeup of the critical assembly of the stand GROG, including a central zone of spheri- cal elements that is under study, a seed zone, and a graphite moderator surrounding them. As in high-tem- . perature gas-cooled reactors, there is a space between the active zone (core) and the upper end reflector. The design of the stand allows sensors of the measuring system to be arranged in any way desired. Be- sides in-line acquisition and processing of experimental results, the multichannel automatic system of data acquisition, storage, and processing, including a computer and terminals, makes it possible substantially to ex- tend the application of techniques directed at fuller and higher-quality research. No. of graphite columns 324 Maximum linear size of core, mm 4000 Maximum number of channels 2304 Uranium enrichment, 96 0.7-10 Maximum zssU charge, kg 250 Volume porosity of core, qo 5-40 Ratio of carbon and uranium nu- clear concentrations 200-2000 Simulated regimes of HTGR operation initial, transient, Number of randomly arranged rods of control and safety system equilibrium up to 24 The experiments on GROG were carried out in a certain sequence determined by the aims and logic of the development of the investigations as well as by the methodological and technical preparedness for each stage. The experiments .started with a study of small, methodological critical assemblies that were simple in compo- sition and configuration; in the first place this decreases the number of indeterminacies upon comparison with the results of calculations and, second, makes it possible for experiments with a large number of versions to be carried out rapidly and simply. Subsequent experiments were aimed at verifying the simulation of the required neutron-physical properties and studying the distinctive features of the HTGR studied, primarily the VG~00. The scope of the investigations on the. methodological assemblies encompassed the development of experi- mental, methods and the technique of work on the stand, verification of the initial (specification) data on the ma- Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 TABLE 2. Composition of Measurements on Methodological Assemblies Parameter studied 1-OII P-OII2-U2~ 2-011 3-O1I .-OS Keff -f- -E- -F- -I- -t- ~- Reactivityeffect -}- -~ -I- -{" -{- -~- Eff. of control and safety -}- -{- +- -{- -{- -}- rod system R eaction rate Thermal-neutron activation: tuaDy ...~ Abs, densityof neutron flux -}- Flux distribution: resonance neutrons -{- -~- fast neutrons -{- Spectral index -{- -{- -}- -{- -~- S (Lu, 1)y) Rid Dy -{- 1s~Au Rate of 238U(n, y) reaction: along radiusofphysical ~- -{- -{- -{- -{- -}- imitator alon height of physical ~ -~- -~- .-f- tor imit Plutonium coefficient f + .+ + .+. .~ .~-. Pen ~' -)- ~- ~' ~ ~- Reactton rate measured with -~- -{- -}- "}- -~- --~ fission and boron chambers ~ i TABLE 3. Effective Neutron~Vlultiplication Fac- tor in Assemblies 1-01 I 2-01 ( 9-01 1 4-01 1,00661 1,00461 1,0551 1,01051 0,0002 0,0002 0,003 0,0005 0,997 1,006 1,046 1,015 terial composition of the assemblies, and approval of some methods and programs for the physical design of HTGR and methodological assemblies (Tables 1 and 2). Most of the measuring methods are used in experiments on different critical assemblies. Two original features of GROG .are scanning of the neutron flux over the volume of the core with the aid of small fission and boron chambers [3] and measurement of appreciable varia- tions of reactivity in physically large breeding systems. The neutron-physical design of the assemblies was carried out in two stages. First we calculated unit cells to obtain homogenized macroscopic constants, which were then used to determine the neutron-physical param- eters of the assemblies. With allowance for the distinctive features of the assemblies of the stand GROG the group constants were calculated with the aid of the programs WIMS [4, 51, MONR1 [6], and PTT [7]. The neutron- multiplication factor Keff and the space-energy distribution of neutrons in the assemblies were determined from the three-dimensional program QUM-3-HER [8]. Some distributions of the reaction rate and estimates of the effects were found from the two~iimensional program PENAP [9] . The specification data for the elements. of the assembly were checked, analyzed, and statistically pro- cessed beforehand. Computational and experimental studies" [10] indicated the homogeneity of the properties of the graphite stack in different directions. The agreement, within the limits of the error of measurement (2~) between the measured neutron