• . ) I OF 1 ORNL P 2397 : robe -1 * . walan At . 1... . . ? 41 MICROCOPY RESOLUTION TEST CHART NATIONAL BUREAU OF STANDARDS - 1963 ORNL -P-2397 Confc660524_ EESTI PRICES 20 221 OPERATION OF THE MOLTEN SALT REACTOR EXPERIMENT HC. $ 1.00: MN Paul N. Haubenreich Oak Ridge National laboratory RELEASED FOR ANNOUNCEMENT ORNL - AEC - OFFICIAL IN NUCLEAR SCIENCE. ABSTRACTS A paper given earlier in this Symposium by Ed Bettis describes some concerts of future Molten Salt Breeder Reactors. It also describes Siuntio the strong incentives which exist for their development. The purpose of my paper 18 to describe the experience at Oak Ridge National laboratory with the Molten Salt Reactor 'Experiment - on: large step on the way to ima.... those Molten Salt Breeders of the future. owling Before summarizing the MSRE experience to date, I would like to ວິຣະປຣະທວນ. **LINO SAOŠANU SINONTHLDUB S give a brief description of this reactor experiment and where it fits . welcominciarieiali into the overall program. .... . . tone . If I had to compress this talk to the absolute minimum, I believe I could choose two pictures which would symbolize the meaning of the MSRE. The first would be the extremely detailed, 3-dimensional phase diagram for the LF-BeFa-ZrF4 system developed at ORNL in the course of deciding on the optimum fuel composition. This quite complete description of this salt system symbolizes the firm base of research and development on which the MSRE rests -- research not only on salt but also on the other materials of the reactor, going back fifteen years to the Aircraft Nuclear Propulsion Project. The second picture would be a photograph taken a few days ago, showing that this basic information has beer. used to build a red-hot reactor. That is literally true — for the photograph is one showing the coolant radiator, operating at an incandescent 1200°F. ....ime.now. ciarheid in die % winter.. Waw ain mo ba naman i e . ni... The point of this picture is this — in the MSRE we have an operating reactor which 18 demonstrating the practicability today of the basic ORNL - AEC - OFFICIAL on.co..., imam. Research sponsored by the U. 8. Atomic Enero Carmission wder con- tract with the Union Carbide Corporatio. .. . -:- - m S S o molten salt reactor features which make the breeder concept look so -- . r e.... ORNL - AEC - OFFICIAL -..'.- 12 * . . exciting. And that is the purpose of the MSRE. Description Let us begin with the flor:Bheet (Fig. 1). The MSRE 18 a one-region, 10-Mw reactor. It 18 not a breeder. The core 18 too small for that, only 5 ft in diameter. The fuel 18 a mixture of ?L1F, BeF2, ZrT4 with less than 1 mole % UF4. Fuel 18 circulated at 1200 gpa ty a sump-type centrifugal pump .. .. .. . .. TH= vs. Hulle located at the high point of the system. The free surface of the salt is ... and the surge space are in the pump bowl, which operates at 5 psig. The vapor pressure of the salt is much less than 1 mm Hg, and the ......nic.insdn 5 psig is merely a convenient pressure for the helium gas which is used . i .icial.m AEC .de.' ". to blanket the salt and sweep out fission product gases. Abcut 65 gpm of the salt is recirculated through a spray ring in the top of the pump tank. The purpose is to give better contact between the salt and helium and allow the relatively insoluble Xe and Kr to come out of the salt. .ini Four liters/min of helium, carrying the fission product gases, is sent . through charcoal beds to allow decay of most of the activity before being W .. inwestow.criclo't it seut up & stack. E- A tiny, motor-driven windlass, shielded and contained, is used to mbilia .. Wanawiam luwer 10-cc sample buckets into the salt in the pump bowl. This device is also used to lower small capsules of HF-UF4 eutectic, each containing 85 8 2350 into the salt to effect on-line refueling. Three flexible control rods in the core permit the temperature to be adjusted as desired. U.2timate shutdota is provided by draining the o le mani .. - ... salt. Incidentally, the fuel is never allowed to freeze. The liquidus temperature of the fuel is 813°F and all the piping 18 preheated electri- ......