B pec 15 w7 G operu!ed by UNION CARBIDE CORPORATION . R '---;_NUCLEAR DIVISION o U S ATOMIC ENERGY COMMISSION ORNI. TM 1960 OPERATION OF MOLTEN-SALT CONVECTION LOOPS IN THE ORR o H C chage ©- v E, Compere -~~~ Jo M, Baker - E. G. Bohlmann s NUTICE Thls documem -contains mformuflon of a prehmmory nuture,-"r—' L ~and was prepared pr '“‘“"‘Y for internal use at the Ock Ridge Nationol "~~~ - _'"Laborutory 1t is subject to revision or correchon und therefore doe;' St 'not represent o fmcl fepon : - o et T . ': %‘%zfi‘fii DI’ TFTS QOC&‘&:;&A v ittt 0 | ittt ks Bt < ot b 4 e e 1.5 Mthe.z PRSI £4 TTL , 1 AR 1- apppene ~uOTSRTY i | v LEGAL NOTICE This report was pnpnrod as an account cf Govcmmom spomorod work. Neither th. Unind Smhs, ' nor the Commission, nor any person acting on behalf of the Commission: o A. Makes any warcanty or representation, axpressed or implied, with respect io the accuracy, - completeness, or usefulness of the information eontained in this report, or that the use of any informetion, apparatus, ‘method, or procou disclosed in this report may not Inftinge privotely owned rights; or " B. Assumes any liabilities with respect to the use of or for domages nsulfing from tl\. use of any information, upparatus, mthod or process disclosed in this report. As used In the above, "persen acting on behalf of the Commission® includes any omploy-t or " contractor of the Commission, or employes of such contractor, to the extent that such employes e ' o, or contractor of the Commission, or employee of such contractor prepares, disseminates, or = ' _ provides .ccoss to, ony information pursuant to hw omploymenr er contract with the Commission, - or his employment with such eonlruclor. ay e -y ORNL-TM-1960 Contract No. W-7405-eng=-26 REACTOR CHEMI STRY OPERATION OF MOLTEN-SALT CONVECTION LOOPS IN THE ORR H. C. Savage E. L. Compere Js M. Baker E. G. Bohlmann Ly LEGAL NOTICE % This report was prepared as an account of Government sponscred work, Neither the United | | States, nor the Commission, nor any person acting on behalf of the Commission: A. Makes any warranty or representation, expressed or implied, with respect to the accu- racy, completeness, or wsefulness of the information contained in this report, or that the use i of any information, apparatus, method, or process disclosed in this report may not infringe . © privately owned rights; or . B. Assumes any linbilities with respect to the use of, or for damages resulting from the - [ ' uge of any information, apparatus, method, or process disclosed in this report. " As used In the above, *‘person acting on behal] of the Commission’ includes any em- | ployee or contractor of the Commission, or employee of such contractor, to the extent that ‘such employee or contractor of the Commission, ol employee of such contractor prepares, disseminates, or provides access to, any tnformatich pursuant to his employment or contract with the Commission, or his employment with such contractor. ’ : DECEMBER 1967 OAK RIDGE NATTONAL LABORATORY U. Oak Ridge, Tennessee o operated by ' UNION CARBIDE CORPORATION for the S. ATOMIC ENERGY COMMISSION o DISTRIEUTION OF THIS DOCUMENT IS uzuffi?rse‘ - | - . \"! e . i 0 A - iii TABLE OF CONTENTS ABSTRACT 1 1. INTRODUCTION ¢ eouuuvseninessnnnosannaesorncosssessasesasnnases 1 2. DESCRIPTION OF REACTOR IRRADIATTON FACTLITY «vveevevenevenennens 2 3. DESCRIPTION AND OPERATION OF FIRST LOOP EXPERIMENT ©eovvececenes 2 2 4 31 kscrlptlon .....‘...’."......'..................I.......‘ 32 Operatlon .....II...l.'."........‘...‘....-‘.'...‘.4...Il.. 4. EVALUATION OF SYSTEM PERFORMANCE IN-PILE SALT LOOP NO. 1 ...... 13 4.1l Temperature CONLYOL «eeevessesesoscascascasssssnsssavensces 13 ® - 4.2 Problems Encountered During In-Pile Operation ..e.eeeveees. 14 i- DESCRIPIIION AND OPERATION OF IN-PIIIE SALT IIOOP NO 2 R EE RN 18 % 5.1 Loop Descrlptlon tetecsanesrrttaresstsenseenensesacessssasss 18 5.1.1 Core Cooling COL1E eveeereoansonnacncessccasaseeanae 18 5.1.2 Salt Sample LINE seevesensoscescnosscercasensseanass 18 5.1.3 Cooler for the Gas Separation Tank secasessencancnes 20 5.1.4 Salt Flow by Convection saeevesecorscsennsasccnnsees 20 5-2 Operatlon Of In"Plle Salt LOOP N0-2 e s s e PRSI OIRNRNEOLELEOSEOESEOLLGOGTES 20 521 OUt-Of-Plle Test Opera'ti()n es e s s re s s e At anr It eI eean 20 522 In-Plle Operati()n Of LOOP NOO 270¢lonioolclonoouoooo 22 6. EXAMINATION OF FAILURE IN CORE OUTLET PIPE Teseerasrecrrsssnnacne 25 7- DISCUSSIONANDCONCLUSIONS -.p--.o-.o.ooooonc--oioooo--oo.-oooouo_ 29 ACKNOW[IEDG‘M:NT .._loo......OO.OOI;OVQOQ‘COQ;ll_t...ql!.Cboooioo.lool.clc 33 -y Figure Number MWD 0 10 11 G 14 iv IIST OF FIGURES Pipe, Molten-Salt LOOP No. 2 (NZOX) Seessssparars g : Page Htle | Number In-Pile Molten-Salt Loop Facility, ORR HN-1 seceevaecnans 3 In-Pile MoltenfSalt Convection LOOp NO. 1 esecsssccssesas 5 Photograph of Partially Assembled Salt Loop NO. 1 sevesss 6 Salt Sampling and Addition System, In-Pile Molten- | | S81t LOOP ressesssscnscresesccansccectccscsccstorscrenns 7 "~ Photograph of Sait Sampling afid Addition Syste@ csesessees 8 Thermocouple Locafions for Molten-Salt Loop No. 1 essceeees 1L Nuclear Heat Generation in Molten-Salt Loop No. 1 eeeeees 12 Photograph of Broken Coolant Coil, Molten-Salt LOOp NO. 1 eocecseesensccannsaancrasnstaennancecccncnes 15 ~ Photomicrograph of Coolant Coil Break, Molten- Salt LOOp NOe 1 cceoesccasscsnsssessacasasccnsssscosccncssss 1O Photograph of In-Pile Mditen-Sa1t_Loop No. 2 eveneieeene 19 Photograph of Partiaily Assembled LOOp NO. 2 eseesesseses 21 Nuclear Heat Generation in Molten-Salt ILoop No. 2 cireeas 23 - Thermocouple Locations forMolten-Salt.Loop NOw 2 eveeses 27 Postirradiation Photbgraph.of;Molten-Salt Loop : No. 2 Showing Location of Leak in Cpre_Ouflet Pipe «.c... 28 Photomicrograph of Crack in Hastelloy N Outlet - o, ‘-P 43 v -y 5 ITST OF TABIES Table ' ‘ - Page Number ' Title Number 1 Operating Time Under Various Conditions for In-Pile Molten-%lt,LoopNo.l'.........'