diffusion length in a continuous graphite medium and the corresponding calcu- lated value, obtained by using the specification data for: graphite elements, gives confidence as to the correct- Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 ness of the latter. The calculations showed that the technological deviations of the parameters of the assembly elements have a weak influence on the principal characteristics of the assembly: no more than 0.1`~ on Keff and no more than 19'0 on the relative deviation of the neutron fluxes [ll]. As follows from Table 3, the calculated values of the effective neutron-multiplication factor in the assem- blies are in satisfactory agreement with the experimental data (the difference does not exceed 19'0). The experimental and calculated values of the spectral parameters agree to within 79o while the difference between the experimental and calculated values in the distribution of the reaction rate may reach 109'0. A transition to experiments on more complex assemblies, close in geometry and composition to the cores of HTGR, will help improve the methods and programs for the design of such reactors. LITERATURE CITED 1. E . V . Komarov, F . V . Laptev, V . G . Lyubivyi, F . M . Mitenkov, et al., "The VG-400 atomic-power en- gineering facility. Possible reactor core designs," At. Energ., 47, No. 2, 79 (1972). 2. A. M. Bogomolov, A. S. Kaminskii, A. D. Molodtsov, et al., "Fuel element of the critical assembly of a reactor," Inventor's Certificate No. 915628, Byull. Izobret., No. 30, 300 (1982). 3. A. B. Dmitriev and E. K. Malyshev, Neutron Ionization Chambers for Reactor Engineering [in Russian], Atomizdat, Moscow (1975). 4. J. Ackew et al., "A general description of the lattice code WIMS," J. BNES, 5, No. 4, 564 (1966). 5. , N. I. Laletin and V. A. Lyul'ka, "On resonance absorption of neutrons in 238U," in: Neutron Physics [in Russian], Part 4, Atomizdat, Moscow (1980), pp. 35-37. 6. A. G: Sboev, "Program for the calculation of resonance absorption of neutrons in reactor cells by the Monte Carlo method MONR1," Vopr. At. Nauki Tekh., Ser. Fiz. Tekh. Yad. Reakt., No. 5(18), 91 (1981). 7. A. S. Kaminskii and L. V. Maiorov, "The PIT program for the calculation of fluxes of slow neutrons by the Monte Carlo method with allowance for thermalization," Vopr. At. Nauki Tekh., Ser. Fiz. Tekh: Yad. Reakt., No. 8(21), 76 (1981). 8. S. S. Gorodkov, M. I. Gurevich, and N. L. Pozdnyakov, "Instruction for the use of theQUM-3-HER program for the calculation of athree-dimensional or two-dimensional heterogeneous reactor," Preprint 1A~~794, Kurchatov Institute of Atomic Energy, Moscow (1977). 9. P. N. Alekseev, S. M. Zaritskii, L. N. LTsachev, and L. K. Shishkov, "The TVK-2D complex of programs," Vopr. At. Nauki Tekh, Ser. Fiz. Tekh. Yad. Reakt., No. 4(33), 32 (1983). 10. L. A. Anikina, A. S. Kaminskii, and E. S. Subbotin, "Measurement of the diffusion length of a graphite .stack by an,improved prism method," Vopr. At. Nauki Tekh, Ser._Yad. Konstanty, No. 4(53); 44 (1983). 11. L. A. Anikina, A. M. Bogomolov, and A. S. Kaminskii, "Influence of spreads of parameters of elements of critical assemblies of the stand GROG on their neutron~hysical characteristics," in: Experimentation in Reactor Physics [in Russian], Moscow (1983), pp. 237-241. Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 STUDY OF MODEL COILS MADE OF SUPERCONDUCTOR INTENDED FOR THE WINDING OF THE T-15 TOKAMAK I. O. Anashkin, E. Yu. Klimenko, S. A. Lelekhov, N. N. Martovetskii, S. I. Novikov, A. A. Pekhterev, The preparation for outfitting the Tokamak-l5 has included several stages of tests on a superconducting current-carrying element (SCCE): study of the current-carrying capability of short specimens, and tests of model coils, full-scale test coils, and regular windings of the facility. Much attention to the behavior o~ SCCE in a winding has been paid since the decision to use niobium- tin multifilamen- taryconductoras the basis of the current~arrying element of the superconducting toroidal-field winding (STFW) for the T 15. Since 1978 tests on windings made of niobium-tin conductors similar in construction to the current-carrying element of the STFW have been under way at the I. V. Kurchatov Institute of Atomic Ener- gy. Some of the results have been published [1-3]. The testing of the model coils was preceded by the construc- tion of special stand equipment and was carried out on a stand designed for tests of niobium-tin windings with a diameter of ~ 1 m [4] . This, on the one hand, gave reason to hope that the tests could be performed fairly rapidly