Contract No. W-T7405-eng-026 ORNI~4371 CHEMICAL TECHNOLOGY DIVISION Pilot Plant Section PREPARATION OF ENRICHING SALT 7LiF-ESBUF‘ FCR REFUELING THE MOLTEN SALT REACTCR Jchn M. Chandler S. E. Bolt¥ MARCH 1969 *Reactor Division. OAK RIDGE NATTONAL LABORATORY Cak Ridge, Tennessee operated by UNICN CARBIDE CCRPORATION for the U. 5. ATOMIC ENERGY COMMISSION LEGAL NOTICE This repori was prepared as an account of Government gponsored work, Neither the United States, por the Commission, nor any person acting cn behaif of the Commissicn: A, Makes any warranty or representation, expressed or implied, with respect to the accu- racy, completeness, or usefuiness of the information contained In thig report, or that the use of any information, apparatus, method, or precess disclosed in this report may not infringe privately owned rights; or B. Assutnes any liabilities with respect to the use of, or for damages resulting from the use of any information, apparstus, method, or process disclosed in this report, As used in the above, ‘‘person acting on behaif of the Commigeion® includes any era~ ployee or conmtractor of the Commission, or employee of such contractor, to the extent that such employee or contractor of the Commission, or employee of such contractor prepares, disseminates, or provides access to, wny information pursuant o his employment er contract with the Commigsion, or his employment with such contractor. iii CONTENTS ADSTYECT & ¢ o « o ¢ o s & © 5 & o e s s & s & o 15 2. 71 Tntroducticn o o ¢ ¢ o o ¢ 6 o s & & o s e & & s o2 233 in the MSRE 233 Considerations for Substituting 2.1 Nuclear Characteristics of the U System 2.2 Chenistry of the Circulating Fuel . « . . 2.3 Fuel Preparation by Remote Means . . « . -« Process Development .« ¢ « ¢ & « o ¢ & ¢ 5 o o 3.1 High~-Temperature, One-Step Process . . . . 342 Low-Temperature, Two=-Step Process .« « « o Description of Salt Production Facility . .1 Descripticn of the TURF . . . « o & « & ll‘elcl Cell G @ s - @ @ - * e 3 - - & & ® 4,1.2 Redioactive Hot Drain — Hot Off-Cas (RED-HOG) v v v v ¢ o o ¢ o« v o o & k,1.3 Cell Ventilation . « « « ¢ « o & & h.1.4 Shielding o « ¢ ¢ o ¢ o ¢ o s h,1.5 Manipulators . « « « o o « « o o . L.2 Process Flowsheets . + « o « ¢ « « ¢ ¢ « k.2.1 High-Temperature Process h,2.2 Low-Temperature Process . . 4.3 Process Equipment . . . « « ¢ « ¢ « o 4.3.1 Decanning Station . . . . « . . . 4.,3.2 Reaction Vessel « v« ¢ o« & o o o « & L.,3.3 Salt Storage and Transfer Vessel . 4.3.4 Shipping Containers . - « « « + & 4.3.5 Scrubbers . . « « o ¢ o ¢ o & 4.3.6 Miscellaneous Equipment . . . h.h Operating Procedures . o o ¢ o o ¢ ¢ o o o Cold Run with Depleted Uranium . . . ¢ o « ¢ ¢ & 5.1 Preoperational Activities . . . « . . . 5.2 Mechanical Operations . . « « ¢ ¢ + « & a ® & & System @ s E__l W W o o = W "N 10 11 11 13 LA W O ON 8. 9. 1C. 5.3 Chemical Operations « ¢ o ¢« ¢ ¢ o o ¢ o o & ¢ o a 7LiF-233UFu Process Operations ¢ o o« s o ¢ o ¢ o o & o = 6.1 Feed MaterialS « o « v o 4 o o o o o o & 8 o o o .2 Radiation Levels of the Oxide Feed . . o « « « o & £.3 Oxide Feed Material Handling o« « « « o o ¢ « o 6.4 Reduction of Uranium OXide « + ¢ o s ¢ ¢ o o« o o a 6.5 Hydrofluorination of U02 e e e « o e+ 5 & e s s o e 6.6 Formation of the Eutectic Salt ¢ « ¢ o « & « o o & 6.7 Purification of the Eutectic Salt . . . . . . . . 6.8 Transfer of the FButectic Salt .+ « ¢ o « « o o & 6.9 Container Disassembly and Preparation for Shipment . 6.10 Transfer of |LiF-“3°UF, Salt Product to the MSRE . 6.11 Material Balance . « o« « « o o o o o o o ¢ o s & @ iv Maintenance Engineering .« « o s o o s o o ¢ o o« o o 7.1 7.2 7.3 7.k Spare Parts .+ v ¢ o s v s o 6 & s s s e ¢ o o o o Redundant Fittings ¢ ¢« o « ¢ ¢ o « o o o s o ¢ s o Tools and Work Tables o . « ¢ ¢ o o ¢ o o o o o Maintenance ProceduresS « ¢« o o « o o s o s o o ¢ o COnCluS ions ° @ ® € e o *® ¢ o @ ¢ ® ° @ - e o & L ° o * Acknowledgments . o o ¢ o o o« © o a o ¢ o s o ¢ o & o o Referenc eS ® o o € [ ® & . ¢ ® & © * ° & @ e o & € e s PREPARATION OF ENRICHING SALT 7LiF—ggBUFLL FOR REFUELING THE MOLTEN SALT REACTOR - John M. Chandler 8. E. Rolt ABSTRACT The Molten Salt Reactor has been refueled with an enriching salt concentrate, 7LiF=233UFq (73-27 mole %). Sixty-three kilograms of this concentrate was prepared in cell G of the Thorium-Uranium Recycle Facility at ORNL. Its preparation in a shielded cell was required because of the high 232U content (222 ppm) of the 233y, In the shazkedown run with depleted uranium oxide, a high-temperature, single-step process was used to reduce the oxide and then convert it to UFL for use in making the eutectic salt. Although this process ylelded a high- quality product, severe damage to the equipment was ob- served. Therefore, 1t was discarded in favor of a low- temperature, two-step process in which the uranium oxide was reduced to U0, by treatment with hydrogen, the UOQ was converted to UFj, by hydrofluorination, LiF was added, and the eutectic was formed by fusing the components. The eutectic mixture, LiF-UF) (73-27 mole %), was purified by treatment with hydrcgen, which reduced the corrosion prod- ucts to metal and subsequently allowed their removal by filtration. The quality cf the precduct was well within the requirements established for the MSRE. The fuel concentrate, containing 39.0 kg of uranium (91.4% 233U), was packaged in nine variable-capacilty (0.5 to T kg of ursnium) shipping containers for addition to the reactor fuel drasin tank and in 45 enrichment cap- sules, each containing 96 g of uranium, for addition to the bowl of the fuel circulating pump. The fuel was shipped in shielded carriers to the MSRE to accommodate the reactor enrichment schedule. 1. INTRODUCTION In July 1966, an ad hoc committee was appointed to study the 235 feasibility of substituting 233U for the U being used to fuel the Molten Salt Reactor. It was the recommendationl of this committee that 233 a charge of eutectic salt containing U be prepared for enriching 235 U The MSRE prior to removal of the from the system. o A single-step process, in which the fuel concentrate could be prepared directly from 233UO and LiFF, in a single reaction vessel 3 located in a hot cell, was considered to be the simplest and most economical appreoach. Bilological shielding would be reguired because 232 233 of the high U content (222 ppm) of the U feed material. The reduired quantity and quality of the enriching concentrate will permit operation of the MSRE st full power for at least one year. This report summarizes the preliminary phases of the wcrk — development, design, and construction — as well as the actual operation ' e 2 7L e 33UF of the process used to prepare the 1 salt. i (73-27 mole %) eutectic 233 2. CONSIDERATIONS FOR SUBSTITUTING U IN THE MSRE 2 35, . . The substitution of 33U for E“SU in the MSRE had tc be considered from three standpoints: (1) the nuclear characteristics of the 233U j system, (2) the chemistry of the circulating fuel, and (3) the preparation 233y 232y and handling of & salt containing 233 2.1 DNuclear Characteristics of the U System An assessment2 of the operation of the Molten Salt Reactor with 233U has indicated that the following could be learned by substituting the new fuel charge: (1) Critical losdilngs would provide a check on the available 233 riuclear data on U for predicting critical conditions in a reactor with a neufron energy spectrum similar to that of the proposed molten salt breeder reactor (MSBR). (2) Measurement of aguU—tom£33U atom ratios over a periocd of sub- stantial burnup might further evaluate available nuclear data and calculationsl methods. (3) Interesting differences would bhe observed in the dynamics cf the reactor because of the smaller fraction of delayed neutrons G L 233 2 available from U fission. Stable operation with 23“U in the MSRE would tend to confirm tentative conclusions regarding the dynamics of the MSBR. ;fig (4) EKnowledge of the fission product yields for 233 U, as compared with those for 235U, would provide a means of more pesitively identifying the resctivity transients that follow changes in power levels. (5) Changes 1In other nuclear characteristics, such as temperature coefficients, neutron lifetime, and control recd worth, would be observed. 2.2 Chemistry of the Circulating Fuel The uranium concentration in the reactecr will be reduced to 0.2 mole %; however, nc significant increase in UF_ concentration or precipi- 3 tation of ursniuvm or U02 1s expected. Most of the properties of the new fuel charge will be very similar to those of the 23?U fuel. No change 1s expected in the compatibllity cf the fuel with the graphite and the Hastelloy K. 2.3 Fuel Preparation by Remote Means A strong incentive exists for develcoping an economical method for reprocessing irradiated fuel for reuse in a power reactor. The high 232 233 U content of the U will make remote processing of a 233U fuel charge mandateory. Therefore, it 1s desirable to prove the feasibility of a simple, direct fuel recycle process without the necessity of high-level decontamination; this would emphasize the advantages of the fuel preparatlon for molten salt reactors. 3. PROCESS DEVELOPMENT The preparation of the enriching salt, 7LiF°233UFa (73-27 mole %), 7 from "LiF and 233UO3 by a direct high-temperature, one-step process, was 2 investigated” in the laboratory and Tound to yleld & satisfactory product. However, during the cold run in Building 7930, the high temperatures necessary for this process were found to promote corrosicn of the equip- ment; thus, an alternative method — a low-temperature, two~step process — was adopted. These processes are discussed in detail in Sect. k.2, 3.1 High-Temperature, One-Step Process The chemical procedures that were developed for the production of the fuel concentratie were similar to those used in the routine prepsara- tion of UFLL from cxides. Necessary modifications included: (1) the use of a vertically mounted, cylindrical reaction vessel rather than trays or fluidized-bed reactors, (2) the addition of LiF to the initial charge of material, and (3) the operation of the process at temperatures sufficient to maintain the LiF in its molten state. Iaboratory-scale experiments were conducted by the Reactor Chemistry Division tc gain irnformation about the rates at which the reactions would occur and alsc to examine possible process control techniques. The reactions investigated wvere: Sintering: Helium atmosphere at 900°C U0, — U0, ¢ + 0.2 O, Reduction: Hydrogen at 900°C U0, ., + C.6 H, — U0, + 0.6 £ Eydrofluorinaticn: Ho-HF sparge at 900-550°C uo, + LHF — UF) + 2H,0 Reduction of impurities: Hydrogen sparge at T700°C MF2 + H2 — MC + 2HF In the reduction step the progress of the reaction could be followed by observing the generaticn of water vapor in the system. The temperature of the gas effluent increased markedly during this periocd. During the hydrofluorination trestment, hydrogen, along with anhy- drous HF, was admitted to the reaction vessel to control the corrosion of the vessel and also to ensure the complete reduction of UO3 to UOE’ The conversion step cf the process was carried out at temperatures sufficiently high to keep all of the LiF in solution. Thus, as UFM was < produced, the liquidus temperature of the fluoride components decreased from 845 to LO0°C, the melting point of the eutectic mixture. In the equipment used in the development runs, the Iiguidus temperature could be measured and the UF& concentration could be determined by comparison with the phase disgram. A final treatment of the melt with hydrogen at T00°C reduced the concentration of nickel and iron to acceptably low ievels. Uponn successful completion of the laboratory-scale experiment, a cold run, using equipment thet more clesely resembled the production eguipment, was conducted by the Reactor Chemistry Division. This run served to further confirm the feasibility of the one-step process and also to train key personnel for the production operatilions at the Thorium-Uranium Recycle Facility (TURF). 