....lllll...'..’l'..... 9 2 Typical Loop Temperatures with Solvent Salt and Fuei Salt with the ORR at 30 Mw for In-Pile Molten-Salt LoOp NOe 1 ceveesossacecsccsassoscsnscasssnscossssnssssssse 10 3 Tabulation of Component Failures Observed in In-Pile Loop NO.l...........‘...-.'.....'-...'.......‘.......... 17 Operating Time Under Various Conditions for In-Pile * 4 MOlten'%lt LOOP NO.2 ® 8 0§ 0 5 T 0 S B Y S S S B 0SSN eSS LE eSS 25 T 5 Typical Loop Temperatures with Solvent Sslt and Fuel Salt with the ORR at 30 Mw for In-Pile Molten-Salt LOOP NO.2.......‘I......I'..l.l...l......-'..l...'...‘.. 26 T - ) 1 -¥ & -’} OPERATION OF MOLTEN-SALT CONVECTION IOOPS IN THE ORR H..C. Savage © E. L. Compere J. M. Baker E. G. Bohlmann ABSTRACT Two molten-salt convection loops have been operated in beam hole HN-1 of the Oak Ridge Research Reactor. Both loops contained molten-fluoride fuel salt (?ILiF-BeF,-ZrF;-UF,) with enriched uranium varying in concentration up to 2.1 mole %. Irradiation of the first molten-salt convection loop experiment in the ORR was terminated on August 8, 1966, after 1484 hr of in-pile ogeratlon and develogment of 1.1 x 10*8 fissions/ce (0.27% 22°U burnup) in the 'IiF-BeF,-ZrF,-UF, (65.16-28.57-4.90-1.36 mole %) fuel. Average fuel power den- sities up to 105 w/cc of salt were obtained in the fuel chan- nels of the core of MSRE-grade graphite. , Irradistion of the second loop experiment in the ORR was terminated on April 4, 1967, after 1955 hr of in-pile opera- tion and develo gment of 8.2 X 108 fissions/cc (1.2%4 237y burnup) in the 'LiF-BeFy-ZrF;-UF; (65.26-28.17-4.84-1.73 moled) fuel. (The uranium concentration was increased to 2.1 mole % for a short time near the end of the experiment.) Average fission heat density in the fuel salt channels of the graphite core was 165 w/cc when at full power. ~ Successful operation of the major heating, cooling, tem- perature control, and sampling systems was demonstrated; how- ever, both loop experiments were terminated because of breaks in the primary loop systems. : " 1. INTRODUCTION The molten~-salt eonveetion loops are designed te'irradiate 8 represen- _'tative molten-salt fuel 01rculating in contact with graphite and Hastelloy N &t fuel fission power den81t1es up to 200 w/ee in the Oak Ridge Research ""Reactor. Long-term 1n-pile operatlon (one year) to achieve high uranium burnup (up to 50%) is an objectlve of the irradiation experiments. ‘Provi- sions for sampling and replacement of both gas and salt permit condltions in the loop to be determined and_torbe altered during operation. -2 - Irradiation experiments in the loop allow us to study the graphite- salt compatibility, Hastelloy N-salt compatibility, fuel-salt stability, and fission-product chemistry. The interaction of fission prbducts with graphite, metal, fuel and gas phases can be investigated, as can the. effects of irradiation on the respective materials. | In order to maintain end controlfitemperatures around the loop circuit, _suffieient heating and cooling capacity is provided to remove up to 14 kw of fission and gamma heat generated at full power Operation and to maintain the salt molten when the reactor is shut down. Operational experience with two in-pile molten-salt loop. experiments is described in this report. 2. DESCRIPTION OF REACTOR IRRADIATION FACILITY Both in-pile molten-salt loops were dperated in horizontal,beam‘hele HN-1 of the ORR (Fig. 1), which is epproximately 8 in. diem and extends 12 £t from a point adjacent to the reactor lattice to the outside face of the reactor shielding. Two shielded equipment chambers contain the neces- sary auxiliary equipment needed for the salt sampling and addition system. Beam hole HN-i and the associated instrumentation were previously used to operate in-pile loop experiments with uranyl sulfate solutinns'and thorium oxide slurries in support of the'Homogeneeus Reactor Program. 3. DESCRIPTION-AND OEERATION OF FIRST LOOP EXFERIMENT - 3.1 Description The main body of the 1oop'aesemb1y'was fabricated of 2-in. sched-40 Hastelloy N (INOR-8) pipe which contained aigraphite core. The graphite core had eight 1/4-in. holes which served as fuel passages. Fuel volume in the graphite core was ~43 cm® in a total loop fuel volume of ~75 cm’. A gas separation tank served as a salt reservoir and provided for a liquld- vapor interface. A return line from the gas separation tank to the bottom of the graphite core completed.the loop circuit. Calrcd electric heaters and cooiing coils imbedded in a. sprayed nickel matiix surrounded the core section, gas separation tank, and return line to provide temperature con- trol and to meintain the thermal gradients necessary to induce convective [2 ] . o) ¥ -) ¥ A [ ] ORNL-DWG 66-T152 s g GAS SAMPLE LINE REACTOR LATTICE LINER [N Tge Syt FROLTINY o A N e SMALL EQUIPMENT CHAMBER LARGE EQUIPMENT (ROTATED S0° CHAMBER SALT SAMPLE LINE- REACTOR SHIELD " Fig. 1. In-Pile Molten-Salt Loop Facility, ORR HN-1. 