but, on the other hand, led to a risk of damage to the niobium-tin filaments of the SCCE, designed for use in a construction with a bending radius three times that of the model coil. This program came into being in connection with the almost formal necessity of making certain that the behavior of the SCCE, mockup speci- mens of which were successfully tested in 1977, would be predictable in the model winding. The first experi- ments, however, gave some unsettling results: The model windings went into the normal state at a current that was only half that used for short specimens of the current-carrying element (Fig. 1). As is seen, the results did not confirm the hopes for ensuring the operating characteristics of the T-15. The main objective of the second series of tests of model coils, which are described in this paper, was to ascertain the causes of the cur- rent degradation, to demonstrate the possibility of obtaining in the windings currents close to the current in a short specimen, and provisionally to verify that the design of the winding adopted for the T-15 facility is ap- propriate for its purpose. The following assumptions about the causes of current degradation in model coils could be made a priori: 1. The existence of macroinhomogeneities of the conductor (weak spots). It could be assumed that in 1979 since the tests were conducted on a limited number of specimens such weak spots were not detected and in tests of the model coil it was those weak spots that determined the critical, current. (Now that 100 short speci- mens have been tested and their total length is several times the length of the model winding and not a single test in a strong magnetic field gave a result below the theoretical requirements, this cause need not be con- sidered.) 2. Damage to the current-carrying element during transportation or winding. 3. Deformation of the winding under the action of ponderomotive forces, leading to heat release exceeding the stability threshold of the current-carrying element. In order to ascertain which of the last two causes was responsible for the observed degradation, a care- ful study was made of the last model of the first series of tests [3], henceforth called model A; its parameters are given in Table 1. The arrangement for the tests did not differ from that described earlier [3] (Fig. 2). A segment of the model winding was placed in the gap of a superconducting niobium-tin solenoid of diameter 240 mm. This solenoid produced a magnetic field of up to 8 T in part of the winding and its interaction with the model winding resulted in a force of up to 1.2 ?105 N that bent the model winding. The solenoid was cooled by immersion in liquid helium and the model winding was cooled by a flow of helium along the hollow current-carry- Translated from Atomnaya Energiya, Vol. 57, No. 6, pp. 401-404, December, 1984. Original article sub- mitted March 6, 1984. 830 0038-531X/84/5706-0830$08.50 ?1985 Plenum Publishing Corporation Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 ~~ ~, a x ?,~ Fig. 1. Results of first series of tests of model windings: O) double pancake coils of diameter 560 mm; O) glued pancake coil of diameter 560 mm; CI) imitation of STFW; model A; x) operating point of glued pancake coil; 1, 2) characteristics of SCCE of glued pancake coil and model A, respectively. TABLE 1. Parameters of Models Tested in the Second Series STENO-2K-9- STENO-2K-11- STENO-2K-11 Material of winding 9225' 9225 14641 Type of winding Double pancake Single pancake Double pancake No. of turns 11 4 8 Winding dimensions , mm: inner diameter 780 761) 770 outer diameter 860 835 850 height 38 78 38 PresenceofpuLsed Yes . No Yes toroidal wtnding Coupling factor, T/ kA: between field and cur 0,15 0,04 0,073 rentof modelwinding between field and cur- rent of xternal solenot~ between field and current of toroidal winding *Superconducting current-carrying element with two cooling channels and nine niobium- tin cores, each containing 7225 filaments . ing element [1]. The model winding was insulated from the liquid helium in the cryostat by fiberglass textolite tape, reducing-the heat exchange with the helium in comparison with internal cooling enough so that it could be neglected. Wound around the model winding was an auxiliary toroidal winding, imitating the field of the plasma column, directed parallel to the axis of the current-carrying element. The model and solenoid windings re- ceived their power supply in series from a lO~cA generator and an auxiliary 3-kA generator made it possible to add current to or subtract current from the