3.2 Low-Temperature, Two-Step Process As an slternative to the high-temperature method, a two-step, low-temperature process was evaluated in a labecratory experiment and subsequently used in the salt production runs. In this process, which is similar to that used to produce the original fuel salt for the MSRE, the temperatures are generally much lower, and the LiF is not added until the conversion of the oxide to the fluoride has been completed by hydrofiucrination. 4. DESCRIPTION OF SALT PRODUCTICN FACILITY The processing equipment was installed in cell G of the TURF, which is located in the Melton Valley area of Czk Ridge National Iaboratory. This faci_litylp contains shielded cells and process supporting systems to permit remote fuel reprocessing. 4.1 Description of the TURF The TURF, Building 7930, was constructed at Ozk Ridge National Iaboratory to help develop and demonstrate economical remote methods 6 for reprocessing irradiated thorium-based fuel and for refabricating the purified fertile and fissile material into fuel suitable for reuse in a power reactor. The use of shielding and remote fabricaticn methods will permit the use of simplified processes yielding only modest decon- tamination factors. The TURF has sufficient space to accommodate equipment for process- ing and fabricating two types of fuel assemblies simultanecusly. The facility is divided into four major areas: (1) an office area adjacent to, but isolated from, areas that contain radioactivity, (2) an operating areea with a development laboratory, chemical makeup area, and equipment rooms for service equipment, (3) a maintenance operating area with ser- vice areas for recelving and storing spent fuels, and (4) a cell complex containing seven hot cells (six shielded and one unshielded). Included in the facility are the services, ventilation systems, crane and manipulator systems, viewing systems, and liguid and gaseous waste disposal systems necessary to support fuel reprocessing. L.1.1 Cell G The 233U fuel charge for the MSRE was prepared in cell G. The interior of this cell 1s 20 ft wide, 16 ft long, and 30 ft high. A false floor was instslled to elevate the process equipment so that the through- wall master-slave manipulators could bve used and maximum advantage could be taken of the viewing capabilities of the shielding windows. This decreased the effective height of the cell to 22 ft, which, in reality, was Turther reduced to 14 ft of effective headroom because of the in- cell electromechanical manipulator system space requirements at the top of the cell. The walls, celling, and floor of the cell are lined with stainless steel. Six cell operating modules (Fig. 1) are built into the walls of the cell; four of these are equipped with viewing windows, and two have window forms fthat are filled with removable shielding. There are mcre than 100 penetrations into the cell for process services — many more than are required for the salt preparation. The cell venti- lation system normally maintains a pressure of =5 in. HEO and a Tlow of 1033 ecfm of air through the cell; the wvessel off-gas system removes ORNL DWG 68— 12283 1" SERVICE PIPE 8"‘0“ o i MODULE fl ‘ SHIELDING e ey e ot e e s e = PLATE ~ (- i SECOND FL. EL. 863° LR B g A g @ s 6 ;” R 1 1 Mg c MANIPULATORS Talalt Wak N % e MANIPULATOR o A MODE L"A'+ h, FMOREL™D" L SLEEVE " | hajlay Y L - ! O : <+ ! = N o A = SMALLITEMS ENTRY PORT O PERISCOPE SLEEVE ELECTRIC SERVICE SLEEVE ¢ WINDOW _ : Mo 6" 8" SERVICE e Ay | SLEEVE —\\,,;E-,_;_,;__:__- = < A, a DRAIN TROUGH — &~ = =~ —~ ., , FIRST FL.EL.849%y L|° . <« . < Fra . . _ B Al Bl B g Ry 8 g i el Dl R ’ 4. .d' Fig. 1. Typical Cell Operating Module. 25 cfm of air from the cell and process vessels and routes it to a system that is maintained at a pressure of =17 in. HgO° A small entry port permits introduction of small tools and miscellaneous items through a glove box and an alr lock. The rocf of the cell contzins a hatch with a 10- by 6-ft opening that is sealed and shielded. The hatch provides access to The cell with the 50-ton building crane. Eight master-slave manipulators were installed in the cell: Tfour Central Resesrch Model A and four Central Research Model D units. A Program and Remote Systems Model 3000 electromechanical manipulator system is mounted on rails to give complete coverage of the cell. 4.1.2 Radicactive Hot Drain — Hot Off-Gas System (RHD-KEOG) The radicactive hot drain--hot off-gas system 1is a combination by- product waste collection system and vessel off-gas system. It has inlet connections in the hot cells and at verious points throughout the facility. The network of stainless steel piping is designed to handle gas and 1liguld in concurrent flow. Liquid waste 1s separated from the off- gas and is collected in tank B-2-T in the belcw-grade and shielded waste tank pit. The gaseous stream flows through a bank of absolute filters and then into the cell exhaust system located upstream of the filter units for that system. The liquid wastes may be pumped either into the Melton Valley waste system or into cell G for recovery of vaiuahble materials. Two connections from cell G te the HOG system were made for the salt preparation process. One connection was made to the capsule drill- ing station in order to collect the particulate matter resulting from the drilling operation. The other was made teo a multipurpose manifold station. With the HOG system pressure controlled at —17 in. HEO’ the flow of air from the cell was adjusted to 25 ¢fm by a manually cperated valve. The gaseous effluent from the process scrubbing system was dis- charged into this manifold station, thereby ensuring dilution, in the case of hydrogen, to less than the explosive limit. In addition to the scrubber dischsrge, the manifold station served the in-cell titration station, the sampling station, the can-opening bcex, the scrubber system \O ....... and sirk liguid drsins, and all of the process vessels. Bome of these connecticns were mede to the menifold station on the cell gide of the manuval flow regulating valve, where the pressure was —5 in. H_ O; scme 2 were made on the HOG system side of the valve, where the pressure was —15 in. H_.O. The location of each connection depended upon the pressure requirenent for that particular use. 4.,1.3 Cell Ventilation Approximately 1033 cfm of eir enters cell G from the cell G pump room via a series of filters, a fire damper, a back-Tlow preventer, and a cell pressure control valve. It discharges into the cell from dif- fusers that are mounted on the cell ceiling and then flows through roughing filters located at the false floor (2nd also at the cell floor level) into the cell ventilation system marifcld in the north valve pit. The flow is contrclled by a manually icaded valve in the exhaust - duct. The exhaust is routed to the filter pit, where 1t passes through two sets of filters in series; from there, it passes through a 30-in. duct to the centrifugal blowers and into the 250-ft HFIR stack for dispersion into the atmosphere. The controls on the ventilation system maintain a pressure of -5 in. H O in the cell under normal conditiocns. Pressures 2 as low as —20 in. HEO or as high as =1 in. HEO can be maintained in the cell under certain emergency conditions. Only a failure of all emer- gency systems would allow the pressure in the cell to rise above —C.5 in. H20° L.1.4 Shielding The shielding of the TURF is designed in such a manner that, during operaticn with radiocactive material having an intensifty of 105 r/hr, the penetrating dose rates in normally occupled areas are no greater than 0.25 mrem/hr, with small hot spots no greater than 2.5 mrem/hr. Dose rates greater than this are permitted in limited-access areas and for short-term, non-routine operations. 10 To satisfy the allowable design radiation levels, the operating cells have 5=l/2—ft—thick walls of normal concrete up to a height of 11 ft, 4-1/2 £t of normal concrete for the remaining portions of the vertical walls, and 5-ft-thick concrete on the roof. This amount of biological shielding was more than adequate to reduce exposure dose rates to < 0.1 mr/hr. The windows are essentially equivalent, in shielding thickness and in their attenuation of penetrating radiation, to the concrete walls in which they are installed. Each window consists of two major assem- blies: the seal glass that is removable from inside the cell, and the tank unit that is removable from the operating face of the cell. EFach window is a composite unit consisting of 7 in. of glass and 58 in. of zinc bromide solution; 1t is well sealed to minimize leakage of air ground its periphery. 4.1.5 Manipulators A Programmed and Remote Systems Model 3000 Manipulator System is installed on a set of rails in cell G. The tube hoist on this manip- ulator has a vertical travel cf 13-1/2 ft and a 1ifting capacity of 1000 1b. The trolley and bridge travel, along with the vertical travel of the tube hoist, provides complete manipulator coverage of the cell down toc the false flcor. This unit provides a1l the motions of the human arm, plus wrist extensicn and ccntinuous rotation at the wrist and at the shoulder. A grip force of 200 1b can be exerted with the fingers. The hand is remotely removable and can be replaced by a hook fixture or an impact wrench. One Central Resezrch Model A master-slave manipulator and one Model D master-slave manipulator are installed at each of the four viewing windows 1In cell G. The Model A and the Model D manipulator have maximun 1ift capabilities of 25 1b and 100 1b, respectively. These master-slave units were installed to operate valves, to make and break tubing disconnects and electrical and thermocouple disconnects, and to conduct the hand operations required in the process. Conventional hand tools were modified for use with these manipulators. 11 The PaR electromechanical manipulator was installed to perform the heavy~duty maintenance, to convey heavy assemblies around the cell, to reach some parts of the cell that were not accessible to the master- slave manipulators, and to provide a "third" hand to simplify certain operations. .o Process Flowsheets Two processes were developed for the preparation of the fuel concentrate: (1) a high-temperature process, and (2) a low-temperature process. The second process was adcpted after conclusion of the cold run when it vecame evident that the origina’l process caused severe corrosionr of the reaction vessel. 4.2.1 High-Temperature Process The chemical flowsheet for the one-step, high-temperature process as shown in block form (Fig. 2} includes the following major steps: 233 (1) Partial reduction of the uraniun oxide, Uté,by thermal means . (2) Further reduction to U’O2 by hydrogen treatment. (3) Conversion of the oxide to UFu by hydrciluorination and con- current dissolution of the UF& in the molten LiF to form the evtectic mixture. (L) Final purification of the eutectic mixture by trestment with high-purity hydrogen. '’ 233 UO3 in the initial cherge to the reaction vessel. The reduction process was then conducted at tempera- The 'LiF was combined with the tures in excess of 845°C (LiF melting point) so that the oxide particles were suspended in the molten fluoride. The moltern LiF served to keep the oxide particles wet, thus reducing the possibility of entrainment of particulate matter in the effluent gas stream. During the hydroc- fluorination step, the progress of the reaction was estimated by determining the liquidus temperature of the molten material, and then LOADING THERMAL RECOMPOSITION Load reaction vessel with about 13 kg 233y (91.5% enriched) as UO3 and about 4 kg 7LiF. REDUCTION TO U0y ORNL DWG, 67-11637A Purge reaction vessel with dry He at 1 liter per min while heating to $00°C, Partial reduction of UOg to lower oxygen content. Sparge with Ho~He mixture through dip lire. Total gas flows 2 to 10 liters per min, Melt temp: ?00°C, Reduction of oxide to near 00% UQ9, CONVERSION TO UF4 Sparge with HF-Hg through dip line. Hg at 10 liters per min, HF at 1 to 2 {iters per min, Periodically adjust melt temp, to about 50°C above tquidus but not below 700°C, ' MELT ANALYSIS REDUCTION OF MF> IMPURITIES When HFqyt equals HFjy and liquidus tempy ap~ proaches 490°C, with=~ draw filtered sample of melt and submit for chemi~ cal analysis, HF