4 flbw. A drawing of the convection loop assembly is shown in Fig. 2, and Fig. 3 is a photograph of the partially assembled loop showing the fuel flow channels in the core graphite. | Tubes of capillary dimensions interconnected the vapbr space of the gas separation tank with.remotely located pressure monitoring equipment and a gas sampling and addition system.. The salt sample line (0.100 in. 0D X 0.056 in. ID) was ~12 ft long end was traced with electric calrod heaters which were imbedded, along with the sample line, in & sprayed nickel matrix. The sample line was routed to the salt sempling &nd eddi- tion system in the shielded equipment chamber at the reactor shield face, A manually operated retraction screw was used to position the loop g0 that the neutron flux and reéultant nuclear power generation in the loop could be varied from the maximum (fully inserted position) to ~1% of méximUm.by retracting the loop package some 9 in. away from the reactor lattice. Figure 4 is a diagram of the sall sampling and addition system, and Fig. 5 is a photograph of the system as fabricated for the second in- rile loop. | 3.2 Operation ~The loop package (shield plug, sempling and addition system, and loop ‘assembly) was bperated in an out-of-pile mockup facility for 187 hr with solvent salt without uranium. Composition of the solvent salt was 7LiF-BeF2-ZrF4‘(64.7-30.1-5.2 mole %). Nominai operating temperatures around the loop ranged from 650°C in the core section to 550°C in the cold leg return line. During this out-of-pile test pericd, thfee salt samples were removed from the loop and five salt additions were made'without diffi- culty. Salt éirculation in the ldop was estiméted to be 5 to lO_cc/min based on heat balance measurements around the cold leg. The loop was removed from the mockup facility and installed in beam hole HN-1 of the ORR &nd brought to operating temperature on June 9, 1966. WHumnflmsM%msfltwflhmdmfilth%l%%Wanfimm~ uranium (as 7IiF-UF, eutectic) was added. In-pile operation was continued until August 10, 1966, when theAreactor was shut down and the loop removed because of a fuel leak from a break in the capillary sample line near its point of attachment to the loop core section. -, 4 ORNL—-DWG 66~—-9865 HEATER . PRESSURE MONITOR LINE SALT FLOW PASSAGE THERMOCOUPLE WELL GRAPHITE CORE COHC) HEATER OB COOLING JACKET - - — W e L Ay e e ke ol P L T N e S — e s com cmu rmmoma—— - wv m e mms s omw .- — m e mesemer e e ———— - -~ by -e -m-a D i R = —— COOLING COIL W S . L9 COLD LEG SALT SAMPLE UINE HEATER In-Pile Molten-Salt Convection Ioop No. 1. SALT RESERVOIR . o Y 7, & 2 =) m 2 II!QHI.“.\\ % o e oo e e i A £ [ e %N > u_ ._ N w i h N [ [ Y m P_ il N # i R i *| n o B o i v ¥ ® X B3 Y 2 GAS SAMPLE LINE THERMOCOUPLE WELL COOLER—I£ = " — 4 Fig. 2. - ,flt ‘» *T -oN dooT 3185 PITquessy ATTBTIIRI JOo uydexSojoyd ¢ *B1d - 2 5 SINVL NOILVEVA3S SV < w) o ) s -} SHIELD TN L E] = " £ oo, BORATED POLYETHYLENE SMALL EQUIPMENT CHAMBER FREEZE VALVE RETRAGTION SCREW PURGE TANK BALL VALVES SAMPLE STATION ORNL-DWG 88-T153 Fig. 4. Salt Sempling and Addition System, In-Pile Mblten—Salt Loop. | PHOTO 173308 Photograph of Salt Sampling and Addition System. Fig. 5. , "} ¥ ) ¥ w) -} During in-pile operation two salt semples were removed from the lbop end three salt additions were made. While preparing to remove salt sample No. 8, a leek was detected ih_the Sampling system. This leak precluded further sampling operation. Even though no additional samples could be taken, the addition of 7IiF-UF,; fuel and one final addition of solvent salt to adjust the loop inventory were subsequently made. A tebulation of the operating time for the first loop is given in Taeble 1. Operating temperatures around the lbop circuit with the reactor down and at full power (30 Mw) and with the loop fully inserted are shown in Table 2. ThermoéouPlellocations for the temperatures shown in Table 2 are noted in Fig. 6. Total hucleaf heat‘genérated in the loop as a func- tion of distance from the reactor tank is shown in Fig. 7. Tbble 1. Operating Time Under Various Conditioné, for In-Pile Molten-Salt Loop No. 1 Operating Time Salt in Ioop Reactor waer‘ (ur) Solvent® 0 ‘_ 330b 30 Mw 1025 Fuel® o 0 27 30 Mw 289 Total 1671 - ®solvent salt'composition =:7L1F—BeF2-ZrF4 (64. 7 30.1-5.2 mole. %). ' | Includes 187 hr of out-of-plle mockup oper- CFuel salt composition = 7LiF-BeF2-ZrF4-UF4 (65. 16 28.57-4.90-1.36 mole %) Table 2. Typical Loop Temperatures with Solvent Salt and Fuel Salt with the ORR at 30 Mw for In-Pile Molten-Salt Loop No. 1 ,‘ Core Section Gas Separation Tank Return Line (Cold lLeg) Thermocouple Location® : Number Solvent Salt Fuel Salt Solvent Salt Fuel Salt Solvent Salt Fuel Salt | Temp, °C Temp, °C Temp, °C Temp, °C Temp, °C Temp, °C 1 Core bottom 624 568 2 Lower fuel passage 635 625 3 Upper fuel passage 670 784 4 Graphite center ' 671 634 5 Graphite OD 656 - 648 6 Core top 657 762 7 Inlet well _ 604 721 8 Outlet well 602 636 9 Gas space | 589 626 10 Top | 597 . 598 11 Center : ' 579 591 12 Bottom | | 514 540 13 At core inlet - 584 . 590 8Rrefer to Fig. 6. 0T -l " -y Fi] -) -’y 11 CRNL-DOWG 67-10453 = : : 73/4 in. GAS LINE COOLING JACKET * GAS LINE 5 EXPANSION ~ TANK SPRAYED NICKEL HEATER COLD LEG * THERMOCOUPLES (2) anD (3) WERE IN TWO DIFFERENT FUEL CHANNELS 12 'SALT SAMPLE LINE . 