model winding. This arrangement made it possible to study the dependence of the critical current of the model on the magnetic field. ? In the second series of tests we used an additional compensator winding that increased. the sensitivity of the bridge circuit for the detection of the normal phase so that the reverse part of the current-voltage (I V) characteristics of the model would be observed. Tensometric displacement gauges were-.also set up to record changes in the shape. of the winding and any possible displacements of the winding relative to the solenoid. In tests of model A no abrupt deformations of the model were recorded up to the level of ~currerit"attained. Recording of the I=V characteristics allowed an unambiguous conclusion to be made about the cause of the de- gradation. The smearing of the I-V characteristic of the transition of the model was several times the smear- ing of the I V characteristic of the SCCE specimen (the parameter of the growth of the electric field with the current was 1050 A for the model and 360 A for the short specimen, Fig. 3). This allows the lowering of the current-carrying capability of the winding to be attributed to local damage to it during winding, "transportation, or previous tests. Identification of the causes of the failure of model A led to the preparation of a new model in which the probability of damage to the SCCE would be reduced to a minimum. The second problem in the preparation of Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 IT-3 IT-1 IT-2 ~-r Fig. 2. Arrangement for tests of model windings: 1) control valves; 2) cryostat; 3) power supplies; 4) solenoid setting up field in segment of winding; 5) pulsed toroidal winding; 6) model winding; 7) bearing flange; 8) heater. Fig. 3 4 6 B f0 Magnetic field, T Fig. 4 3 ao .. 3 7S f0 ~ 5 m x 1 P 6 B f0 Carrent, kA Fig. 3. Comparison of the parameters of the growth of the electric field with the current for the model (a) and the short specimen of SCCE from which the model was made (b) . Fig. 4. Results of second series of tests of model windings_O) model B; O) model C; x) operating paint of T-15 STFW; 1, 2) characteristics of SCCE of models B and C . Fig. 5. Results of tests of models by the method of "extrapolation to zero heater" (the squares correspond to the case when the critical current was limited by mechanical perturbations in the winding):. O, O, O) models B, C, and C';- = ?- ?-) expected behavior of model C' in the absence of mechanical perturbations. model B was, that :of demonstrating the possibility of producing a winding in which not only degradation but also conditioning of ;the SCCE would be eliminated. The latter is essential since at the design current of 5.2 kA the T 15 wind,ing_ (STFW) is not stabilized in a stationary state and the beginning of conditioning of such a large magnet. system can be tantamount to a limitation of its working capacity. Model B (see Table 1) was aone-layer pancake coil, glued on either side to 0.8 mm stainless steel sheet. To diminish the risk of breakage of the niobium-tin filaments, immediately after the application of a stabilizing layer of electrolytic copper [5] the cur- rent~arrying element was wound on a receiving cassette with an inner diameter of 0.8 m. It was not subse- quently rewound. Before the gluing the tension in the SCCE was reduced and this permitted fiberglass textolite insulating insert to be placed between the windings. Because of the bandaging scheme adopted model B has an Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 d=B,ST ~- % ~,6 U4 0,2 0 -Q2 -0,4 -J,6 -UB -1 0 1 -1,0 Tensile_deformation, % Fig. 7. Fig. 7. Dependence of critical current of the SCCE on the tensile-compressive deformation: O) extrapolated critical current of model B. Fig. 8. Dependence of the stability of the SCCE on the amplitude of the pulsed magnetic field, the relative ori- entation of I and Bo during cooling with two-phase heli- um at a pressure of 0.13-0.15 MPa (~) and supercriti- cal helium at a pressure of 0.4-0.5 MPa (O). extremely great rigidity with respect to bending in the plane of the winding. The tests were carried out accord- ing to the scheme indicated above. Without any conditioning a reversible transition to the normal state was at- tained at a current substantially above the critical current of model A (Fig. 4). Further on we discuss why the current of the short specimen is not attained in small~liameter models. In any case the slope of the I V curves, which coincides with the slope of the I V curve of the short specimen, indicates that the SCCE in model B was