-, ORNL - AEC - OFFICIAL cally before the salt 18 put in. .co . com .. . When the reactor operates at power, hent is removed from the fuel and discharged to the atmosphere by a secondary loop containing another salt - LiF and BeF2. The circulation rate here is 850 gpm. ORNI - AEC - OFFICIAL Fig. 2 shows the physical layout. The primary system and the drain tanks are in underground cells. Mainter.ance is done with long-handled tools through openings in the roof. The main salt piping is 5-inch schedule-40 pipe. All salt piping and vessesl are made of a nickel-base alloy called INOR-8. This alloy is now available commercially as Hastelloy-N. The core is made up of graphite bars, in direct contact with the fuel salti' Fig. 3 shows the core being assembled. The 2 x 2 stringers are made of a special grade of graphite, practically impermeable to the salt. Experience In very brief outline, our operating history has been as follows. We began preliminary testing of MSRE components and systems in the fall of 1964. In January, 1965, we began to circulate salt in the fuel and coolant loops. On June first, last year, the reactor was first made critical. We carried out low-rower experiments in June and July. Then after several months of final preparations, the reactor was operated at a significant power (1 Mw) in January. We ran into some difficulties with our offgas system that delayed our escalation of the power, but we have now operated at powers up to 5 Mw. We expect to be at 10 Mw within a couple of weeks. Now, what have we learned from the operation thus far? Nuclear Characteristics - First, let's look at the nuclear charac- teristics, itemized in Table I. We can summarize by saying that the charac- teristics are extremely close to the predicted values. After this table ORNL - AEC - OFFICIAL Je .. 411 9. WH . . ' . N can was made, we conducted dynamics experiments which confirmed our prediction ORNL - AEC - OFFICIAL that the system 18 stable at all power levels. While we're on the subject of reactivity we might mention that the xenon poisoning has been quite low. The two main variables affecting - ----- - - - the poison are (1) the mass *ransfer coefficient for xenon migration into the graphite, and (2). the fraction of the dissolved xenon which is stripped from the salt passing through the spray ring in the pump bowl. We determined mass transfer in the core last year by an experiment which involved saturating the fuel and graphite with krypton, then purging it out. Because of the time constants we could not measure the efficiency of the gas stripper in this experiment. Our experience with xenon poisoning has now given us cn indication of the stripper efficiency. With no stripping we would have about 1.7% 6k/k equilibrium poisoning at 5 Mw. We actually got only 0.3% 8k/k which corresponds to about 60% stripping efficiency. Corrosion - Here is a subject that is a real pleasure to talk about. We've just had practically no corrosion. This is really no surprise because loop experience with fluoride salts in INOR-8 has always shown very low corrusion at MSRE temperatures (around 1200°F). The corrosion 2 - that does occur 18 reaction of salt impurities such as FeF2 with the - - INOR-8 to remove chromium, as CrF2, from a thin surface layer. This attack : -- : stops when the impurities in the salt are used up. We put in quite pure : -:- . salt so there has been very little CrFa formation. We have had fuel salt in the MSRE since April, 1965. (Before that we circulated flush salt.) It has been molten all those 13 months, and during that time it has been circulated for 2400 hours. The chromium concentration in the salt started at 38 ppm; today it is only 45 ppm. This increase corre- sponds to removal of chromium from a layer only 0.006 mil in thickness ORNL - AEC - OFFICIAL over the loop surfaces. Confirmation