STRIPPING Increase melt temp, to 700°C and sparge with H2 alone at 5 to 10 liters per min until HF in gos effluent becomes <0,01 milliequiv per Hter Ho. Sparge melt with He at § liters per min for about 2 hours fo remove residual HF and Ho. SALT TRANSFER Note: If results of chemical analysis {Step 5) do not meet specifications, continue HF~Hy treatment (Step 4). Fig. 2. Chemical Flowsheet for Preparation of the Heat transfer line and filter fo 600°C, Transfer salt to storage vessel, Cool all vessels and {ines to room temp. 233U Fuel Concentrate for the MSRE. A ol 13 the furnace temperature controller was adjusted to approximately 50°C above this value. Frequent titrations of the reaction vessel effluent gas stream yielded valuable informaticon cn the progress of the conver- sion of the oxide to the fluoride since the utilization of HF remained guite high and nearly constant until the conversicn was completed. Chemical analyses of filtered samples of the melt were made to determine whether the final process — the purification of the product by hydrogen sparging = should be initiated or whether the hydroflucrin- ation process should be continued. h.2.2 Low-Temperature Process The low-temperature process (see Fig. 3) was used to produce the - fuel concentrate. It differs in only one major respect from that used in the production of the original 235U;bearing enriching salt for the MSRE. In the earlier process, the starting material was UFA; in the low-temperature process, the starting material is UO The oxide is digested at 550°C and then reduced to U62 by treatmeit with hydrogen at temperatures ranging from 400 to 550°C. Helium is used as a diluent gas during the reduction step until the major portion of the excthermic reaction is completed. The oxide is then converted to UFLL by hydro- fluorination at temperatures ranging from 400 to 630°C and at HF (in hydrogen) concentrations varying from 5 to MO%H The final steps of the process involve mixing and fusing the fluoride salts, contactling with HF and hydrogen to remcve residual oxides and corrosion product im- purities, and filtering the molten salt during its transfer to the storage containers. The reduction and conversion processes were monitored by a thermo- couple array that was inserted into the powder in the reaction vessel and by measurements of hydrogen utilization during the reduction step and of HF utilization during the conversion step. Unfiltered and fil- tered samples of the melt were withdrawn for oxide, petrographic, and metal impurity analyses. ......... CHARGE UOss HEAT TREAT U@3§ HYDROGEN REDUCTION: U03 — UO2 HY DROFLUORINATION: U02 — UF‘4 EUTECTIC FORMATION: UF, + LIF —UF - LiF EUTECTIC PURIFICATION: MO + HF —= MF + HyO MF + Hy — MO + HF PRODUCT PURITY: UO3+H 14 CRNL DWG 68~-9515R1 OVERALL REACTION ot 4HF — UF4 + 3H2O‘I UF4 + LiF — UF4- LiF 27% — 73% EUTECTIC COMPOSITION ~13.2 kg U AS UO, 3 TO 5 hr DIGESTION AT 550°C; COOL TO 400°C, START 5% Hp AT 400°C AND INCREASE TO 50% Hg TEMPERATURE RISES TO 490°C; TREAT AT 500-550°C AT 100% USAGE OF Hg; COOL TO 400°C. START 5% HF IN Ho AT 400°C; INCREASE TO 40% HF IN Hp; TEMPERATURE INCREASES TO 450°C; WHEN HF USE DECREASES BELOW 80%, INCREASE THE TEMPERATURE TO 630°C STEPWISE UNTIL HF LUSE BECOMES O; COOL TO 150°C, ADD EXACT QUANTITY OF ?LEF; MELT UNDER 30% H2; DIGEST AT 850°C FOR 3 TO 5 hr; COOL TO 700°C, PURGE MELT 24 TO 30 hr AT 700°C WITH 20% HF IN Hg; TREAT WITH Hy FOR 75 TO 150 hr. UNFILTERED SAMPLE ANALYZED FOR OXIDE CONTENT, FILTERED SAMPLE ANALYZED FOR METALLIC IMPURITIES, Fig., 3. Chemical Flowsheet for the Low-Temperature Process for Preparing the MSRE Fuel Concentrate. The temperatures encountered in this process are generally 200°C lower than those measured in the single-step, high-temperature process. The adoption of this process was felt to be Justifiable without the benefit of a fuli-scale cold run because of the: (1) success of the laboratory-scale experiment, (2) similarity of this process to the original concentrate prod- uction process, (3) improvements made in the vprocess monitoring ilrstruments and technigues, (L) +the time schedule involved. 4.3 Process Equipment The equipment flcwsheet shown in Fig. 4 is a simplified presentation of the msjor compeonents redquired in the process. These components are: (1) the fuel decanning statiomn, (2) the reaction or oxide treatment vessel, (3) the salt storage and transfer vessel, (4) wvariocus containers for shipping the preduct, (5) the off-gas scrubbers. In addition, the process requires many other smaller pleces of equipment, such as: the oxlde can preparation equipment, the in-cell titration assembly, the furnaces for the vessels, the enrichment capsule driliing and weighing station, disconnect stations for electrical and instrument lines and process gas lines, and work tables and tool racks. Figures 5 and 6 show some of this equipment after installation. All services and reagent scurces are located in the penthouse - outside the cell. 16 ORNL DWG &7-il638 R1 1-2 g/min, HF GAASNifiMéP[‘gE e - ‘ == [ Y : ‘ B RESTHOYOR FUEL CAN ENTRY CRUTE 1 Acggaus ADDITlONI chLROD GAS FILTER CELL_"_G_" U036 kgs —iiF4 kgs [~ 1~ p——He PURGES ~1/2 slm CAN vOG AT CUTTING i TOOL CAN CHUCK AND DRIVE CAN CPENING A T DUMPING BOX ° . GELL VENT . % z ‘2ig‘§fm GAS SAMPLE ANALYSIS FLOW + SALT 7 OVERFLOW AR B | ARARRA, ALARM ¢ CALROD | CALROD SALT FILTER ‘ F-2 F-3 — FURNACEZ| | FURNACE AQUEQUS KOH | SCRUBBING SYSTEM ; T-1 T2 T-3 REACTION SALT ENRICHING CAPSULE VESSEL STORAGE AND SALT CAN FILLING VESSEL VESSEL RHD SYSTEM Fig. 4. Simplified Chemical and Engineering Flowsheet for Preparing 233U Fuel Salt, 7 BHORNL PHOTO NO. 90483 A CAN OPENER AND DUMPING BOX SALT TRANSFER LINE™ SALT SAMPLING BOX = REACTION VESSEL FURNACE . Fig. 5. Main Reaction Vessel Furnace and Auxiliary Equipment for 2J3U Fuel Salt Preparation - Cell G, Building 7930. CRUBBER OFF GAS TITRATION o rTiE EQUIPMENT _ REACTION VESSEL CAPSULE AND SALT FURNACE RNACE SALT STORAGE SEL FURNACE 233 Fig. 6. Auxiliary Furnaces and Equipment for Building 7930. U Fuel Salt Preparstion - Cell G, ORNL PHOTO NO. 90464 A QT 19 4.3.1 Decanning Station The decanning station (Fig. 7) is a rather complicated work box that is required for opening the double~container cans, extracting the oxide powder, feeding the powder to the reaction vessel, and disposing of the spent cans. To accomplish this, the station treated each can in the following manner: (1) Received the full can and fed it into a rotating chuck. (2} Tightened the chuck to grip the can while the grocving tool formed a groove deep enough to fix the inner and outer cans together. This was done twice for each can. (3} Pushed the can into the box, displacing the previously opened car. (L) Ejected the previcusly opened can and 1id. (5) OCripped the can, supported it, ancd rotated it while the cutting tocl cut it open. (6) Removed and retained the cut-off 1id while the can was emptied of cxide by vibration and brushing. (7) Inspected the inside of the can and 1id for remaining pcwder. (8) Transferred the powder from the box into the reaction vessel by vibration and use of in-box tools. The decanning station was designed for alpha containment. All penetraticns fto it were sealed by O-rings and boots. kach spent can acted as a sesal until it was displaced by the next can. A ball valve and charging hopper situated on top of the box was used to introduce the LiF to the system. 4.3.2 Reaction Vessel All of the chemical reactions of the process were conducted in the reaction vessel (Fig. 8). The cold-run vessel wag a right cylinder, 7-5/16 in. ID and 36 in. high, fabricated from type 30kL stainless PHOTO 91345 Oc 21 ORNL DWG-67-11674 Material: Vessel & Pipe 3041 SST Liner, Diffuser & Dip Line: Nickel 20! /—-Sheil 8" Sched 80 Pipe 36 -3/4" h——Liner 7 5/16" 1.0.x.36" Lg.x1/8" Thk. Diffuser & Dip Line Fig. 8. Reaction Vessel T-1. 22 steel; it contained a free-standing liner, 1/8 in. thick, constructed from type 201 nickel, and was designed, on the basis of stress rupture data, for a specified service 1life. The vessel contained nozzles to permift powder addition, salt sampling, gaseous effluent discharge, and product transfer. Upon completion of the cold run, the design of the vessel was modified to include a truncated conical bottom te provide better contact of the reagent gases. The production runs were conducted in this vessel. Design information for this vessel and the other process vessels 1s listed in Table 1. %.3.3 Salt Storage and Transfer Vessel The transfer vessel (Fig. S) was a right cylinder, 4 in, in diameter and 36-1/2 in. tall, constructed of type 201 nickel. It contained five nozzles, one c¢f which was a spare. The function of this vessel was to receive a filtered, purified batch of the enriching salt from the re- action vessel and to dispense it to the various product shipping con=- tainers. It was necessafy for this vessel to store salt for 24 hr or more during the changeout of some of the shipping containers. L.3.4 Shipping Containers The shipping contaliners were arranged into three arrays for the Tilling operstions. Later, upon completion of the filling operation and freezing cf the salt, each array was disassembled into individual ccntainers for shipment to the MSRE. The first array (Fig. 10) consisted of 45 enrichment capsules (see Fig. 1i), each of which was 3/L4 in. in diameter and 6 in. long and designed to contain 96 g of uranium. The capsules were connected in series and arranged in three 15-capsule decks. An overflow pot containing ligquid level detection elements and thermocouples was the last vessel in the series. The tubing connections into, and out of, each capsule were precisely positioned so that uniform filling and subsequent blowback of the overfill were possible. Two holes were drilled in each capsule to permit the salt to flow out when the capsule 23 I 1 Table 1. Equipment Design Infofmation: MSRE 233U Fuel Salt Prepaiation Reaction Reaction Salt Furnace Furnace Furnace Vessel Vessel Salt Storage Addition Enrichment Liner Liner Liner Item or Requirement T-1 Outer Inner Liner Vessel T-2 Can® Capsule F-1 F-2 F-3 Type of material 304L S8 201 Ni 201 Ni 30LL SS 200 Ni 300 SS5 300 88 304 SS EIC Maximum operating temp., °C 900 900 700 600 600 900+ 900+ 900+ Maximum 0pera£ing pressure, 56# at 9OO°Cb Free standing 20 20 20 - Open Open Open psig ' 20# at T00°C” 1liner in reaction | vessel; no pres- g sure difference | developed ? Maximum operating vacuum, 15 - 15 15 15 - - - psig Design temperature, °C 900 - - - - - - - Design pressure, psig 6# at 900°C - 20 - - - - - _ (Limiting condition) Calculated maximum stfess, | T27 psi at Maximum stress is 2560 péi at - - - No load carried No load carried psig - 6 psig that due to heed 20# gage by bottom plate by bottom plate : of salt in liner or wall or wall and is negligible Allowable maximum stress 1150 psi at - 3000 psi at - - 870 psi at - - at maximum operating - 900°C 700° CP 900°C temp. Overall height, in. 37 35-7/8 36-1/2 Variable 6-3/8 37 36-7/8 36-7/8 Bottom plate thickness, 1/2 1/8 1/h 5/16 - 1/2 - 1/4 1/8 1/8 in. (reinforced with welded beam on diam) Flat head thickness, in. 3/h - 1/ 1/4 - - - - Outer diameter, in. 8-5/8 T7-9/16 h-1/2 2-1/2 3/h 10-3/k 4-7/8 4-7/8 Wall thickness, in. 1/2 1/8 15/64 0.065 0.035 0.359 1/8 1/8 Inner diameter, in. 7-5/8 7-5/16 b-1/32 2.37 0.680 10.032 4-5/8 4-5/8 Top flange OD, in. - - - - - 16 4-5/8 L-/58 Top flange thickness, in. . - - - - - 3/8 1/4 1/4 Hemispherical bottom wall ‘ thickness, in. - - - - 0.035 - - - Hemispherical top, in. - - - - 3/8 solid | - - - ®Designed by MSRE. blO,OOO hr creep-rupture data; design based on service << 10,000 hr. 24 ORNL DWG-67-11675 T Material: Nozzlgs & Vessel: Nickel Nozzles: 3/8" Sched 40 Pipe 36 172" _— Vessel : 4" Sched 40 Pipe Fig. 9. Transfer Vessel. AN b‘ 25 PHOTO 91466 Fig. 10. Filling Array Consisting of 45 Capsules. 5 5/8" ORNL DWG-67-11676 RI v Material : Nickel 200 3/4 «— 3/16"x0.025" Wail Tubing /8" T ‘/———m 3/4" x 0.035" Wall Tubing 5.375" Fig. 11. Typical Enrichment Capsule. 27 is lowered into the MSRE pump bowl. A cable-and-latch assembly was sttached to each capsule for use iIn the sampler-enriched operation at the reactor. The second array (Fig. 12) consisted of four 2-1/2-in.