13 1 2 Fig. 6. Thermocouple Locations for Molten-Salt Loop No. 1. . . " ORNL—DWG 67—10154 10,000 7 5000 FISSION AND GAMMA 2000 1000 NUCLEAR HEAT (watts) 500 FISSION 200 400 ' - 2.3 4.3 6.3 83 {0.3 123 LOOP POSITION, DISTANCE FROM REACTOR TANK TO CENTER OF GRAPHITE CORE (in.) Fig. 7. Nuclear Heat Generation in Molten-Salt Loop No. 1. - By -y 13 4. EVALUATION OF SYSTEM PERFORMANCE, IN-FILE SALT IOOP NO. 1 Several,failures of eomponeht parts of the loop and associated systems occurred during'in-pile operation. Finaelly, leekage of fuel salt from a break in the salt sample line caused the first loop experiment to be ter-. minated. All perts of the system which failed were examlned in hot cell facllltles to determine the cause of failure before proceeding with design and fabrication of loop No. 2. These failures and the loop performance are discussed below. 4.1 Temperature Control Heaters. The molten-salt loop package used 21 heaters to control salt temperature in the loop and to heat the salt sample line and associated sampling and addition‘system. All heaters were 1/8 in. OD, Inconel sheathed, magnesium-oxide insulated, with a Nichrome V heating element. These heaters are designed for continuofis.operation at temperatures to 870°C. The heaters on the loop circuit were operated continuously at verious power levels while those on the sampling and addition systems were used intermittently as reqfiired. There were no heefer failures during the 1671 hr of ldop opera=- tion. | S Coolers. Four separate coolers were used to remove the 8.8 kw of fis- sion and gamma heat produced wheq-the loop was fully inserted and, in con- juncfiidn with the electric heaters; provided temperature control. Two of these coolers coneisting of 1/4=-in. X 0.035-in. wall, 304 stainless steel cooling coils, which used air and/or en air-water mixture as coolant, sur- ~rounded the core section-where the maximum nuclear heat generation occurred. The two coolers prOV1ded for countercurrent coolant flow. Coolant for the No. 1 cooler entered at the top. of the core sectlon and exlted at the bot- tom. Coolant for the No. 2 cooler entered,at ‘the bottom and exited at the top. Both coolihg coils were wrapped around the core section in machined ' _grooves, tack welded at each end to hold the coil in place, and then bonded -to the Hastelloy N core body with nickel-sprayed material. . Another cooler consiéting-of 8 3/16-1n. OD X 0.035-in. wall, Inconel cooling coil, which used only air as coolant, was used on the cold leg. For the gas separation tank an annular jacket cooler of 1/16-in. thick 304 14 stainless steel with eir as the coolent medium wfis used. Although calcu=- lations indicated that this cooling method would be adequate, air alone proved to be inadeguafie to maintain the temperature of the gas separation tank at temperatures below 600°C and a water injection system was added to the incoming air after in-pile operation had commenced. The heat removal rate of the loop coolers was entirely adequate, except for the gas separation tank as noted above, to remove the 8.8 kw of fission end gamma heat generated when the reactor was at its maximum power of 30 Mw and with the loop in the fully inserted position. Even after the loss of one of the two cooling coils (see below) eround the loop core section, the loop could be operatedkét full power (8.8 kw). 4.2 Problems Encountered During In-Pile Operation Shortly before the addition of uranium to the loop, tests indicated that the No. 1 core cooler was leaking &t a point near the loop (inside the loop container can). This cooler wes removed from service by plugging off both ends. However, by referring to Table 2, it can be seen that tempera- tures.in the top section of the loop core were quite high (up to 784°C in ~the upper fuel passage) because of the loss of the No. 1 cooler.‘ Subse- quent examination of the loop'in the hot cells showed that this cooling coil had broken at the point of attachment to the core body on the exit end. A photograph of the bresk is shown in Fig. 8, and a phétomicrograph of the break is shown in Fig. 9. After the uranium addition, the reactor was brought to 30 Mw and the loop was 1nserted in incremental steps over a period of ~160 hr in order to measure nuclear heat generation and to test operation. After ~92 hr of operation in the fully inserted position, a lesk in the cooling Jjacket around the gas Sepafation tank (using an air-water mixture as coolent) allowed water to enter the loop container causing erratic temperatures in the bottom part of the loop — eSpecially at the salt semple line. Subse- guently, fiater entered the smsll equipment chamber at the face of the resctor shielding where it wes detected by @ water level probe. The reactor was shut down, the equipment chamber opened and dried, and reactor operation resumed. Because of the lesk in the jacket cooler, water - Photograph of 15 Brdkén_Coolant Coil, Molten-Salt 16 Fig. 9. Photomicrograph of Coolent Coil Bresk, Molten-Salt Loop No. 1. - 17 injection could not be used in the cooling air to the gas separation tank, and this loss of cooling capacity limited the loop operation to a position at 1 in. retracted (~70% of full power). A few hours after reactor startup the loop was retracted out of the high flux region ‘and the salt was frozen when high radlationrlevels were observed in the charcoal trap in the loop container off-gas line — indicating fission—product leakage from the loop. Preparatibns for loop removal were begun, and on August 10, 1966, the reactor was shut down and the loop removed from besm hole HN-1. Examination of the loop in hot cell facilities showed that fuel salt " had lesked from the loop at a break in the salt sample line near its p01nt _of attachment to the loop core section. It appears that this failure can be attributed to the water leakage described above, which caused the nickel spray bonding the heaters. to the salt sample line to break, leaving the _capillary sample line (0.100 in. OD X 0.050 in. ID) unsupported. This line then falled because of excessive mechanlcal stresses. Table 3 is a tabulation of the component failures which occurred during 1n—p11e operation of loop No. 1.,‘ Table 3. Tabulation of Component Failures Observed in In-Pile Loop No. 1 ,Descriptioufof:Failure" Material -t.Probsbie Cause Break in & capillary tube in @~ Hastelloy N Mechanical stress the sampllng and addition : ' system : Break in l/4-1n. tublng used 304 SS Unknown, but probably for loop cooler No. 1 at point _ assoclated with mechanical of attachment to -loop S - forces from thermal expan- . sion of cooler discharge 1ine : Ieak in seal weld of cooler - 304 85 ~ Unknown, but probably due Jjacket around gas separatlon o ~ to peor quality of seal _tank R _ Lo o weld Break in salt sample line near ‘Hastelloy N Mechanical stress resulting - point of attachment to loop == = from loss of support when nickel spray matrix sur- - rounding line cracked off 18 5. DESCRIPTION AND OPERATION OF IN-PILE SALT LOOP NO. 2 5.1 Loop Description The design of the second in-pile salt convection loop was essentially identical to the first loop experiment and is shown in Fig. 10. Problems encountered in the first experiment, described previously, and subsequent poétirradiation hot-cell examination led to modifications to the second loop which fiere designed to eliminate these problems. These modificationa are described below. 5.1.1 Core Cooling Coils The material for the i/4-in. OD X 0.035-in. wall core coolant tubes was changed from 304 stainless steel to Inconel for the second loop. Al- though stainless steel tubing gshould have been entirely adequate for the intended service, InéOnelris the preferred material for exposure to high- temperature steam (~400°C) generated when air-water mixtures are used as coolant. Since the rupture of the No. 1 core cobler,occurred ad jacent to a point where the tube was tack welded to the core wall, the tack weld was éliminated in favor of a mechanical strap attachment. Further, an expansion loop to relieve stresses was included in each of the coolant tube outlet lines. A test of the adequacy of the modified core cooling coils was made by_bperating a moékup of the rédesigfied coil with air-water mixtureé'as coolant for more than 400 hr at temperatures expected in-pile, including 120 thermal shock cycles (600-350°C), with no sign of difficulty. Thermal c&cling occurs during a reactor setback and startup, and it was estimated that no more than about 20 such thermal cycles would oceur dfiring a.year of operation. 5:1.2 Salt Sample Line The two failures which occurred in the capillary tube (0.100 in. OD X 0.050 in. ID) used in the salt sampling and addition system of loop No. 1 resulted from excessive mechanical stress. Consequently, the wall thick- Hless of these tubes was increased (0.170 in. OD X ‘0.060 in. ID), and sddi-. ctional mechanical support was added - particularly on the section of the line for & distance of ~9 in. from its pbint'of'attachment to the loop core. Fig. 10. \ 19 Photograph of In-Pile Molten-Salt ILoop No. 2. 20 5.1.3 Cooler for the Gas Separation Tank The cooling Jacket of l/l6-in.-thick stainless steel surrounding the reservoir tank used in the first loop was replaced by an Inconel tubé wrappéd around the outside of the tank and attached by means of sprayed-on nickel. Also, provisions for use of an asir-water mixture as coolent were added since it was found that air alone did not provide sufficient cooling in the first experiment. 5.1.4 Salt Flow by Convection Continuous salt circulation by thermal convection was hot maintained in the first experiment, and salt flow rates of 5 to 10 cm?®/min were sub- stantially below the calculated rate of ~45 cm3/min for a temperature dif- ference of 100°C between the salt in the loop core section &nd in the cold leg. It was concluded that the occasional loss of salt circulation was caused'by gas accumulation in the top of the core section where salt from the eight l/4-in. fuel passages in the graphite was collected in an annular ring before entering the gas separation tank. The low sélt flow rate could also be partially attributed to flow restriction caused by the design of -the top and bottom salt flow passéges in the graphite core. Accordingly, the salt flow channels at the top and bottom of the eight 1/4-in. holes in fhe graphite core were redesigned to provide a better flow pattern as shown in Fig. 11 which can be compared fiithithe original design (Fig. 3). Further, thé top. and bottom of the core section, horizontally oriented on the first loop, were inclined at 5° to minimize trapping of gas. 5.2 Operation of In-Pile Salt ILoop No. 2 5.2.1 OQut-of-Pile Test Operation The loop package was operated in the out-of;pile mockup facility for 248 hr. In order to rembve potential conteminants