O ——— T ay Wl OR T MASTER CoRPY OAK RIDGE NATIONAL LABORATORY operated by UNION CARBIDE CORPORATION NUCLEAR DIVISION CARBIDE for the U.S. ATOMIC ENERGY COMMISSION 1 0 ORNL- TM-2696 2 © | - COPY NO. 130 A ¥29 OCT 24 =oH DATE - September 3%, 1969 BSE BBS0ID: ASSESSMENT OF MOLTEN SALTS AS INTERMEDIATE COOLANTS FOR IMFBR'S H. F. McDuffie H. E. McCoy R. C. Robertson Dunlap Scott R. E. Thoma Abstract Several molten salts were considered as intermediate coolants for IMFBR's., Included were fluoride, chloride, carbonate, nitrate-nitrite and fluorcborate salts. Chemical reactions that could occur between sodium and fluoroborates lead to the conclusion that carbonates might be a better choice for LMFBRs. Use of carbonates avoids the safety considerations and related costs that arise from the reactions of sodium with water if a steam generator fails and with air if & coolant pipe ruptures. In the absence of these safely considerations, sodium is clearly superior to the molten salts as an intermediate coolant for ILMFBR's because the lower thermal conductivity and higher viscosity of the salts would result in higher equipment costs. Keywords: coolants, fast-breeder reactor, liquid metals, fused salts, molten salts, fluoroborate, carbonate, sodium. o NOTICE This document contains information of a preliminary nature ond was prepared primarily for internal use at the Oak Ridge MNational Laboratery. It is subject to revision or correction and therefore does not represent o finol report. - LEGAL NOTICE This report was prepored as an eccount of Gevernment sponsored waork. Neither the United States, nor the Commission, nor any person acting on behalf of the Commission: A. Mokes any warranty or representotion, expressed or implied, with respect to the accurocy, completeness, or usefulness of the information contained in this report, or that the use of any information, apporatus, method, or process disclosed in this report moy not infringe privately owned rights; er B. Assumes any licbilities with respect to the use of, or for domages resulting from the use of any information, apparatus, method, or process disclosed in this report, As usod in the above, '""person octing on beholf of the Commission' includes oany employee or contractor of the Commission, or employee of such contractor, to the extent that such employee or controctor of the Commission, or employee of such controctor prepares, disseminutes, or ! provides access to, ony informotion pursuont to his employment or contract with the Commission, or his employment with such cantrocteor. 1 fi— - Table of Contents Introduction .« « ¢« o o ¢ ¢ ¢ « « o s o o 2 o o o o Summary and Conclusions . . . . . . . . Requirements for an LMFBR Intermediate Coolant . Absolute Requirements ,:. . . . Trade-0ff Requirements . . . . . . . A Survey of Possible Molten-Salt Coolants . . . Fluorides . o ¢« ¢ ¢ o o o ¢ o o s o s o o o Chlorides « + o ¢ v ¢ v ¢ o o o o o o o o Carbonates . « « « & v ¢ v ¢ ¢« ¢ & s o o o Nitrate-Nitrite Mixtures . . . . Fluoroborates . « « & o o o« + o o + « o 0 Evaulation of Fluorcborates to Illustrate Use of Molten Salts for IMFBR'sS + + « & & Suitability of Coolant Salt for Operating Conditions of ILMFBR Cycle and Engineering Design Changes Required for Its Use . . . . Compatibility with LMFBR Materials Including Effects of Radiation . . . . . Effects of Safety and Accident Conditions . Availability and Cost of Fluoroborate . Development Requirements for LMFBR Use on Molten Cycle Choice . . . & & &« v o o0 Development Program for the MSBR . . Evaluation of Carbonates for IMFBR Use . . . . References v ¢« v &4 o o o o o s o o o o o o o \n ® @ -3 ~3 -3 O 11 11 16 19 21 2l 30 31 32 55 33 33 3L 55 - = _“— ASSESSMENT OF MOLTEN SALTS ‘AS INTERMEDIATE COOLANTS ‘FOR IMFBR'S H. F. McDuffie H. E. McCoy R. C. Robertson Dunlap Scott R. BE. Thoma Introduction The Division of Reactor Development and Téchnology'of the AEC asked ORNL to assess the use of molten salts as possible coolants for the inter- mediate loop of an LMFBR. Consequently, a group conéisting of the authors of this report was constituted to prepare the assessment. ' Initially we assumed that the fluoroborate-fluoride mixtures that appear to be of most interest for molten salt reactors would be good choices for IMFBR's, and most of the effort was directed towards evaluat- ing the use of fluoroborates forAfast reactors. Much of a report was pre- pared discussing fluoroborates and the status of the development program that will qualify them for use with molten salt reactors. As the assessment proceeded, it became clear that salts other than - fluoroborates might be more appropriate for IMFBR's. The report was re- vised aécordingly, but some of the already prepared material on fluoro- borates was left in because it illustrated the factors that must be con- sidered in the design of a molten salt intermediate system and indicates ‘the types of development activities that would be required for evaluation of any molten salt for IMFBR use. Ssummary and Conclusions 1. The use of molten salts as heat-transfer medie is well -founded on long-standing technology.