-diam oy 34-in.~-long cans connected in series for filling operations. Each was designed to contain 7 kg of uranium and had an instrumented overflow pot. The cans were later cdisassembled, and the top and bottom plugs were removed; then the cans were fitted with 1ifting bails and bottom "stopper-type" plugs and weighed. They are being shipped, one at a time, to the MSRE, where the bottom plug of each can is removed before the can is lowered into the drain tank. In this tank, the salt melts and runs out of the container. The third array (Fig. 13) contained a group of six 2-1/2-in.-diam variable-length cans (see Fig. 14) that were similar in design and arrangement to those in the second array. One of these cans was used to store excess product material that was blown back from the other five cans after they had been filled to overflow. The latter cans were designed to contain 0.5 te 3 kg of uranium (2 cans, 0.5 kg each; one can, 1 kg; one can, 2 kg; and one can, 3 kg) . 4.3.5 Scrubbers © A caustic scrubber system was installed to neutralize the HF effluent from the reaction vessel before it was discharged to the TURF HOG system. The scrubbers, namely, four 13-gal polyethylene bottles filled with 10% KCH, were connected in series and fitted with the necessary fill-and-drain connections. The first bottle in the series was kept empty, and the dip lines in the subsequent bottles were posi- ticned at an exact depth. Thus the first bottle had an adegquate volume to retain any liguid that might flow back toward the reaction vessel (in the event that the pressure distribution in the system became reversed). Also, a vacuum bresker was installed to prevent the occur- rence of a vacuum, which could cause collapse of the bottles. 28 PHOTO 91470 Fig. 12. Second Filling Array: Four Cans, Each Containing 7 kg of Uranium. PHOTO 91471 Fig. 13. Third Filling Array Containing Six Shipping Containers of Miscellaneous Sizes. 30 ORNL DOWG-67—11677 i Material: 304L SST Nozzie Salt i Capacity Lo DIMENSION &' Kg ugs ’ m\\fl H 33.592" 7.0 ' [4.802 3.0 | 10.102 2.0 5.402 1.0 3.052 0.5 2 2-1/2" x0 065" Wall o -_fl/////m__-Tumng Uy £ bl = il O 5" 0D, | Orain .‘-‘-NOZZ|G Fig. 14%. Salt Addition Can. 31 Raschig rings had been installed to prevent ccllapse of the bottles in the scrubber system used in the cold run; however, these rings prevented uniform mixing and were subsequently replaced by a vecuum breaker. 4.3.6 Miscellaneous Equipment Oxide Can Lengthening Equipment. — The 23jUO decanning station 3 was designed by using can drawings and sample cans from the Savannah River Laboratory. After the decanner was fabricated, it was discovered that the actual cans were 1-3/& in. shorter than denoted on the drawings. (The measurements of the actual cans had not been checked closely be- cause of the high radiation ievels of the oxide.) Since modification of the decanner was impractical, we decided to lengthen each can 1-1/2 in. by cementing an extension on it. Three items of equipment, as shown in Fig. 15, were required to accomplish this: (1) 2 machining fixture to bevel the end of the can, (2} a press tc force the extensicn onto the can, and (3} a fixture to cure the epoxy resin used in the cemented Jjoint. In-Cell Titration Assembly. — The off-gas titration assembly consisted of a train of three 3-in.-diam by 6-in.-tall glass vessels, a wet tegt meter for measuring gas flows, and the necessary tubing and valves to permit reagents to be added and flushed from outside the cell. Furnaces and Heaters. — Three high-temperature, electrical- resistance-heated furnaces were installed in cell G for temperature control of vessels used in the treatment, storage, and transfer of the salt. A 24-kw, 12-in.-cavity Turnace was used for the reaction vessel; a T-1/2-kw, 4-7/8-in.-cavity unit was used for the transfer vessel; and a 10-kw, Q-?/Snin.—cavity unit was used for the shipping container assemblies. Locally fabricated clamshell heaters using tubular resistance-heating elements provided heat to the nozzles located cn the tops of the vessels. The salt transfer lines were heated by at- taching tubular heating elements to the tubing and then applying thermal insulation. The off-gas line (and filter) was heated with electrical, resistance-heating tape to prevent condensation of HF or water vapor in the line. Can Lengthening Equipment. PHOTO 91472 2t 33 Enrichment Capsule Drilling and Weighing Fixture. — Two holes were drilled in each of 45 enrichment capsules to permit the salt to flow out when the capsule is immersed in the salt in the reactor pump bowl. Upon disassembly of the capsule filling array, each capsule was inserted into the work box (Fig. 16); here, a bottom outlet heole and a top vent nole were drilled in the capsule, the capsule was weighed, the lifting bail was tested for integrity, =snd the capsule was inserted intc the shipping carrousel. The transparent box served as containment during the driiling operation. Off-gas from the capsule drilling box passed through an absolute filter and was then discharged to the plant vessel cff-gas system. Disconnects. — The electrical and thermocouple disconnects used in the cell were standard commercisl units. Generally, one-hzlf of the disconnect was mounted rigidly to facilitate making and bresking cpera- tions with the master-slave manipulators. Several types of pipe and tubing disconnects were used. In genersl, the gas service lines utilized ball-check quick disconnects. Compression fittings were used in the salt transfer lines and In the main off-gas lines. Conventional screwed pipe connections were used in areas where remote operaticon of the Jjoint was not considered necessary. Locally designed and fabricated connections were used in several large-diameter Joints that would possibly reguire remote maintenance. Work Tables and Tool Racks. — The prccess equipment was pecsitioned in the cell to obtain maximum advantage of the capabilities of the eight . available master-slave manipulators. Extensive use was made of support- ing stands and mounting framework. Tocl racks for the required hsnd tcols were leocated at the major work centers. Where possible, the vecid spaces between the process equipment and the cell walls were filled with trays te provide extra work areas, space for temporary tool storage, and a surface on which to catch tools and small equipment items that werer drcpped. A large work table (Fig. 17) was installed to accommodate the product shipping container disassembly, the shipment preparation opera- tions, and the capsule drilling and weighing fixture. Fig. 16. Capsule Drilling and Weighing Station. PHOTO 92787 - Fig. 17. In-Cell Work Table. Y -84919 36 The tables, racks, and supports were fabricated from carbon steel; the trays were fabricated from stainless steel. L. L Operating Procedures Operating procedures were developed for each of the 25 separate operations required for the fuel production. (These procedures were written before the cold run was made.) Brevity and conciseness were stressed. Complete revision of most of them was necessary as & result of the experience gained in the cold run. Review and revision, on a 233 day-to-day rasis, continued throughout the U production runs; the minor equipment failures and malfunctions that occurred during these runs led to extensive procedural changes to ensure continued production of a high-quality fuel concentrate. Tables 2 lists the runsheets that were prepared. 5. COLD RUN WITH DEPLETED URANIUM The run with depleted uranium provided full-scale testing of the chemical process, operational procedures, and of most of the equipment in the absence of a penetrating radiation field. In this run, 14.6 kg of uranium oxide (11.6 kg of 238U) pilus 3.4 kg of LiF were charged to the system to produce 18.7 kg of 1’_,:3'.,16''U"E‘LL (73-27 mole %) eutectic.5 5.1 Preoperational Activities The period from September 15, 1967, through January 15, 1968, was devoted to facility testing; process eguipment fabrication, installation, and testing; operator training; preparation of procedures and manuals; and preparation of the process safety review, the criticality review, and the radiochemical plant safety analysis. 5.2 Mechanical Operations The run with depleted uranium began Januvary 15, 1968, with the a38U0 -filled cans into cell G from the shielded carrier. e loading of 3 37 2 Table 2. List of Runsheets Prepared for 2"‘O)U Fuel Salt Prcduction 1he 15. 16. 17. 18, 25, 26, 27 28. 29, Packaging of LiF. Removing Empty Fuel Carrier from Building 793C Penthouse Pedestal. Preparing and Transferring Empty Fuel Carrier from Building 7930 Penthouse to Building 3C19 Penthouse. Preparing and Transferring of Loaded Fuel Carrier from Bullding 3019 Penthouse to Building 7930 Penthouse. Checkout of Cell G. Preparation of KOH Scrubber Systen. Preparation of Inert and Reagent Gas System. Pressure Test of Helium Header. Pressure Test of H2 Manifold. Pressure Test of HF Manifold (before any run). Pressure Test of T-1 and T-2 Systems. Preparation for Operation: Valve and Rotameter Settings. Discharge oif Fuel Cans from Carrier to Cell G. Capping Fuel Cans. Decanning and Charging Fuel to T-1. Charging LiF to Reacticn Vessel. Transfer of 0dd-Lot LiF tc Cell G. Eutectic Formation. Hydrofluorination Reaction Progress. HF Titration (Inlet). HE Titration (Outlet). oample Procedure. Salt Transiter from Reaction Vessel T-1 to Transfer Vessel T=-2. Salt Transfer from T-2 to T-3, Run No. 1 (3/L-in. Capsule Assembly) . Salt Transfer from T-2 to T-3 (Partial Filling of T-3). Replacing Salt Vessel Assemblies (T-3). Capsule Drilling and Weighing. Weighing and Removal of Salt Shipping Vessels from Cell G. Transfer of Salt Shipping Vessels to MSRE. 38 The UO3 had been processed at the Y-12 Plant, and the empty inner and outer cans had been obtained from the Savannah River Laboratcry. The cans had been filled with the oxide and sealed by a magnetic forming technique to simulate the product cans that would later be used in the production runs. The can opening box did not perform satisfactorily; therefore, it was removed from the cell and extensively modified. The alignment tolerances were relaxed, components were strengthened, addi- tional equipment was installed to aid in transferring the powder from the box, and the viewing capablilities were increased. Difficulty was encountered in transferring the LiF from the de- canning box to the reaction vessel. The avallable vibration was in- adequate, and no provision had been made to sweep or "rod" the powder out and down through the powder addition line. The vibrators had been located on the equipment in such a way that most of their force was imparted to the fixtures and mechanisms in the box. This resulted in some damage to the equipment. Because of this, the vibrators were used sparingly and with cautiocn. 5.3 Chemical Operations Chemical operations began on January 2k, 1968, with the sintering operation, which was conducted at 900°C with a helium flow of 1 std 1iter/min through the bed. The sintering treatment was immediately followed by the hydrogen reduction step at the same temperature. A 2-liter/min ficw of gas, consisting initially of 0.1 liter of hydrogen and 1.9 liters of helium per minute, was attsinea initially. The com- pcsition of this gas was adjusted periodically until a flow of pure hydrogen was obtained. Then the flow rate was increased, in increments, to 10 1iters/min, where it was maintained for 5 hr. During this treat- ment, the sintered-metal off-gas filter became plugged. ©Subsequent examination of the filter unit revealed the presence of a hard filter cake. Water wvapor, which is evolved during the reduction proccess, coculd have condensed on the filter in the event that the filter heater was 1mproperly operated; this would have permitted the filter temperature to drop below the dew point of the effluent. 