not damaged. This is also indicated by the results of tests using the so-called method of "extrapolation to zero heater" (Fig. 5): the current at which a resistive zone appears in the winding is studied as a function of the power of a local heater arranged on it or as a function of the readings of a thermometer placed together with the heater. In the absence of damage in the SCCE and perturbations in the winding, this dependence is plotted as a smooth, almost straight line and the point corresponding to the transition of the winding when the heater is switched on lies on this line (models B and C). If the SCCE has been damaged and the heater turned out not to be at the site of the damage, it is unlikely that characteristic kinks would appear on this line. If the transition is caused by perturbations, e.g., of a mechanical origin, then starting from a certain value of current at which the acting forces reach an appreciable value the critical current ceases to depend on the power of the heater and the transition takes place abruptly, without a recorded resistive segment of the I-V curve (model C'). Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 The fact that a current satisfying the requirements of the STFW for the T-15 tokamak was attained in model B with rigidly fastened winding without any conditioning stimulated the construction of model C, whose design imitated the design of the T-15 STFW. The current-carrying element was wound on a rigid stainless steel ring with an L-shaped cross section (Fig. 6). In this winding, too, we managed to approach the current of a short specimen, but only after a single conditioning. This is a cause of some concern since the forces acting in the STFW of the T-15 will be several orders of magnitude greater than in the model and the number of con- ditionings may increase sharply. The fact that in models B and C the critical current is somewhat below the level of current in the short specimen can be explained if the dependence of the current-carrying capacity of the SCCE on the deformation is taken into account (Fig. 7). During winding of aheat-treated SCCE of radius 0.4 m the inner layer of the transported cores turns out to be compressed.0.5go and their current-carrying capacity should diminish. This decrease cannot be compensated by an increase in the current-carrying capacity of the stretched cores of the outer layer since the cross resistance of the conductor is too high for the current to be redistributed between the stretched and compressed cores over the transposition length (80 mm) on which the cores change places. The critical current of the bent SCCE practically coincides with the critical current of the compressed cur- rent carrying element. The earlier hypothesis that the sensitivity of the SCCE to perturbations by a pulsed magnetic field de- pends on the relative direction of transposition, the current in the winding, and the pulsed field has been checked on models A and B. Since the transposed cores of the SCCE form helices about the axis of the conductor, a change in the longitudinal magnetic field can induce in them a current in the same direction as the transport current or in the opposite direction. At the same time eddy currents in the stabilizing copper warm up the superconductor slightly. Depending on the resultant current that flows in the superconductor, this warming up either causes the transition of the superconductor to the normal state or is insufficient to do so. Figure 8 gives the results of a study of this effect on model A. A current slightly below the critical value (6 kA) was introduced into the model winding. Then a current of 0 to 1200 A was introduced into the toroi- dal winding wound around the model winding, up to a certain value, starting from 50 A. After this the current was removed from the winding approximately exponentially with a time constant of 7 10 msec; this imitated the nature, amplitude, and rate of change of the magnetic field of the plasma during tearing of the plasma in the T-15 facility. If with such removal of current the model winding remained in the superconducting state, the cur- rent in the toroidal winding was increased by 50 A and if it went into the normal state the corresponding value of the longitudinal field was recorded as the maximum attainable for the given value of current in the model winding. The experiments were conducted in three SCCE cooling regimes, corresponding to two~hase and one- phase supercritical helium in the channels. It was shown that cooling with two-phase helium slightly increases the stability of the winding.. Thus, after a second series of tests of model windings of the current-carrying element intended for the STFW of the T-15 tokamak it has been established that the material of the winding is suitable for use and has a considerable reserve of stability in stationary regimes and in pulsed perturbations by a longitudinal magnetic field. The current-carrying element can be used in the STFW of the T-15 tokamak, but the requirements as to ensuring the rigidity of the winding are much more stringent. 