- < . This report was written in 1968 but not issued as a formal document. Because of expressions of interest in the subject, it 1s being issued now with minor corrections but without revision. Oon ~_. 2. The use of lithium-beryllium fluoride in the MSRE has been fully - satisfactory, but it would be desirable for large reactors to have a coolant that has a lower liquidus temperature and a lower cost. 3. The use of fluoroborate-fluoride salt mixtures appears attractive for large scale molten salt reactors on the basis of low liquidus tempera- tures, low cost, low vapor pressure, and good compatibility with Hastelloy N. Development is in progress in connection with the proposed demonstration of fluorcborates as suitable intermediate coolants for molten salt reactors. i, The use of fluoroborate-fluoride salt mixtures as intermediate coolants for an IMFBR would eliminate the possibility of a Viblent reaction ~of sodium with water due to a leak in the steam generator. However, an equally exothermic reaction (to give insoluble boron and soluble N;F) could occur if a leak in the primary heat exchanger allowed sodium fo get into the fluoroborate salt. The implications of such a change in.the loca- tion and nature of a potential hazard need to be considered. _ ~ 5. The use of molten carbonate salt mixtures for intermediate cool- ant in LMFBR's deserves serious consideration because of their cofibination of low cost,rreasonably low liquidus temperatures, low vapor pressuré, compatibility and probable freedom from violent reactions with either sodium or steam. _ | | 6. s_bme' consideration should be given to the possible use of nitrate- nitrite heat_tfansfer fluids as,intermediéte coolants because of the good match of their physical prgperties to the temperature range of interest to ~IMFBR's and because of the extensive industrial experience with the ac- ceptance of thesé fluids fdr heat transfer. Again, the probable exother; mic reaction of metallic éodium with the melt represents the tfansfer of a hazard from the steam generator to the primary heat exchanger. 7. The use of chloride or other fluoride mixtures does not appear attractive at the present time. | 8. An effective progrém to develop a molten-salt intermediate cool- ant system for LMFBR's could be performed by ORNL in conjunction with its present development of coolants fbr MSBR's. 7 Requirements for an ILMFBR Intermediate Coolant In assessing intermediate heat transfer fluids it is possible to group the significant pafameters roughly as follows, Absolute Reqguirements 1. The salt must be compatible with the container materials and adequately stable to the radiation which it will encounter. 2. The melting point and vapor pressuré of the salt must be such as to permit the system to be operated within the temperature limits desired. | 3, The viscosity and thermal properties must permit the use of acceptable heat exchangers, steam generators, and coolant pumps. L. The consequences of an accidental mixing of the salt with sodium or steam must be Wifhin the design capabilities and not imply catastropic situations. 5. The consequences of an accidental cooling of the system must be reversible. Trade-0ff ReQuifemefits 1. The corrosion rate of the container should be low. 2. The liquidus temperature of the coolant should be low. 5. The vapor pressure of the system should be low and any condensed vapor should not be a solid with a high melting point. 4. The viscosity and density of the coolant should be low. 5. The thermal capacity and thermal conductivify of the coolant should be high. i 6. The price of the coolant should be low. 7. Large amounts of the coolant should be available in high purity. 8. It should be possible to separate the intermediate coolant from the primary sodium coolant if they are accidentally mixed, and the con- sequences should not be such as to leave neutron absorbing poisons or moderating elements in the primary coolant.circuit; | | 9. It should be possible to make up for coolant losses due to radiation decomposition. 