39 The conversicn of the oxide to the fluoride began January 27, 1968, Anhydrous hydrogen fluoride mixed with hydrogen (10 meq of HF per liter of hydrogen) was injected through a dip line to the bottom of the UO2 bed at the rate of 2.2 liters/min. This reaction began at 900°C, and the temperature was pericdically decreasged to maintain the melt at about 50°C above the liquidus temperature. The in-cell titration appa- ratus was used to monitor the HF in the effluent, but the results were inconclusive; thus another titration unit that permitted precise addi- tion of the equilibrating reagents from outside the cell was designed and installed. The data obtained with the new equipment showed that the utilization of HF was essentially zerc (which indicested that the conversion of the oxide was complete) after 295 hr of hydrofiuorination. An attempt to sample the melt was unsuccessful because of the presence of a hard top layer (or crust). No noise was evident from a listening rod installed to detect bubbling from the dip lire. It was concluded that the dip line had broken, by corrosicn or melting, above the liquid interface. This would account for the lack of HF utilization and the slow progress In the conversion of the oxide. Subsequent examination of the dip tube showed a large hole approximately at the interface. A new dip tube, zlong with a sensitive sound detector, was installed, and the HF treatment was resumed. Bubbling in the melt was easily detected; and, for the first time, reascnable values for the HF utilization were obtained. After 250 hr of additional HF treatment, the dip line be- came plugged at a location about 1m3/h in. from its open end. Upcn exemination, the plug material appeared to be uranium oxide. Four sub- sequent lines plugged approximately at the gas-liguid interface within the tube; in these instances, the plug materizl also appeared to be uranium coxide. It was obvicus, then, that excessive moisture was being introduced into the melt by the gases used in the process. This problem was sclved by installing a column filled with magnesium perchlorate in the helium system and by installing a catalytic recombiner; a molecular- sieve drying column (with regenerative capabilities), and a moisture analyzer in the helium system. e The effluent gases from the reaction vessel were directed to a caustic scrubbing system, which consisted of polyethylene bottles filled with 20% KOH solution and Raschig rings (to prevent collapse). At an early stage in the process, the dip lines in the scrubber botties plugged, and a white precipitate appeared. Frequent sparging to promote mixing, frequent recharging cof the caustic solutiocn, rinsing with hot water, and a decrease (from 20% to 10%) in KOH concentration permitted use of the system for the remainder of the cold run. It was felt that the Raschig rings were a major contributor to this problem in that they promoted stratification of the solution; this was confirmed by the colored interface that appeared in the bottles when phenolphthalein indicator was present. At the conclusion of the cold run, the scrubber system was completely replaced by one that contained spare lines and a vacuu breaker (to prevent possible bottle collapse), and permitted remote replacement and flushing of all lines. Although it gave satisfactory results in the small-scale development runs, the technique of measuring the liguidus temperature of the melt and subsequently determining, from the binary UFu—LiF phase diagram, the progress of the conversion reaction did not prove successful in this system. The heat capacities and thermal insulation of the systemn, along with poorly located thermccouples, obscured the inflection in the temperature decay curve of the melt upon cocldown. However, the HF utilization data, alcng with material balances for the system, provided sufficient reaction progress information. When the reactlon was Jjudged to be complete, an unfiltered sample of the melt was withdrawn and ana- lyzed, chemically and petrographically, for cxide content. The purification cf the melt, by the reduction of the metal fluo- rides to metal, was conducted without incident. Titration of the HF in the effluent gas proved to be a reliable means of determining the end | point of this reaction. Filtered and unfiltered samples were taken at the conclusion of the purification step. The Tiltered samples were considered to be representative of the packaged product since the melt was filtered in the transfer operation. The cold run product analyzed as follows: b1 Concentration (wt %) U O Ni Fe Cr AT S 61.0 0.02 0.C065 C.007 C.26 0.1 0.002 The product wag filtered into the intermediate storage vessel; then it was transferred to the capsules and shipping containers by subjecting the salt to = helium overpressure of approximately 12 to 15 psig and forcing it out of the vessel through & dip line. This in- volved several operations because the capsules were, first, filled and, then, replaced in the system by the shipping can container assembly for filling. The compression-type tubing fittings used for this operation were very difficult to open after thermal cycling and after salt had contacted them. (Also} it was difficult to make up these fittings properly and to obtaein leak-tight joints, using the master-slave manip- uwlators.) The approach toc a solution of these difficulties consisted of: (1) modification of the intermediste transfer vessel by additicn of spare nozzles, (2) reorientation of the vessel to improve accessibility, (3) use of redundant fittings when necessary, (H) use of a high-temperature thread lubricant on all fittings, (5) modification of the operating procedure to reduce the number of times fittings had to be opened, and (6) training operators in use of the manipulators on a simulated piping arrangement. In summary, the cold run with depleted uranium was invaluable in that it: (1) 1led to a process that was more simple to operate and that was compatible with the equipment, Lo (2) pointed out design and fabrication deficiencies in the equipment, (3) provided on-the-job training for the operators and supporting personnel, (4) tested the operational procedures, showing the need of exten- sive changes in them, and (5) showed that comprehensive remote maintenance procedures would probably be mandatory to ensure successful completion of the 233 U fuel concentrate production runs. 6. (1a7-%3UF, PROCESS OPERATIONS Three production batches, using the low-temperature flowsheet, were required in order to prepare 63.4 kg of the fuel-enriching concen- trate, TLiF—EngFh. This concentrate contained 3¢ kg of uranium (35.6 kg cf 233U), The first run began May 9, 1968, and the third run was completed July 30, 1968. The 45 enrichment capsules were filled with a portion of the first-run product; the four T7-kg shipping containers were Tilled with the remainder of the first-run product, all cf the second-run product, and a portion of the third-run product; the mis- cellaneous shipping container assembly was filled with the remainder of the third-run product. The ten salt shipping contsiners and the 45 enrichment capsules were delivered to the Molten Salt Reactor as required in the reactor enrichment schedule. 6.1 TFeed Materials The 233UO3 had been prepared in batches by using 7 M NHAOH (in excess) to precipitate hydrous uranium oxide from solutions that con- and NH, NC,. 3 43 The uranium in the feed solution had been purified and isolated in 1964 tained 10 to 40 g of uranium per liter and were 1 M in HNO and 1965 by solvent extraction followed by ion exchange. These treat- ments decreased the concentrations of plutonium, thorium, fission prod- ucts, and corrosion products (iron, nickel, and chromium) to acceptably low levels. Table 3 shows typical analyses of the oxide feed material. Table 3. Compositionsa cf Teeds for Preparing MSRE Fuel Selt Run No. RU-33-1 RU=-33-2 RU-33-3 Oxide, g 15,929.4 15,901.7 16,376.7 Total U, g 13,082.4 13,105.8 13,533.0 233y, o 11,966.9 11, 969.7 12,377.8 Impurities, ppm Chromium 50 &7 Ll Tron 372 588 271 Nickel < 16 L7 < 13 232 £D5 218 ool Isctopic Analysis of Uranium, at. % 233 A 33y 01.L47 91.35 91.46 ) 23 'y 747 7.4 7.46 232y 0.809 0.948 0.816 236U 0.058 0.0701 0.059 238U 0.191 0.245 0.202 Data furnished by manufacturer. Ly The LiF feed (99.97% TLi) had previously been densified +to 1.2 g/cm3 by hydrogen treatment st elevated temperatures. 6.2 Radiation Levels of the Oxide Feed The high radiation levels of the oxide feed material (see Table &) result from the daughter activity of the 232y (222 ppm) in the 233y, Alpha, beta, and gemma radiation is emitted in the transitions (see Fig. 18). From the shielding standpoint, the 2.6-Mev gamma radiation from the QOBTl and the 2.2-Mev gamma radiation from the 21251 are the most important. Table 4. Radiation Levels™ of Storage Cans Containing Oxide Feed Material® Time After Rediation Level {(r/hr) Purification Distance from Source (months) 0 1 ft T m 10 ft 20 ft o 1 o6 10 1.1 0.120 0.028 36 250 25 0.120 Lo 300 25 0.200 "Measured with 2 paper shell cutle ple. Prach can contained 450 g of 233U. 6.3 Oxide Feed Material Handling The 233U-’O3 had been packaged in double-walled alurinum containers, containing approximately 450 g of uranium. The cans, 2.8 in. 0D by 8 in. long, had been sesled on one end by welding, and on the other end by a magnetic forming technique. They had been stored in Building 3019 at ORNL. 233 To prepare the U fuel concentrate, we removed 89 cans from the facility and transported them, in six loads, to the TURF in a shielded carrier. The carrier was positiconed on a pedestal atop celli G for unloading the cans intoe the cell. * 232% (n,2n) 233 * Y232 |y RECYCLE FUEL Ui U ~80% FROM ngzz B13d ~ 0% FROM U a T4y pg®>2 Gamma : Energy (n, ¥ Nuclide ' (Mev) |Yield ’ ! Pb2!2 | 0.250 |0.40 § pg23! TH228 0720 0.19 § 0830 0.19 § B25h .03 006 | 24 Bi2!2 | {34 005 | Th™ 6.i3h al9y 181 007 22 00358 v 0510 025 § 232 228 224 : Th Ac Ra 1208 | 0582 080 | 0859 | 0.15 § 262 | 1.00 § a 1.39x10'§ BTy a 3.64d v ¥ RQZZB RnZZO Po2ié B 64 % @ 54.5s 3.04x1077s 4 ¥ p02|6 BEZIZ prOB 606 m (stable) a b 36% a 0.158s " B3.Im h Y ¥ pb2!2 TIZOB Fig. 18. Decay Chain and Gamma Activity in %06-233[}02 Fuels. Le No spread of contamination or excessive radiation exposure to perscnnel occurred during the removal of the oxide cans from the storage facilit during their transfer to TURF, or during their discharge from s g 3 g the carrier to cell G. In cell G, the cans were gaged, and any burrs or excess weld metal were removed by filing. Three separate operations were required to elongate the cans to the 9—1/2—in. length required for proper operation of the can opening box. First, the cans were machined on one end tc "true” them. A premachined end cap (with epoxy cement applied) was inserted into the cell through the small-items entry port. This cap was installed on the can in a pressing fixture, and then the can (with cap) was placed in a curing fixture, where five cans at a time were cured under longitudinal pressure for 1 hr at a temperature of 110 to 120°C. The can preparation equipment functioned satisfactorily, except on one occasion in which the drive motor on the can machining rixture burned out. The remaining cans for that run were dressed with a file, and the caps were installed without Incident. ILater, a new motor was installed remotely in the fixture. The elongated cans of oxide were opened, one at a time, in the decanning staticn. This was a time-consuming operation involving the use of equipment that had been recognized during the cold run to be marginal. The decanner had a rugged chuck assembly that held and ro- tated the cans while a cutting wheel opened tThem inside an alpha-sealed box. Most of the cother attachments in the box were of little value, and in some instances, actually hindered the operation. The major problems in the decanning operation can be sumarized as follows: (l) Alignment and clearance tolerances were tco close for hot-cell operation. (2) Visibility in the decanning box was limited. (3) The powder failed tc flow ocut of the box easily. W7 (k) A vibrator shaft seal failed, leading to loss of oxide powder from the box. This caused rather extensive contaminaticn of the cell. Modifications were made remotely during the fuel concentrate production to increase visibility, repair the seal, and to provide tools inside the decanner to assist in trans- ferring the powder to the reaction vessel. 