of practically no corrosion came when we examined the fuel pump internals last fall and found no sign ORNI - AEC - OFFICIAL of attack. Semple specimens exposed in the core have given the same picture. Fuel Stability - A subject that is of great importance in any fluid-fuel reactor 18 fuel stability. One asks, "Can high temperature or radiation or some contaminant cause the fuel to break down, separate or Yorm precipitates?" The fluoride salts are not susceptible to any such effects with one exception: gross contamination with oxygen or water produces oxides which are not highly soluble. The MSRE 16. designed to prevent oxide formation and we have been signally successful. The oxide content of the fuel, after 13 months in the reactor, is only 50 ppm compared to & solubility of about 700 ppm. Equipment Performance and Reliability - The components that are perhaps most central to the salt operation are the circulating pumps. So far, the fuel pump has circulated salt for 3300 hours and the coolant pump has operated with salt for 3500 hours with no sign of any trouble whatsoever. We have no mechanical valves; instead we blow air on a flattened section of pipe to form a freeze plug. These "freeze valves," once frozen, are reliable. We had some trouble getting the controls properly set up to freeze as quickly as we would like, but this hasn't been a real hindrance. The sampler-enricher 18 another component that comes into contact with the salt. Last week & short in the electrical wiring to the motor put the sampler out of service for the first time. Before then, we had taken 120 samples and added 87 enriching capsules without incident. We ORNL - AEC - OFFICIAL decontaminated the sampler, repaired the trouble, and it should be back in service today or tomorrow. VULUI UNI T . * .." . . . One component that did require a good bit of work to get it to operating properly is the coolant radiator enclorure. The doors that ORN - AEC - OFFICIAL drop to stop heat removal had to be modified before they would operate reliably. . All the other other components and systems are more or less pe- ripheral to the salt operation. One system that is less perlpheral than others 18 the offgas system. In the past 4 months, during power cperation, we have been bothered a good bit by partial plugging at some points in the offgas system. The most notable places are at the pressure control valve and at the inlets to the charcoal bed. We opened up and inspected the trouble spots. In every case the lines looked clean, but there were small quantities of solid or viscous liquid on the valve trims. Analysis was hampered because of the intense radioactivity but the material was snown to be almost exclusively hydrocarbon with practically no salt. The total quantity was only a few cc and we think we know the source of the material. The holdup volume in the cffgas line is normally at the tempera- ture of the reactor cell (about 130°F). But when the reactor is operated at power, the fission product gases heat the pipe up above 200°F. We believe that a small quantity of oil accumulated in the holdup volume before power operation began; that the heating due to the fission product gases evaporated part of the oil and the intense radiation converted part of it to crosslinked polymers which plugged the valves. We have . recently installed a filter and trup assembly in the offgas line up- stream of the pressure control valve and this seems to be working 2 effectively. . * . Of course there 18 much more equipment than I have mentioned and ORNL - AEC - OFFICIAL . we have had the usual amount of trouble with conventional components. For example: The motor on a space cooler in the reactor cell failed ORNL - AEC - OFFICIAL and when we examined it, we found that the rotor had slipped on the shaft. This 18 an almost unheard type of failure but it did happen and it succeeded in shutting us Jown until the motor could be replaced. All in all, however, mechanical difficulties with equipment has caused very little delay in the program. Summary In summary then, our experience has been as follows: 1. Nuclear characteristics are as predicted. 