10. The consequences of mixing the coolant with steam or water should be easily reversible. 11. A-leak of intermediate coolant into the primary coolant circuit should be readily detectable. ' 12. Engineering scale experience with the coolant should be available. 'It_is obvious that questions of economics, maintenance lifetime, operating inconveniences, etc.,'afe trade-off items whichvmust ultimately be balanced againét the various technical items. There are many such trade-off items for every coolant considered; it is important not to ex- clude any candidate from further consideration until it is clear either that the absolute requirements cannot be met or that the trade;off items are overwhelmingly unfavorable. A Survey of Possible Molten-Salt Coolants Fluorides Manj fluoride mixtures meet the absolute requirements stated earlier. The lithium-beryllium fluoride mixture used in the MSRE was selected be- cause some of.it leaking into the fuel would not contaminate the fuel salt with nuclides that make it unusable. It was also very satisfactory chem- ically, and most of'its physical properties were acceptable as seen from inspection of the values in Table 1. Table 1. Properties of LieBeFu Melting temperature (peritectic) - 857°F (L458°C) Liquid density (538°C)(1000°F) 1o4.1 1b/£t° p(g/cm3)n= 2.21h14.2 x 107" ¢oc Crystal density (X-ray) | 2.168 g/cc Coefficient of thermal expansion 2.1 x 1077 (°c)™t Surface tension (857°F)(L58°¢C) 250 dyne/cm Vapor pressure (857-1200°F) <0.l torr Viscosity 1200°F (649°C) 6.8 centipoise : 1000°F (538°C) 11.9 centipoise Liquid thermal conductivity 0.011 watt (c:m-°C)-l 0.64 Btu/hr-ft-°F TEMPERATURE (°C) ORNL-DWG 66-7632R 900 °4e \ 800 - 700 LiF+LIQUID 500 ' - - \ / | - . BeF, (HIGH QUARTZ TYPE) | | \\ / 2 Teluin . 400 - \\Z 360 2LiF-Bef,+Bef, (HIGH QUARTZ TYPE) LiF+2LiF-BeF, 300 30 o~ ' | 280 @ 2LiF-BeF, u | LiF-BeF, + BeF, (HIGH QUARTZ TYPE) w 4 D | |iF-BeF, + BeF, (LOW QUARTZ TYPE) 220 2 LiF-BeF, & | l . \ - | — i ] l : 200 . _ — . - ‘ LiF 10 . .20 30 40 50 . . 60 70 - - 80 90 = Bef, BeF, (mole %) Fig. 1. Phase Diagram of the System LiF—BeF2. 0T Although the liquidus temperature can be lowered further by the addi- tion of a higher percéntage of beryllium fluoride (as‘seen from the phase diagram in Figure l), this is at the expense of a rapidly increasing viscosity, which would impose severe economic penalties. The cost and the inconvenience of dealing with beryllium would handicap the use of lithium- beryllium fluoride as an intermediate coolant for an IMFBR, and there is no advantage to using only lithium and beryllium for fast reactor coolants. Coolant compositions which have liquidus temperatures below 400°C (752°F) can be found in the NaF-BeF,, systeml and in the N’aF-LiF-BeF2 system.2 In the latter system, temperatures as low as 315°C (599°F) have been reported. These materials are almost certainly compatible with Hastelloy-N and possess adequate specific heats and low vapor pressures. They should not undergo violent reactions on mixing with sodium or water; sodium‘should reduce the beryllium to metal and water would generate HF and precipitate BeO, but these consequences would be reversible by ap- propriate clean-up treatfient except for the possibility of deposition of metallic beryllium in an inaccessible form. The viscosities of these fluo- ride salts at low temperatures are certainly higher than are desirable. It is possible that substitution of ZrFLL or AlF5 for some of the BeF2 will provide liquids of lower viscosity at no real expense in liquidus temperature. The eutectic composition of lithium-sodium-potassium fluoride (i46.5- 11.5-42.0 mole %) melting near 455°C (851°F) is quite well known and should be relatively stable to mixing with metallic sodium or with water. Its liquidus temperature is probably too high for consideration. Stannous (tin II) fluoride, SnF,, which melts at 215°C (419°F), has been suggested several times as a fuel solvent or coolant in molten salt technology. It is available in large quantities and in high purity, largely as a result of its use in toothpaste. We have excluded considera- tion of it, nevertheless, because of its ease of reduction or, alternatively, its high oxidizing and corrosive power; it is similar to PbF2 and BiF5 in this respect and could not be contained in nickel-based or iron-based alloys but would require something