6.4 Reduction of Uranium Oxide !‘) The expanded bed of ;33UO3 was heated for 2 hr at 550°C in a helium atmosphere to remove, by pyrolysis, any traces of ammonium com- pounds or cother volatiles remaining from the chemical processing. The bed was then cooled to U00°C before hydrogen treatment was started. This temperature was sufficiently low to accommodate the temperature rise expected from the exothermic reaction of hydrogen with U03: U0, + By = U0, + H,0 + 72,000 cal/g-mole. The concentration of hydrogen (in helium) was adjusted initially tc 5 vol % and was gradually increased to 50 vol % during the first 4 hr of treatment. The temperatures within the bed were monitored by an internal probe with 12 thermocouples that were placed at 2-in. intervals along the vertical axis of the bed. These temperatures rose in response to an increase in hydrogen concentration and then became constant after the initial excursion. The procedure of incrementally increasing the hydrogen concentration was repeated until the hydrogen:helium volume ratio was 1:1; the gaseous flow rate was 2 liters/min. The location of the reaction zone and the zone movement 1nside the bed were clearly defined by the temperature profile. As the reaction progressed, the reaction zone rose, in the form of a band, up through the powder bed. Figures 19 and 20 show plots of the temperature at 6 in. and 16 in., respectively, from the bottom of the 2Lk-in.-high bed during the uranium recuction phase cf the three production runs. The rise in the 880 560 S 540 g B 520 - e & & 500 & = [P = 480 Bad = 2 460 ® 5 Z 440 | 420 400 380 Fig. 19. ORNL DWG 68-9477 "RUYK no. | He = 0.4 sim purge RUN po. 2 ———— ) Ha = 1.0 sim RUN no. 3 mn—rm- S e He = 1.0 sim \ \! \ \l \ \‘ \ \ \ \ 4 \ \ \ % Vo \ \ \ \ \ \ I e e e 4 8 i2 6 20 24 28 32 36 40 44 48 52 56 60 TIME (hrs) Temperature 6 in. from Bottom of Oxide Powder Bed During Uranium Reduction. 8N 580 560 ~ 540 [ 8] L & 520 > }— = £ 500 O. F i = 480 [FY] = 9 460 x [T1] Q. 440 Q. 2 420 400 380 Fig. 20. ORNL DWG 68-3476 He = 0.4 sim purge Hz"‘ 3.0 SIm He = 1.0 sim 8 12 ie 20 24 28 22 36 TIME (hrs} 40 44 48 52 56 60 Temperature 16 in. from Bottom of Oxide Powder Bed During Uranium Reduction. ot 50 temperature of the lower zone during the first 8 hr of treatment is due {to the exothermic reaction (the reaction vessel furnace was set to con- trol at 400°C). After the temperature excursion resulting from the initial increases in hydrogen flow had subsided (approximately 12 hr), the temperature of the furnace was increased, at the rate of 30°C/hr, until the bed temperature was 525 £ 25°C. The reduction operation was continued at this temperature, with 50 vol % hydrogen=--50 vol % helium, until 50 to 100% excess over the stoichiometric amount of hydrogen had been passed through the bed. Hydrogen utilizetion within the bed was 100% until the reaction zone approached the top of the powder bed; then a slight decrease was observed. Hydrcgen usage was determined by a material balance of the gas Tlowing into the reaction vessel, as measured by rotameters, and the gas outflow, as measured by the in-cell wet test meter. 6.5 Hydroflucrination of o, Upon completion of the oxide reducticn step, the bed was allowed to cool to LO0°C. The coaversion of Uo, to UFH by hydrcfluorination, using HF gas diluted with hydrogen, began at LOO°C and was completed at 625°C. TFive to seven days were required for the conversion. The EF gas wes supplied to the process by withdrawing it from the vapor space of a heated 100-1b HF cylinder. The cylinder heater was thermostatically controlled to provide a vapor pressure to 12 to 14 psig. A differential-pressure Transmitter across a capillary restrictor in the HF gas supply line was used to monitor the flow. The gas was passed through =& maze of tightly packed nickel wire in a 2-in.-diam nickel tube that was maintained at 625°C to remove sulfur from the stream. It was then mixed with 2 metered amount of dry hydrogen, filtered, and intro- duced to the reaction vessel through a dip tube that extended to the bottom of the U02 bed. At the beginning of the hydrofluorination step, the composition of the gas was 95 vol % hydrogen--5 vol % HF; the flow rate of the mixture was 2 std liters/min. Over a period of 3 to 4 hr, the HF concentration 33 was incrementally increased to 40 vol % as the exothermic reaction caused the bed temperature to rise from 400 to 450°C. During these initial hours, the temperature within the bed was constantly monitored to determine when the temperature excursion resulting from each HF flow adjustment had ceased and when another adjustment could be made. ATter the HF concentration of the hydrofluorinsting mixture had reached 40 vol %, the reasction zone traveled, in the Torm of a band, up through the bed (in a manner similar to that observed during the hydrogen reduction) as the U0, was converted to UFH' The reaction uo, -+ LHF — UF) + 2E,0 + 144,000 cal/g-mole is more exothermlic than the reducticn reactlon, but it does not have as great a tendency to cause thermal excursions. Probably, this is the result of differences between UO UOE’ and UFLL with respect to bed 3) permeabllity and thermal properties. The reaction-zone temperatures for the three production runs are plotted as a function of time in Fig. 21. In runs 1 and 2, several furnace temperature adjustments were made during the first five days of treatment in an eiffort tc increase the bed temperature to L75-500°C (vecause the HF gas is less corrosive to nickel at the higher temperature). In run 3, the furnace operatec satisfactorily and did not reguire adjustment. The vrogress of the reaction was followed by observing the reaction zone travel through the bed and also by cobtaining a material balsnce of the HF in the system. The HF utilization was essentially 100% for the first five days and then decreased sharply as the reactlion zone reached the top of the bed. Then the temperature of the bed was increased to 625°C, where it was held for two days to ensure completeness of the reaction. The HF utilization did not increase at the higher temperature; instead, it continued to decrease, suggesting that the rezction had been complete at the end of the fifth day of hydrofluorination. A total of 13.5 kg of wranium, as U0, was converted to UFM in o ‘ each of the three runs; only very minor differences in the runs were noted. 5e ORNL DWG 68-9478 680 660 |- 640 . He = 0.4 sim purge g20 | He = 1.2 sim -] \}{ |/ B ! | | ! | | | L HF 600 - 40% (C.8sim} in Hy 580 - 560 — 540 — 520 500 — 480 |- 460 — REACTION ZONE TEMPERATURE (°C) 440 42Q 400 - 380 | | 1 ! 1 | | 0 24 48 72 96 120 144 ieg 192 TIME (hrs) Fig. 21. Maximum Temperature Within Powder Bed During Hydrofluorination of UOQ.. c 53 £.6 TFormation of the Fubectic Salt The eutectic mixture UF)-LiF (27-73 mole %) (Fig. 22) was formed by adding the stoichoimetric quantity of 1ithium fluoride powder to the uranium tetrafluoride powder and fusing the mixture. Dry LiF powder was added to the reaction vessel by the fellowing series of operations: (1) it was poured into a hopper located on top of the decanning station, (2) it was dumped from the hopper intec the decanner box, and (3) it wae transferred from the box to the vessel by vibration and brushing. The temperature of the reaction vessel containing the stratified UFLL and LiF powders was then increased to 855°C in order to melt the 1ithiun fluoride (mp, 835°C). The melt was digested at this tempera- ture for 3 hr while it was sparged with hydrogen (at a flow rate of 0.2 1iter/min) to reduce any extraneous compounds that might have been introduced during the LiF addition. A sound detector was atliached to the reaction vessel dip line to sense the bubbling of the hydrogen sparge that would occur in the presence of a liguid in the bottom of the reaction vessel. This was an excellent device for determining when the meltdown had started. The internal thermccouple probe confirmed that, at meltdown, a very wide range ol temperatures existed along the axis of the melt. A rapid cooldown to 65C°C was observed in the lower region of the vessel where ligquid existed. At the same time, temperatures as high as 850°C were measured in the upper regions of the vessel where fusing of the com- pounds had not yet taken place. As more liguid was formed, a gradusl shift in the temperature profile occurred, indicating a rising liquid level in the vessel. The temperature of the pool eventually reached 850°C, the set point of the furnace temperature controller. The initial presence of the ligquid and the rising level were also indicated by changes in the dip line pressure as The back pressure on this line increased to overcome the rising liquid level. Differences, with regard to conditions during initial meltdown, were noted in the runs. In runs 1 and 3, the UFA and LiF had to be 5k ORNL-LR-DWG 4745TA $40C // 900 P 7 o c. & 80O \ > \ / < o ! / & 700 E ! ; - z t 600 N <& W < 500 S W v W 4LiF -UF, | - - 400 | f LiF i0 20 30 40 50 60 70 80 90 UF, UF, (mole %) Fig. 22. LiF-Tfl‘)_P Phase Diagram. 52 heated to about 850°C before melting occurred. In run 2, initial meliting occurred at 650°C, nearly 200°C below the melting point of the lithium fluoride. The low-temperature initial melting must have resulted from the presence of a sizable heel of salt (mp} hg0°C) that remained in the reaction vessel from run 1. Probably, this heel acted as =a "seed" to permit fusing of the UFlL and LiF powders at the lower tempera- ture. The G=-in.-deep pool of eutectic salt (mp, LO0°C) was next treated with 20 vol % HF--80 vol % hydrogen (flow rate, 2.4 liters/min) for 24 hr at a temperature of 700°C to remove the last traces of oxide from the salt prior to the hydrogen purification procedure. At the conciu- sion of this ftreatment, an unfiltered sample of the salt was withdrawn =nd analyzed for oxide content. (A 1/2-in.-0D x 2-1/L-in.-long nickel cup was immersed in the salt to withdraw a 25-g sample.) After the sample had solidified, a l-in.-long section was cut from the center and analyzed petrographically and chemically for the presence of er' The remaining portion of the sample was submitted for = complete chemical analysis. Hydroger. treatment of the salt was started prior to obtaining the results of the analyses. The more rapldly obtained petrographic results were used to determine whether hydroflucrination should be resumed or whether hydrogen purification of the melt should be continued. In each of the three runs, the UC. content was reported to be less than the 2 lower limit of accuracy (200 ppm) for the petrographic appraisal; thus, subsequent HF treatment was unnecessary. Chemical analyses of the same samples showed oxide contents (in the product sslts) of 62, 34, and 32 ppm for runs 1, 2, and 3 respectively. £.7 Purification of the Eutectic Salt The eutectic salt was purified by bubbling pure hydrogen gas (3 to 10 ppm HOO), at & flow rate of 2 std liters/min, through the 10=in.