such as oxygen, water, etc., the empty loop was flushed with argon gas and vacuum pumped at 600°C " for 20 hr. The loop was charged with solvent salt (without uranium) end roperated for 77 hr and then drained to flush the loop. A second charge of - - - 21 .. . . w “ PHOTO 74244 22 golvent salt was subsequently added, and operation at temperature was con- tinued for an additional 171 hr. During these mockup operations, 15 salt samples were removed from the loop and;12-sa1t édditions were made. No. problems were encountered. ' | Salt circulation by convective flow was estimated to be 30 to 40 cc/min ‘as determined both by heat balance measurements around the cold leg_return line and by adding an increment of heat in a stepwise fashion to one point in the loop and recording the time required for the heated sa;t to traverse '_ & known distance as recorded on thermocoupleé around.the.loop'circuit. This fldw rate is a five-fold increase over the rate observed in loop No. 1 and is attributed to the modifications described previously. However, occasional loss of flow still occurred. One possible explanation for this is that a sufficient temperature.difference was not meintained between the salt in the hot and cold legs. This is supported by the fact that flow, when lost, could be restored by adjusting the tefi@erature around the loop circuit. Since occasiohal flow loss did’not adversely affect the loop operation in-pile, this was not éonsidered to be a_pfpblem of sny serious consequence. After satisfactorily completing out-of-pile testing, the loop was trans- ferred to the ORR, installed in beam hole HN-1, end in-pile operation was begun on January 12, 1967. 5.2.2 In-Pile Operatidn of Loop No. 2 | At the start of in-pile operation, the loop contained the solvent salt charged to the loop during the mockup operation. In-pile operation with this solvent salt continued for 417 hr during which the loop was operated at various distances from the reactor lattice to determine operating param- eters and to measure gamme heat generation. Uranium ass 'IiF-UF, eutectic (93% enriched) was added on January 30, 1967, to bring the uranium concentra- tion in the salt to 1.72 mole %, which was expected to. produce afi average fission-power generation of 200 w/cm® in the 43 cm? 6f fuel salt in the core ‘section. This estimate was basel on an expécfiéd average thermal neutron flux of ~2 X 10*? in beam hole HN-1. Subsequent meesurements of the nuclear power ‘generation as 8 function of distance from the reactor léttiée gave 8 value of 165 w/cm for average core f1831on-power density with the loop fully inserted (Fig. 12). This indicated that the flux was ~1 X 10%3, NUCLEAR HEAT (kilowatts) 23 ORNL-DWG 67-10155 20 10 FISSION AND GAMMA nN FISSION - o o 0.2 041 _ 2.3 4.3 6.3 8.3 10.3 423 LOOP POSITION, DISTANCE FROM REACTORV TANK TO CENTER. OF GRAPHITE CORE (in) Fig. 12. ‘Nuclear‘Heat'Genera.tion‘in Molten-Salt Loop No. 2. 24 In order to inérease the fissioh-power density in the loop, a second eddition of enriched uranium (as ’IiF-UF, eutectic) was made on March 7, 1967, to bring the uranium concentrétion in the fuel salt to 2.1 mole %. This addition was expected to increase the average fission-power density to the desired fialue of 200 w/cm3 average in the loop core (grafihite) region. However, as a result of a rearréngement of the fuel loading in the ORR Jjust prior to the second addition of uranium, there was essentially no. increase in fission power. This rearrangement of the reactor fuel reduced the thermal flux in beam hole HN-1 in an amount sufficient to compensate for the increased uranium in the fuel salt. Previous rearrangements presumably also accounted for the lower than anticipated neutron flux observed initially. Loop operation was continued and the ORR was brought to full power (30 Mw) on March 11, 1967. On March 14 the reading of the radiation monitor on the charcoal trap in the loop container sweep gas line had inéreased-to' 18 mr/hr from the normsl level of essentially zero. Some 8 hr later the rédiation level had increased to 3.4 r/hr. This reading did not increase further until March 17 when it increased rapidly (over a period of ~3 hr) ~to ~100 r/hr which indicated leakage of fission products from the loop into the container can surrounding it. At this point the loop was'retfacted ofit of the high flux and the loop temperatures were reduced to ~400°C_to'freeze the salt. This caused the radiatidn in thelcharcoal trap to decrease to "~1 r/hr over a 15-hr period. ‘ | - From March 17 to March 23, 1967, the loop was operated in a position where the flyx levels were 1 to 24 of that when the loop was fully inserted. Durifig this fime, the fuel salt was kept frozen (~400°C) except for brief periods when it was melted in an attempt to locate the point of leakage. It was concluded.that the leak was in the vicinity of thé‘gas separation tank and continued operation of the loop was not poséible. From March 27 to March 31, 1967, the fuel salt was drained from the - "loop by sampling. By this procedure, requiring removal of 10 samples (12 to 25 g per sample), the loop inventory was reduced from 216.8 g to 2.1 g. The ORR was shut down on April 4, 1967, and the loop removed from beam hole HN-1 and transferred to hot cell facilities for examination. . Hours of operation with both solvent salt and fuel salt and with the ORR at zéro power and 30 Mw are tabulated