1. N. Chernoplekov, "Status and trends of superconducting magnets for thermonuclear research in the USSR," IEEE Trans., MAG-17, No. 5, 2158 (1981). 2. D. P. Ivanov et al., "Study of the properties and stability of a superconducting current~arrying element for the Tokamak-15 facility with respect to pulsed poloidal fields," in: Proceedings of the Al1~Jnion Con- ference rnthe Engineering Problems of Fusion Reactors [in Russian], Vol. 8, A. D. Efremov Scientific- Research Institute of Electrophysical Apparatus, Leningrad (1982), pp. 94 103. 3. D. P. Ivanov et al., "Study of model coils of the superconducting current-carrying element of the Tokamak-15 facility," Preprint IAE-3715/7, Institute of Atomic Energy, Moscow (1983). 4. I. L. Zotov et al., "Test stand for large superconducting magnet systems," Prib. Tekh. Eksp., No. 6, 176 (1974) . 5. V. Agureev et al., "Electroplated stabilized multifilament superconductor," IEEE Trans., MAG=11, No. 2, 303 (1975). Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 OSCILLATIONS IN THE CONCENTRATION OF ARTIFICIAL RADIONUCLIDES IN THE WATERS OF THE BALTIC AND NORTH SEAS IN 1977-1982 D. B. Styro, G. I. Kadzhene, UDC 551.464.679 I. V . Kleiza, and M . V . Lukinskene As a result of extended atmospheric tests of nuclear and thermonuclear weaponry a certain "background" of artificial radionuclides has been formed in the environment [1] . Up to the present time this background has decreased significantly in the waters of the Pacific Ocean. Oscillations of it have turned out to be primarily associated with discharges of radioactive wastes by undertakings of the nuclear industry [2, 3] or with the ef- fects of individual atmospheric nuclear explosions carried out up to the present by the People's Republic of China [4, 5] . The Baltic and North Seas are contaminated with various radionuclides, of which the most radioactively dangerous are 137Cs, 90Sr, and to a lesser degree, 144Ce. If contamination of the waters of the Baltic Sea is caused mainly by global fallout, the waters of the North Sea are contaminated predominantly by the wastes of undertakings of nuclear industry [2, 3, 6] . Regular observations of the radioactive contamination of the water areas of these seas have been made by us since 1973 [7 10]. The natural data obtained have permitted estimating the possible values of the concentra- tion of various radionuclides, investigating with the help of a mathematical method (the so-called objective anal- ysis) the structure of the radionuclide concentration fields [10-12], and revealing the possible source of the contamination [13] . The isolation of radionuclides from seawater was performed by the radiochemical methods described in [14, 15] and with the help of a very rapid sorption procedure for is7Cs [16]. We have determined the maximum possible error of a test from the results of numerous measurements of the concentration of the radionuclides of one and the same sample of water. In the determination of the concentration of is7Cs and 144Ce it amounted to 2590 and for 90Sr - 159'0? The results of measurements obtained on the basis of the sorption procedure under natural conditions were selectively monitored by radiochemical analyses during each series of observations. The average values. of the concentration of the radionuclides investigated are not constant in the waters of the Baltic Sea, and especially large variations in the concentration of 90Sr are characteristic (Fig. 1). The concen- trations of the other radionuclides are subject to oscillations to a lesser extent. It has been .established in the analysis of the published data [4, 5, 17-20] that the variations in the concen- tration of radionuclides in the surface waters of the Baltic Sea are caused by changes in the global fallout due to atmospheric nuclear tests conducted by the People's Republic of China. The time of the nuclear tests corre- lates well with the time of increase in the average concentration of radionuclides in the waters of the Baltic Sea (mainly 90Sr). Amore noticeable increase in the average concentration of 137Cs and 