2. Xenon 18 effectively stripped. The system is stable at all powers. 4. There has been practically no corrosion. 5. The fuel 18 chemically quite stable. 6. There have been some equipment difficulties, but nothing basic or very difficult to overcome. Several of us who are now operating the MSRE participated in the 11l-fateå aqueous homogeneous reactor experiments at Oak Ridge. You 1 may be sure that we really appreciate how smoothly the molten-salt system operates, and how free it is from worrisome basic limitations. It certainly looks good in comparison. But no matter what the standard of comparison, we are confident that the MSRE will live up the name which We think goes with the initials M. S. R. E. – Mighty Smooth Running Fbxperiment. LEGAL NOTICE + - - . The report nuo preparar, u w accont of Gorenmat sponsored work. Neither the United Hatre, por the Commisslot, por may person acting ol bolall of the compladou: A. Makes my warranty or reprenotation, expreted or implied, with respect to the acco- macy, completenes, or nahulness of the information coatalpod to this report, or that the of may waloraation, appenat, molhad, or process ducloued to the report may not latring privately owned ricots; or B. Aasunen nay liabillies with respect to the one of, or for dumuru ralties from the un of an: taforation, appuntos, method, or proceso dircloud laws reports ORNL - AEC - OFFICIAL Al wood to the abome, "porno stag od behalf of the Coanda" loladas may on- plogue or contractor of the Councuskoo, os emplogue of much contractor, to the attent that fruch employee of costructor of the Conninaloa, or employme of much controcor properne, dionautnatos, or provides access to any tabormation purreant to Me omaptogenast or contract with the Counterton, or No employment will such contractor. ie women .. . exercer 20,- Hym.:-.:,:.. ...53m- ..., Oral-DWG 65-4140 FUEL PUMP SAMPLER- ENRICHER COOLANT PUMP SAMPLER LEGEND FUEL SAU COOLANT SALT HELIUM COVER GAS RADIOACTIVE OFF-CAS TO ABSOLUTE FILTERS - M 1325 F ABOUT Iouem 850 G.P.M. OFF-GAS HOLOUP HEAT EXCHANGER 1225 °F OVERFLOW TANK 1175 OF AIR FLOW: 200,000 cm ABSOLUTE FILTERS 1200 G.P.M. El W 1100 F BLDG. VENTILATION REACTOR VESSEL FREEZE FLANGE (TYP.) opens ace with whom a STACK FAN FROM COOLANT SYSTEM -FREEZE VALVE (TYP.) RADIATOR smo.com te is...-15. letu . cov ю ABLETE FUTERS WATER WATER STEAM MANN CHARCOAL STEAM DRUM BED STEAM ORUM AUX. CHARCOAL BED QOQLANTI DRAN TANK 6 . FUEL ORAIN TANK NO.1 U10-U10168 FUEL STORAGE TANK FUEL ORAIN TANK NO 21 FUEL FLUSH TANK SODIUM FUORIDE BED . Fig. 1 .. ,-- ..ver TE . ORNL - AEC - OFFICIAL UNCLASSIFIED ORTIL-OWG 63.12091 REMOTE MAINTENANCE CONTROL ROOM REACTOR CONTROL ROOM f . 1. REACTOR VESSEL 2 HEAT EXCHANGER 3. FUEL PUMP 4. FREEZE FLANGE 5 THERMAL SHIELD 6. COOLANT PUMP 7. RADIATOR 8. COOLANT DRAIN TANK 9. FANS 10. DRAIN TANKS II. FLUSH TANK 12 CONTAINMENT VESSEL 13. FREEZE VALVE MSRE Flow Diagram. HU ORNL - AEC - OFFICIAL Fig. 2 . ***. PRILI O 17.; W * ata . m . menos. LIKsh. Salinan mpya ***ille; o arieomt *- ......... . .. : . . . .. Fig. 3 -10- rote wen. 1 RX mit den . ha tendraadtuk anasto ..... , . aims .... *D ra na .. W : : ! ww.com mais 1 w LogoW.7110 ORNL - AEC - OFFICIAL ORNL - AEC - OFFICIAL - - - - - - - - - -- - - - - - - VIJI:10 - 33V - INYO WIDISJO-734-1Ndo . ORNL DWG. 65-11991 MSRE NUCLEAR CHARACTERISTICS Comparison of Predicted and Measured Values Predicted Initial critical concentration of fuel salt (g 235U/l). 33.06 Measured 32.85 +0.25 235U conc. coeff. iscic 0.234 -7.0 x 10-5 0.22 0.223 -7.3 x 10-5 0.212 Temperature coeff. (Ok/k) (°F Reactivity effect of circulation (% ok/k) Rod worth at initial critical loading, 51-inch travel (% ok/k) Rod 1 Three-rod bank 2.11 5.46 2.26 5.59 Table I . E * S4 WN M. . NI X. 771 BILA , . de 7 M .: :. . .." ti END DATE FILMED 11 / 1 / 66 . . . MOH ..