in Table 4. Typical opersting 25 temperatures around the loop with fuel salt and solvent salt when the loop was in the maximum thermal flux position are shown in Table 5. Figure 13 shows the location of thermocouples on molten-salt loop No. 2. Table 4. Operating Time Under Various Conditions for In-Pile Molten-Salt Loop No. 2 Operating Time Salt in Loop Reactor Eower ) Solvent® 0 3250 , 30 Mw 341 Fuel® - 0 168 30 Mw 1369 2203 ®Solvent salt composition = 7LiF-BeF,-ZrF, (65.7-30.1-5.2 mole %). bInclu.d.es 248 hr of out-of-pile mockup oper- ation. w Cpuel salt composition = 7LiF-BeF,-ZrF,-UF, (6503-28 02-4 08-1-7 mOle %) and (65 -4"'27.8"4 08-201 mole %). | 6. EXAMINATION OF FAILURE IN CORE OUTLET PIFE Fdllowing its removél-from the reactor, the loop package was trans- ferred to hot-cell facilities where the convection loop was removed from ~its container cafi for éxémination. No evidencé of salt leakage from the loop was seen by visual Examinatiofi. ‘The loop was then firéssurized to ~100 psig with helium and "Leak Tec" solution applied to the external sur- faces of the loop{r'By"thié-teéhn1QUe a gas leak was observed in the core outlet tube sdjacent to the point where it was attached to the core body. - Pigure 14 is a photograph;cf the loop takeh_in the hot Cell and indicates the point where the'gas'léSRVWEé seen . SubseQuently;'sectioné'of the loop were cut out for metallographic examination and a crack through the wall Table 5. Typical Loop'Tbmperatures with Solvent Salt and Fuel Salt with the ORR at 30 Mw for In-Pile Molten-Salt Loop No. 2 Core Section ~ Gad Separation Tank Return Line (Cold Ieg) Thermocouple Location? _ ‘ Number Solvent Salt Fuel Salt Solvent Salt Fuel Salt Solvent Salt Fuel Salt ' | Temp, °C Temp, °C Temp, °C Temp, °C Temp, °C Temp, °c 1 Core bottom - 590 - 535 2 Lower fuel passage - 619 - 588 3 Upper fuel passage 661 . 655 4 Graphite center _ 651 543 5 Graphite OD 648 | 510 6 Core top - 698 - 672 7 Core outlet pipe 668 730 _ 8 Inlet well o 524 543 9 Outlet well - 577 578 10 Gas space ' 529 450 11 Top - 535 . 535 544 12 Center | - 548 _ 548 560 13 Bottom . 586 | 586 617 14 At core inlet 575 ' 575 577 | aRefer to Fig. 13. A 27 ORNL-DWG &7-10156 | 73/, in. 2’ [ o | GAS LINE GAS LINE- -~ 10 L EXPANSION TANK SPRAYED NICKEL HEATER __ COOLING COIL ‘COLD-LEG * THERMOCOUPLES (2) aND (3) ' WERE IN TWO DIFFERENT FUEL ~ CHANNELS : 20° 14 LT SAMPLE LINE Fig. 13. Thermocouple Location for Molten-Salt Loop No. 2. 28 Fig. 14. Postirradiation Photograph of Molten-Salt Loop No{,2 Showing Location of Leak in Core Outlet Pipe. 1Y -t 29 of the Hastelloy N pipe (0.406 in. OD X 0.300 in. ID) was found. Figure 15 is a photomicrograph of the crack which extended almost completely around the circumference of the pipe. There was no evidence that fuel salt had leaked through the crack and only gaseous fission products had escaped. 7. DISCUSSION AND CONCLUSIONS The four fallures encountered durlng 0peratlon of loop No. 1 were ~ examined in hot—cell facilltles. Based on this examlnatlon and the oper- ating history of the loop, correctlve measures were taken in the design and construction of loop No. 2 (refer to Section 5). None of these failures were encountered during the operatlon of in-pile salt lOOp ‘No. 2. Analysis of the causesrof the failure of the outlet pipe in loop No. 2 had led to the conclusidn'that this failure was probably caused by excessive stresses resulting from differentlal thermal expanS1on of the loop compo- nents (core, cold 1eg, gas separatlon tank, and outlet plpe) Computer code MEC-21 (ref. 1) was used to determlne the stresses developed due to the thermal expansion of the piping system. Calculations of the piping stresses in the.loop'werermade for two-conditions: (l)lfor_the_temperature profile around the loop at -full power operating cOnditions,'and (2) for the temper- ature profile changes observed ‘during & reactor setback._' For both condltions (1) and @) the piping stress analy51s indicates that the max1mum stress from thermal expansion occurs in the core outlet pipe where the failure occurred. For the normal operating condition the bending_momefit produces a stress of ~10,000 psi in thegbipé wall (tension on the top and compression'on'the bottom). For.the temperatures encountered during a reactor setback' the direction of the bending moment is reversed causing a stress of ~17, 000-psi in the pipe wall (compre351on on top and tension on bottom). The entire loop was. fabrlcated on Hastelloy N (INOR- 8) which is also the material used for the MSRE. Materials used in the loop were obtained | from the MSRE stock of speciallyfordered heats of Hastelloy N. Data on the 1James H. Griffin, A Piplng Flexibility Analy81s Program, LA-2929 (July 1964). 30 TOP OF CORE CORE QUTLET PIPE Fig; 15. Photomlcrograph of Crack in Hastelloy N Outlet Pipe, Molten-Salt Loop No. 2 (~200X) ¢ AT 31 properties of~Hastelloy N at temperatures of interest in the Molten-Salt Reactor Program and the effect of irradiatippion these properties have been summarized by R. B.,Briégs,2 | Date contained in the re'ferenced_report2 indicate that,:finfa tempera- ture of 1200°F (650°C) and for en irradiation dose of 5 X 1019 nvt, stresses of 8000 to 10,000’psi.would-produce fupture after 10,000 hr. Stress-rupture properties of Hastelioy N at the 1350°F (732°C) temperature of the core out- let pipe and efter an irrediation dose of 5 X 10° nvt are below those at the temperature of 1200°F (650°C) used for design purposes.’ For the in-pile fiolten—salt loop there are no significaht primery stresses since the loop is operated at or near the ambient pressure (loop pressure is maintained between 12 and 20 psie). Thermel stresses, although important, are usuvally of less concern because once encountered they tend to be self-limiting provided the material has sufficient ductility. How- ever, tests indicate that the_ductility of Hastelloy N is reduced such that strains of 1 to 3% can result in fracture at temperatures of 1200 to 1300°F and an irrediation dose of 1 X 10%° ‘nvt or more. For the design of the in-pile salt loop, thermal stresses in the core wall (Hestelloy N) and in the core cooling coil (304 stainless steel for ‘loop No. 1 and Inconel for loop.NO. 2) were evalusted. Based on heat flow “at 10 kw of nuclear heat gerieration, these stresses were estimated to be | about 10,000 psi. For the core wall at 650°C and & dose rate of up to 1 X 10*3 n/cm?-sec, this thermal stress was considered acceptable for oper- ating times to 10,000 hr or more than the one year projected as the:maximum time of in-pile 0peration;- Stresses caused by differential_thexm@l expansion of the loop pipe were'not.galcu1afed prior to in-pile opefgtion; For normel operating conditions temperature differences of 50 to 100°C around the loop circuit did not. seem sufficient to produce undue stresses. In particuler, 'A no evidence of stress or any bther failure was observed during in-pile oper- etion of'lodp No. 1. 2R. B. Briggs, Effects of Irradiastion on Service Iife of MSRE, ORNI~CF-66-5-16 (May 4, 1966). 3H. E. McCoy, Jr., and J. R. Weir, Jr., In- and Ex-Reactor Stress- Rupture Properties of Hastelloy N Tubing, ORNI~TM-1906 (Sept. 1967). 32 ft now appears that several factors could_have caused the failure in 'the core outlet pipe. First, the temperature of the section of pipe where failure occurred was at a temperature of ~1350°F (732°C). Thus, a thermal stress of ~10,000 psi celculated to exist in the outlet pipe may have been sufficient to cause failure. A second and more likely cause of failure is ‘the rapid stress reversal (+10,000 péi to}—l7,500 psi)~calculated for the thermal shock causea by a reactor setback. Approximately six such cjcies were encountered;during-in-pile'operétioh. In particular one such cycle occurred on March 3 after a dose accumulatioh.of ~2 X 10*° nvt. It was on March 11 that evidence_of fission-product leaskage from the 1oop was first observed. Whether or not such thermal cycles caused the failure is specu- lative, but the stress reversal resultingAfrom such cycles would certainly Vappear likely to'Contribute to failure at the point of maximum stress where the temperature was 1350 F and after accumulation of a radiation dose suffi- - - cient to affect the strength and ductility of Hastelloy N. Some ‘thought has been given to possible design changes that mlght elim- inate or at least reduce thermal stresses — possibly expansion joints for exampie. H0wever, such designs, as well as the present loop configuration, requite a material with more strength and ductility than the present Hastélloy N possesses at temperatures and radiation doses anticipated for meaningful in-fiilevloop eXperiments. Therefore, @ material exhibiting “bettbr physicél properties under these conditions is needed for future in- pild loops designed to obtain data at high fuel fission power and long-term ope?ation. Improvement in the physical properties of Hastelloy N — espe- cially improving its resistance to neutron irradiation — is being given ma jor attention.* Work has shown that additions of titanium, zirconium, " and hafniwm vill reduce the radiation damage of Hastelloy N. In-pile and out-of-pile tests are being run on these modified Hastelloy N alloys. To date, laboratory-size vacuum melts and small 100-1b commercidl melts are being evalqeted. A commercial melt of an improved Hastelloy N containing 1/2% Ti addition has been ordered, apd it is anticipated that this matbrial will_be used for the next in-pile'eenvection loop assembly. | - “H. E. McCoy, Jr., and J. R. Weir, Jr., Meterlals Development for Molten-Salt Breeder Reactors, ORNL-'IM-1854 (June 1967). C ot 33 ACKNOWLEDGEMENT Credit is due to Mr. C. W. Collins of the Reactor Division for assis- tance in the thermal stress analysis of the loop piping.- ¥ W '-—l. ovooIowumdMhW 11. 13. 14. 15. 16. 17. . 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 61-75. 76. 35 INTERNAL DI STRTBUTION F. Baes M. Beker F. Bauman E. Beell Bender S. Bettis F . Blankenship G. Bohlmann E. Boyd B. Briggs . W. Collins L. Compere H. Cook L. Culler H. Devan Ditto Epler Ferguson Grimes Grindell o - Kasten Kedl Kelly King Kirslis Tamb Litmen PONHESINPEONNGIENONOEEER RGO UHREN NP EEE D Haubenreich 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40-41. 42. 43. 44 . 45. 46, 47. 48. 49. 50. 51. 52, 53-54. 55=-56. 57-59. 60. SRR EEDEETE L R L. . I. E. W.. . McCoy -McDuffie aERaprpnAs E. " ORNI~TM-1960 Long Tundin MacFPherson McClung Meyer Moore Nicholson Qakes ' Perry Rosenthel Savage Sevolainen Scott H. J. E. M. R. E. C. I. Shaffer Skinner Thoms Watson Weir Whatley White Wyatt Central Research Library Document Reference Section Ieboratory Records Ieboratory Records (IRD-RC) EXTERNAL DISTRIBUTION Division of Tbchnical_Ihformation Extension (DTIE) Iaboratory and University Division, ORO