144Ce has been noted in the fall of 1980 after thermonuclear tests in the People's Republic of China. [5] (see Fig. 1). One should empha- size that the increase in the intensity of global fallout in recent years has been accompanied mainly by an in- crease in the concentration of 90Sr in the Baltic Sea; it has reached its own "na.tural" background values in a more extended period of time than has 137Cs. The contamination of the North Sea waters is caused mainly by discharges of wastes by undertakings of the nuclear industry [2; 3] . Therefore the variations in the global fallout do not cause noticeable variations in the is7Cs concentration, as has been observed in the Baltic Sea. The average concentration of is7Cs in the sur- face waters of the North Sea has increased especially strongly since 1976: thus from 1975 through 1976 it in- creased from 36 to 60 Bq/m3, in 1977 it increased to 108 Bq/m3, an~1 it maintained approximately the same value up to 1981 (106 Bq/m3). A decrease to 84 Bq/m3 was detected only in the fall of 1982. Translated from Atomnaya Energiya, Vol. 57, No. 6, pp. 405-407, December, 1984. Original article submitted April 4, 1984. 0038-531X/84/5706- 0835$08.50 ?1985 Plenum Publishing Corporation 835 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 .ra i 200 1978 9979 Fig. 2 Fig. 1. Average values of the is7Cs (O) , 90Sr (~) , and i44Ce (x) concentrations in the surface waters of the Baltic Sea. Fig. 2. Variation of the is7Cs (O), and 90Sr (~) concentrations in the surface waters of the North Sea near the bay of the Firth of Forth. TABLE 1. Concentration of Radionuclides in September-October 1982 in the Surface Waters of Seas, Bq/m3* Baltic 25/7 34/8 16/5 North 233/32 37/9 ,48/6 Norwegian (south- ern part) 8/4 9/3 33/5 *The numerator denotes the maximum concen- tration, and the denominator denotes the mini- mum concentration. The most contaminated waters evidently enter the North Sea from the bay of the Firth of Forth, which is connected by systems of canals with the Irish Sea. Many years of observations have permitted detecting a con- stantly increased 137Cs concentration in this region of the sea [2, 10, 13]. One should note that here the in- creases or decreases in the is7Cs and 90Sr concentrations correlates well, although the absolute values of the soSr content are lower by an order of magnitude (Fig. 2). In this region the concentration of both radionuclides have their largest values as a rule. Consequently, the waters of the North Sea are contaminated by both radio- nuclides due to discharges of wastes by undertakings of the nuclear industry. We shall dwell in more detail on the results of measurements of the concentration of radionuclides in'the surface waters of the Baltic and North seas in the fall of 1982, since they have been published first. The total content of the radionuclides under discussion in the surface waters of the Baltic Sea decreased significantly in comparison with 1981 (see Fig. 1) except for 137Cs, whose average concentration remained the same (15 Bq/m3) [10] . However, the absolute values of the concentration of these radionuclides in the surface waters measured at different points of the Baltic Sea vary within rather wide limits (see Table 1). Figure 3 illustrates the nonuniformity of the distribution of radionuclide concentrations in the surface waters of the Baltic Sea. The greatest nonuniformity in concentration is characteristic of 90Sr (Fig.. 3b). Here an interesting regularity is revealed: the is7Cs concentration decreases mainly at the points where the 90Sr and i44Ce concentrations increase. Natural oscillations of the radionuclide concentrations in the waters of the Baltic Sea significantly exceed the errors of the measurement results. The cited method -objective analysis using the experimental information obtained -was applied to the investigation of the structure of the concentra- tion fields of radionuclides in the reservoirs under investigation.- The structure of the is7Cs concentration field in the surface waters of the Baltic Sea in the fall of 1982 is given in Fig. 4a. This structure has changed noticeably in comparison with the spring of 1981 [10], although 836 Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1  Declassified and Approved For Release 2013/02/22 :CIA-RDP10-021968000300050006-1 7, d a _ ?3 ,~ ,' 2~ x ,,191 >0 -S 0 S 70 a 70 '.S 0 b ~_ 5 70 iii i 0 dG, Bq/m3 c Variations of the (a) 137Cs, (b) 90Sr, and (c) 144Ce concentrations iri the surface waters of the Baltic Sea in the fall of 1982 (vertical solid .lines denote the maximum experimental erro