-deep meit, The temperature of the molten salt during this reaction, G MF + 1/2 H, — HF + M, 56 is 700°C. The progress of the reaction was followed by titrating the effluent gas with the in-cell titration assembly. These data are plotted in Fig. 23. The high rates observed initially can be attributed to the evoiution of the soluble HF in the melt. The first inflection and plateau of each curve correspond to the reduction of the nickel fluoride; this is followed by a second inflection and plateau, which represent the reduction of the iron fluoride. The end point of the purification was evident from the leveling off of the HF evolution rate at 0.025 meq per liter of hydrogen. The hydrogen flow rate was in- creased on several occasions during the process, and the reaction rate seemed to be almost independent of the hydrogen concentration. The effect ¢f the hydrogen treatment on the concentration of the corrosion products is shown in Table 5. The chromium concentration in the melt was not affected by the hydrogen treatment. Higher initisl values of nickel in the third run, and of iron in the second and third runs, were the result of residual material left in the bottom of the reaction vessel from previous runs. Apparently, the metals had pre- cipitated and thus had not been filtered cut during the transfer operation. The levels to wnich the iron and nickel concentrations were lowered by the hydrogen reduction of the metal fluorides are believed to be neay the 1imit of accuracy of the sampling system and the laboratory analyses. The actual concentrations of these Impurities in the en- riching concentrate, as delivered to the MSRE, should be somewhat lower than the values reported because an additional 24 hr of hydrogen treat- ment was provided in each run after the filtered samples were withdrawn for analysis. Table 6 gives the chemical analyses of the fuel shipments to the MSRE; Table T gives the isotepic analysis of the uranium. During each of the three runs, the reactiocn vessel was exposed for 2C days to 40 vol % HF--60 vol % hydrogen at temperatures ranging from 400 to 850°C. Approximately 5 g of nickel (from the nickel liner of the reaction vessel) was lost to each melt. This corresponds to = uniform corrcosion rate of slightly less than 0.001 in./year -~ g 1low rate for this type of process. HF OUTLET (meq HF/I Hp) 0.20 018 o.oe 0.06 0.04 .02 0.00 ORNL DWG 68-948 -. Ho = 2.0 sim \ \ He = 0.4 sim \; | ‘\ TEMPERATURE = 700°C A .‘ o RUN no. | ‘ \\" RUN no, 2 —— — — L ] \ N RUN no. 3 ——-—-— 24 48 Fig. 23. 72 96 120 144 168 192 TIME (hrs} Purification of the Eutectic Mixture. 2is LS Table 5. Concentrations of Corrosion Products in the Eutectic Salt Nickel (%) ‘Type of Run Sample 1 2 3 Unfiltered: Before H? Treatment 0.0191 0.0038 0.430 Filtered: After H_ Treatment 0.0066 0.0050 0.0160 2 0.0295 0.0750 Chromium (%) Run 1 2 3 M o 3 Table 6. Analyses™ of the 2 jUF)_L-?LiF (27-63 mole %) Fuel Concentrate Cans B, C, D, and B Capsules (45) Cans ¥, H, J, L, N, P Chemical Spectrographic Chemical Spectrographic Chemical Spectrographic Element Analysis Analysis Analysis Analysis Analysis Annlysis Al B < .01 < 0,01 ca < 0.004 < .00k Ce < 0.033 < 0.021 < 0.038 Cr 0.0057 < 0.002 $.0055 < 0.002 0.0045 < 0.01 Cu Fe 0.0104 0.047 0.0051 < 0.004 0.01k45 < 0.09 Gd < 0.003 < 0,002 < 0,00k Li 4.83 4.91 L.8 Mn < 0.0004 < 0.C00k Mo < 0.004 < 0.C04 Ni 0.0082 < 0.04 0.,0066 < 0.002 0.01E0 < 0.09 0, b ppm 62 ppm 32 ppm Pb Pr < 0.030 < £.010 < ¢.038 8i s < 20 ppm < 20 ppm < 20 ppm Sm < 0.007 < 0,004 < 0.008 Sn Ti < 0.002 < 0.002 . i Zn < 0.002 < 0.002 U 61.38 61..8 61.19 a . . In wt % unless given in ppm. 65 60 Table 7. Isotopic Analysis of Uranium in MSRE Fuel Concentrate Capsules (U45) Cans Cans B, C, D, E F, H, J, L, N, P Isctope 233y 91.465 01.378 91.466 23%; 7.185 7.463 7.466 235y 0.803 0.87h 0.810 236 0.058 0.065 0.0585 23'8U 0.186 0.218 0.196 Total uranium, kg 4.300 28,059 6.583 Average at. wt 233.100 233.150 £33.090 Total uranium, wt % 61.800 61.380 61.190 6.8 Transfer of the Eutectic Salt Eight transfer operations were necessary to convey the three 13.5=kg batches of eutectic =ssl1t mixture from the reaction vessel to the intermediate transfer vessel and then to the variocus shipping con=- tainer assemblies. The transfers were made by applying helium gas pressure cover a pool of molten salt having a temperature of 600°C. The salt was forced out of the dip line and thrcugh the transfer line and salt fillter tc the receiving vessel, which was vented. The transfer lines were heated with tubular resistance heating elements, insulated toc conserve heat and avoid cold spots, and equipped with thermocouples. The transfers ranged in size from the 13.4 kg of uranium (4.7 liters of salt) in the production batches to 4.3 kg of uranium (1.5 liters of salt) that was reguired to fill the 45 enrichment capsule array. The primary transfer line (3/8-in.-0D x ©.065-in. wall type 200 nickel tubing), which connected the reaction vessel to the lk-in.-diam storage vessel, alsoc served as the treatment gas supply line to the dip 61 tube in the reaction vessel. During the processing cperations, this line was ccnnected to the ftreatment gas supply line instead of the transfer vessel. In each of the feed batch transfers; a compression- type Titting had to be broken and the transTer line had tc be connected to & new salt filter that had previously been installed in a nozzle atop the storage vessel. The sintered nickel filfer was 3 in. in diameter, and the pore size was 20 p. Transfers through tnhis line re- gquired an overpressure of 15 psig to overcome the static head (18 £t HEO) and the restriction to flow of the filter. Time required to com- plete the transfer varied from 78 to 140 min. Initial movement of the salt through the line was obvious from the rapid response of the attached thermocouples to the salt tempersture. The progress of the transfer was followed by observing the rising back pressure on the helium purge to the storage vessel dip line; this indicated a rising liquid level. The rush of gas through the line at the end of the transfer was audible from the sound detector. These conditions were verified by a sharp drop in the reaction vessel pressure and a rapid increase in gas flow through the salt transfer line. Five transfers were required to fill the three-product container assemblies. They were made via a 1/4-in.-OD nickel transfer line that connected the assembly to a dip line in the storage vessel. Since filters were not involved in these transfers, the transfers were com- pleted in approximately 10 min. Thermccouples cn the transfer line and on the shipping containers indicated the movement of salt. An overflow pot at the end of the containers (in series) was egquipped with liquid level detection probes and internal thermocouples to indicate the preg- ence of sglt. A blowback of gas then ensured that the contasiners were filled to the proper level. The transfer operations were conducted at a salt temperature of 600°C, (This temperature had been arbitrarily selected, and the con- : tainers had been fabricated to contain the desired quantity of fuel at this temperature.) The filled containers were allowed to cocl, and the salt was allowed to solidify before the disassembly operaticn began. 62 6.9 Container Disassembly and Preparation for Shipment The shipping container assemblies were stripped of the thermo- couples, overflow pots, heaters, and supporting hardware. The containers were separated from each other by cutting the interconnecting tubing. The 45 enriching capsules were clipped and trimmed, were tested (i.e., their lifting bails were tested) with & 15-1b pull, were drilled for draining and venting, and were ldentified and weighed. Then they were packaged, in groups of six, in carrousel shipping containers. The equipment, which was designed to do these operations remotely, functioned satisfactorily. The 2-1/2-in.—diam product cans were easlily removed from the filling arrays; nhowever, removal of the top and bottom plugs from each can was guite difficult. An inordinate amount of force was necessary to break the bottom plug locse from the thin salt Tilm that had formed in the annulus between the plug and the bottom nozzle. Several fixtures de- signed for this task proved to be inadequate and had to be replaced with sturdier units. A lifting bail was installed on each can. Also, a stopper was inserted in the bottom of the can to minimize the loss of salt during handling and shipment. The cans were welighed by using a beam balance that was located in a glove box mounted cn top of the processing cell. The glove box be- came a part of the cell contaimment for this operation. A long cable, with the 1ifting bail of the can attached, was suspended from the balance into the cell. The gross weight of the cans ranged from 1.6 to 15 kg. All of the shipping containers were then stored in cell G to await shipment tc the MSRE. 6.10 Transfer of TLiF-233UFLL Salt Product to the MSRE Six of the 2-1/2-in.-diam product salt containers (total uranium content, 33 kg) were delivered, individually, to the MSRE in the 10-ton Pu-Al carrier. The Pu-Al carrier was slightly modified for this opera- tion. The three-barrel magazine was removed, and a winch-and-cable 63 assembly was installed on top of the carrier in a sealed housing &o that it became a part of the carrier containment. The carrier was placed on top of the pedestal over a 6-in.-diam port to cell G, and beeame.a part of the cell containment when the slide drawers in both the carrier and the pedestal were opened. The cable hook was lowered into the cell to receive the product can and then raised to withdraw the can from the cell into the carrier. The pedestal drawer and the carrier drawver were closed, isolating the car- rier and cell from each other and from the ambiIent atmosphere. In this manner, containment of the fuel was maintained at all times. The car- rier was then remcved from the pedestal, sealed by the installation of the cover plates, and cleaned of any surface contamination. The fuel carrier was then transported tc the MSRE, where it was positioned over a turntable (Fig. 24) that was designed for charging the enriching concentrate to the reactor fuel draln tank. Measurements, at the surface of the Pu-Al carrier, of the radiaticn emitted by one T-kXg uvranium fuel can are given in Table 8. The neutrons are proauced by (a,n) reactions on lithium and flucrine (of the eutectic) by the high~energy alpha particles emitted by several of the daughters of the =37, The enrichment capsules were shipped tc the MSRE in a smaller carrier that was designed to accommodate six capsules in a carrousel fixture. The carrcusel served as a support Tixture during trsnsport and as a holder for six bottom plug assemblies that became a part of the sampler ~ enricher system at the MSRE as the capsules were removed Trom the carrier through a l-l/2=in.-diam hole in the top. A small glove box was fabricated for use in removing the carrousel assemblies from cell G. This box was mounted on top of the small car- rier, which, in turn, was mounted on the pedestal above the cell. Then the pedestal slide-valve drawer was opened, and the glove box and the carrier became a part of the cell containment. A cable from a winch inside the box was lowered into the cell to receive the carrousel. The carrousel was withdrawn into the carrier, the pedestal drawer was closed, Tank. 6l ORNL-DWG 68-967 TURF CARRIER -GRAPHITE SAMPLING s SHIELD PURGE GAS SUPPLY— e e e o . P MAINTENANCE SHIELDz S e T ~ CONTAINMENT ENCLOSURE / AND STANDPIPE ASSEMBLY . ~ Fig. 24. Arrangement for Adding 233U Enriching Salt to Fuel Drain 65 Table 8. Radiation lLevels, Measured at Surface of Pu-Al Carrier, of TLiF—233UFlL F'uel Containing 7 kg of Uranium Instrument Loecation Along Radiation Level (mrem/hr) at Specified Vertical Axis Type of Distances from Cerrier Surface of Carrier Radiation 0 12 in. 5 £t Gamma, <2 < 1 < 1 Ny 5 50 12.5 12 in. below T top of carrier I\T C:3 0.3 0.2 T I N, o+ 25 15 5 Gamna <7 NF 200 75 10 Center of carrier NT C.6 0.3 0.3 - _ N ¢ 75 Lo 7.5 Gamma. < 1 <1 < 1 L \ 18 in. =bove NF 2 o0 = bottom of N, 0.3 C.6 0.2 carrier ~ = v Nint 25 17.5 6.5 aN?, NT, Nint = Neutron radiation (fast, thermal, and intermediate). and then the carrier was removed from the pedestal and placed on 1its base. The glove box was removed from the carrier, and shield plugs were inserted in the l/Emin.mdiam and l—l/z-in.—diam access openings in the top of the carrier. The carrier contained a S5-in.-diam cavity with 4 in. of lead shielding around it. When the small carrier housed six capsules, each containing % g of uranium in approximately 150 g of fuel concentrate, the gamms radi- ation level was 200 mrem/hr at the surface and decreased to 20 mrem/hr at 6 f£. The neutron radiation levels at the surface and at 6 ft were 105 mrem/hr and 10 mrem/hr, respectively; measurement made at the open 66 bottom of the carrier during removal from the pedestal indicated s radiation level of 50,000 mrem/hr. The removal of the fuel from cell G and its subseguent transfer to the MSRE were characterized by very low radiation exposures to personnel and essentially no spread of contamination. In fact, only the top surface of the pedestal and a small area immediately adjacent to it had to be decontaminated. £.11 Meterisl Balance A total of 39.72 kg of uranium, in the form of uranium oxide powder, was transferred to cell G of the TURF. Of this amcunt, 38.95 kg of uranium, as LiF-UFh, was packaged as the finished product. The can-opening operation accounted for a loss of 245 g of uranium as uranium oxlde. DMost of this, 200 g, was lost through the empty can discharge chute because excessive powder had accumulated in the box. A lezking sezl on the vibratory screen drive shaft accounted for 45 g. Remote modification to the decanning station and changes in operating procedures eliminated these sources of loss after the first batch (29 oxide cans) had been processed. Sampling accounted for a loss of 100 g of uranium. Two 15-g and two 5-g salt samples were required in each of three production runs. In addition, several extrs samples were required for special purposes. The presence of heels in certain vessels and the holdup in transfer lines and filters were responsible for the majority of the loss. The heel in the reaction vessel amounted to 27 g. However, the storage vessel contained 346 g of uranium as heel as a result of a dip line replacement. The dip line was installed improperly and, as a result, about 1/3 in. of unrecoverable salt heel was left in the vessel. A summary of the uranium accountability is given in Table 9. er Table 9. Uranium Material Balance Uranium (g) (Wt %) Received as oxide 39,721 100 Finished product, LiFmUFA 39,952 98.1 Measured loss tc heels, semples, etc. Li6 1.1 Estimated loss to dusting, lines, filters, etc. 322 0.8 7. MAINTENANCE ENGINEERING Early in the run with depleted uranium, it became obvious that some of the equipment in cell G could not be considered reliable and would have to be repaired or replaced before the conclusion of the three production runs. Many of the deficiencies were corrected in the interval between the depleted-uranium run and the 233U’ run. Since it was impractical to consider major equipment modifications at this time becsuse of the fuel delivery schedule reguirements at the MSRE, emphasis was placed on: (1) storage of spare parts in the cell, (2) installation of redundant fittings, (3) procurement and modifica- tion of tcols and work tables in cell G, and (4) formulation of main- tenance procedures. 7.1l Spare Parts Because of the difficulties involved in obtaining access tc the hot cell, a generocus supply of replacement parts was stored inside the cell and cataloged Tor easy retrieval. Where applicable, the parts were prefitted to ensure compatibility with the system. Threads on the fittings and nuts and bolts were treated with a lubricant that is suitable for use in high-temperature environments. 68 7.2 Redundant Fittings The compression-type fittings used on the salt transfer lines and on the gas supply lines at the vessel nozzles were extremely difficult to make and bresk after they had been thermally cycled from room tempers- ture to 600-700°C. The passage of molten salt through one of these fittings made the operation even more difficult beczuse the salt formed a seal around the protrusion of the tubing into the fitting. BSince several of these fittings had to be routinely operated in the product transfer operations, there were several lines in the system for which no alternate had been provided. In these cases, redundant fittings were installed in such a manner that a second Titting was available 1if the primary fitting could not be operated. Although complex fitting arrangements are normally conducive to leakage, no such prcblems were encountered here. On two occasions, use of the standvy fitting was necessary in order to replace the sinbtered metal filter in the salt transfer line. 7.3 Tools and Work Tables A large assortment of hand tools, medified for use with the manip- ulators, were stored on tool racks in the cell. Usually the medification consisted of welding grips to the tool to aid in handling by the master- slave manipulators. A few special-purpose tecols, such as air=-operated snips (for cutting the interconnecting tuking of the capsules) and an impact wrench (for use with the electromechanical manipulator), were also stored for later use in the cell. End wrenches and other fixed- size tools were color coded by sizes for ease in identification. Work tables and trays were located between the viewing windows and the major pieces of equipment. They proved to be invaluable In the many mechanicael and maintenance operations conducted in the three sslt production runs because of the limited reach ©i the master-slave manip- ulators and also because of the limited field of wision available through the windows. A secondary, but very important, function of the trays involved retrieval of tools and miscellaneous items that were dropped S because of wear of manipulator parts or because of aswkwardness cn the part of the operator. When an item fell to the Tloor, rather than onto a tray, it usuzslly had to be replaced since retrieval from the floor in many locations cof the cell was impossible. The large heavy-dulty work table (Fig. 17) designed for the product container and capsule array disassembly operations served well in these functicns and was also used extensively as a general-purpose work bench. T - 7.4 Maintenance Procedures Remote maintenance of the process equipment was recognized, during the run with depleted uranium, to be a formidable task that could not be left to the operator’s discreticn when the need arose. Thus an engineer wac assigned, on a full-time basis, to develcop detailed main- tenance procedures that would be applicablie during any conceivable failure as well as during the planned maintenance operstions. A comprehensivé set of close-up photographs was taken of the equipment in cell G before the cell was sealed to process the radioactive material. These were later sugmented by photograpins taken through the chielding windows and the monocular viewers. Altogether, more than 300 photographs were made. Fifty-five detailed, step-by-step maintenance procedures were written by using the photographs and field sketches {more than 300) that had been used in the fabrication and installstion of the equip- ment. A complete list of tocls and materials necessary for the job was included in each procedure. Where appropriate, the maintenance proce- dure was referred to, cor inecluded in, an operating procedure to ensure continuity of a process cperation. Although it was not necessary to use all of these procedures in the fuel production, they were instru- mental in ensuring a safe and orderly completlon of this task. &. CONCLUSIONS 1. Reactor-grade eutectic salt concentrate, 7LiF- Jur, (73-27 2 27 233 UO2 or UFA 23 mole %), can be prepared by remote means from Li¥ and 33 by the process described in this report. 70 S 2. Hydrogen reduction and hydrofluorination reaction rates of oxide beds can be controlled satisfactorily by adjusting hydrogen and HF flows to avoid excessive temperature excursions from the exothermic resctions involved. Temperatures can be determined by placing a number of thermocouples within the beds. 3. Equipment for a remote process of this type should be designed with a minimum number of close-tolerance fits. It should alsc be de= cigned on & modular concept sco that components can easily be replaced, thus eliminating, if possible, remote repair operations. L. 1In general, commercial pipe, tubing, and electrical fittings are satisfactory in this type of operation. However, they are not satisfactory in instances where high temperatures or exposure to molten salt are expected. In these cases, special-purpose fittings should be developed. 5. Concentrations of the corrosion products nickel and iron can be reduced to satisfactorily low levels by hydrogen treatment of molten salt at 550°C. Chromium was not reduced by the process described. . 6. A binary mixture of UFu-LiF having a eutectic composition can be fused at a temperature of 650°C in the presence of a small amount of ligquid eutectic heel remaining in a vessel. In the absence of the lig- uld heel, it is necessary to heat the materials to the melting point of the LiF (835°C) to achieve liquefaction. 7. Tc chbtain a product containing less than 40 ppm of oxygen and to avold plugging of dip lines by the formation of oxides within them, it was necessary to install & catalytic recombiner and a molecular- gleve drying column in the hydrogen supply system. The use of Linde type L1k 1/16-in.-diam sieve material in a regenerative drying column reduced the moisture content of commercial-grade hydrogen to less then 3 ppm. 8. Maintenance procedures are necessary to ensure orderly com- pletion of a process of this nature. They are vital when entry into a contaminated cell 1s necessary and when equipment and the prcduct must be removed from the cell. Ig! 9. ACKNOWLEDGMENTS The authors express their appreciation to the following percons whose assistance was essential to the successful preparation of the enriching 233U Tuel concentrate for the MSRE: G. I. Cathers and J. H. Shaffer (Reactor Chemistry Division) for process development; E. L. Nichclson and W. F. Schaffer, Jr., for process equipment design; S. Mann for assistance in operations and data analysis; J. P. Jarvis (on loan to the Metals and Ceramics Division from the General Engineer- ing Division) for process equipment installation and modification, and for development of maintensnce procedures; and J. W. Anderson and D. M. Shepherd (both of whom are on loan to the Metals and Ceramics Division from the General Engineering Division) for facility engineering and supporting services. 10. REFERENCES 1. P. N. Haubenreich, private communication, December 1966. 2. J. R. Engel, MoRE Design and Operations Report, Part XI-A, ORNI~TM-2304 (September 1968). 3. Chem. Tech. Div. Ann. Progr. Rept., May 31, 1968, ORNI-U272, pp. 28«33, L. J. W. Anderson and J. M. Chandler, Safety Analysis for the Thorium- Uranium Recycle Facility (TURF), Building T793C, ORNI-4278 (%o be published). 5. C. J. Barton et al., "Phase Bquilibria in the Alkali Fluoride- Uranium Tetrafluoride Fused Salt Systems,” J. Am. Ceram. Soc., h(2), 63-69 (1958). 6. P. N. Heubenreich et al., MSEE Design and Operaticns Report, Part V-A, Bafety Analysis of Operations with 233U, ORNL-TM-2211 (February 1968), p. 37. T3 ORNL-4371 UC-80 — Reactor Technology INTERNAL DISTRIBUTIOR 1. Biology Library 76, C. E. Larson 2-l4, Central Research Library 77T. X. H. Lin 5-6. ORNL = Y-12 Technical Library 78. R. B. Lindauer Document Reference Section 79, A. P, Litman T-41, Laboratory Records Department 80. A. L. Lotts 42, Laboratory Records, ORNL R.C, 81. M. Lundin L3, R. G. Affel 82, H. G. MacPherson Wi, J. L. Anderson 83. R. E. MacPherson 45, J. W. Anderson 84, 5. Mann 46, C. F. Baes 85. H. E. McCoy 47, 8. E. Beall 86. J. R. McWherter L8, M. Bender 87. R. L. Moore 49, E, S. Bettis 88. J. P. Nichols 50, E, G. Bohlman 89. E. L. Nichoclson >1. 5, E. Bolt 90, A. S. Meyer 52. G. E. Boyd 91. A, M. Perry 53. R. B. Briggs 92-93, M. W. Rosenthal 54, R. E. Brooksbank 9L, Duniap Scott 55. W. D. Burch 95. W. F. Schaffer 56, W, 4. Carr 96. J. E. Shaffer 57. G. I. Cathers 97. M. J. Skinner 58-59, J, M, Chandler 98, J. W. Snider 60. F. L. Culler, Jr. 99. D. A. Sundberg 61. 8. J. Ditto 100. R, E. Thoma 62, W. P, Eatherly 101, W. E. Unger 63. J. R. Engel 102. J. E. Van Cleve, Jr. 64. D. E. Ferguson 103. A. M. Weinberg 65, L. M. Ferris 104, J. R. Weir 66. J. H Frye, Jr. 105. M. E. Whatley 67. R. E. Gehlbach 106, G. M, Watson 68. W. R. Grimes 7 107. J. C. White 69, A. G. Grindell 108. R. G. Wymer 70, R. G. Guymon 1069. Gale Young T1. P. N. Haubenreich 110. J. L. Youngblood T2. W. H. Jordan 111. P. H. Frmett (consultant) 73, P. R. Kasten 112. J. J. Katz (consultant) Th, J. W. Koger 113. J. L. Margrave (consultant) 75. C. E. Lamb 11k, E. A. Mason (consultant) 115. R. B. Richards (consultant) EXTERNAL DISTRIBUTION 116, D, E. Bloomfield, Battelle-Northwest, Richland, Washington 117. J. A. Bwartout, Union Carbide Corporaticn, New York 118, Laboratory and University Division, AEC, ORO 119-338, Given distribution as shown in TID-4500 under Reactor Technology category (25 copies — CFSTI)