more noble such as molybdenum or graphite. There are essentially no other fluoride mixtures melting below LOO°C (752°F) which do not contain either beryllium fluoride, hydrogen fluoride, 11 or ammonium fluoride as a component; consequently, it is believed that it would be unprofitable to concentrate a search in the field of fluorides beyond the limits already outlined. Chlorides Chlorides have always been considered potentially useful heat transfer fluids.5 It would certainly be possible to fifid mixtures with low liquidus temperatures and low viscosities. The thermal properties should be com- petitive with those of fluorides. The vapor pressures are likely to be higher. The corrosionLL and radiation stability are likely to be less favorable. | o Many chloride mixtures are known which melt below 200°C (392°F); these usually cpntain a relatively volatile chloride, such as_ZrClu, I\TbCl_5, and AlCl5 o3 PbClg, or GaCl_.. Either of these factors makes the mixtures less attractive for 5 use in an LMFBR. The consequences of accidental leakage of chlorides into , Oor an easily reduced chloride such as C4Cl fluoride fuels, sodium, or water are likely to be worse than those of a fluoride leak. The effect of chlorides on stréss corrosion cracking in. the steam generators would be a matter for considerable concern. We believe that a satisfactory intefmediate LMFBR coolant will not easily be found among the chloride mixtures and, if one fiere found, it would only be after a large development effort to demonstrate compatibility. Carbonates Molten carbonate mixtures have been used extensivély as heat transfer baths in metal working, and consideration has been given to their use as coolants for molten salt reactors.5r Due largely to the work of Janz and his associates at Renssalear Polytechnic Institute a number of properties of molten carbonates have been established. Figure 2 presents a phase diagram of the ternary system LiECO5-NaECO5-KéCOZ. The eutectic of the composition 4%.5-31.5-25.0 mole % is reported to melt at 397°C (TL7°F) by Janz et al., but the composition 26.8-42.5-30.7 mole % is reported to melt at 393°C (739°F) by Rolin et al.! ternary composition, m.p. 397°C (T747°F) was L40-30-30 mole %. Janz and Saegusa8 reported that the 12 ORNL-DWG 64-T146 Na,CO3 858 A - 800 COMPOSITIONS IN mole % 750 TEMPERATURES IIN °C ~700 \ MIN. 705° E-497TA 550 / NE =397 450 S -~ ) | S T o | O ~ 500 \ & Q3 LiKCOx NE-498 901 mp 504.5 °C Fig. 2. The System LiQCOB—NaeCO3-K2003; Modified from Janz and Lorenz, J. Chem. and Eng. Data Vol. 6, No. 3, 321-323 (1961). S 9 092 0™ — ‘009 Q LinCO3/ \O\/ \ VYo 726 E£-482 ~ 15 Subsequent studies by Janz have indicated that the dissociation pres- sures over carbonate melts should not exceed one atmosphere in the temper- ature range of interest (1lithium carbonate has a pressure of 501 mm at 8u3°C (1550°F). The ternary carbonate mixture has been used for a‘number of years and is known to be noncorrosive to steel at 1LOO°F (760°C) over many months of exposure; no obvious corfosion was observed after about 4000 hr of exposure at 1200°F (649°C) to INOR-8. In tests at ORNL, Bettis reported9 at temperatures of 900-1000°F (u482-538°C). ORNLlO has reported the enthalpy and the viscosity of a terhary carbonate mixture (Li-Na-K, 30-39-32 wt Q) (41-36-23 mole %) over the temperature range (887-1319°F) L75~715°C with the liquidus temperature indicated as being near 390°C (734°F). The derived heat capacity of the that molten carbonate was apparently stable toward molten lead salt was 0.413 cal/g°C; the kinematic viscosity, based on efflux-cup measurements, was reported to be given by the expression v = 0.02h exp(u4818/T°K) centistokes and the density was estimated, assuming ideal solution, to be 5 = 2.212 — 0.00039 T°C grams /cm From this, the viscosity was calculated to be 33.5 centipoise at L60°C (860°F) and 5.98 centipoise at 715°C (1319°F). The mixture was proposed for use in out-of-pile development studies relating to the MSRE because of ité similarity in properties to the fluoride salts and because it is essentially noncorrosive to stainless steel without a protective atmosphere. Figure 5 shows the viscosity of this mixture as a function of the tempera- ture predicted by the early ORNL workers. , ‘ Janz and Saegusa8 have reported considerably lower values for thé viscosity of the ternary eutectic mixture, LiQCOE-NachB-K2003 (L0-30-30 mole %), m.p. 397°C: C VISCOSITY (Ib/hr - ft). ORNL -LR-DWG 73993 - 30 70 " N 25 0 ~ | —20 - Qo wn . ‘g \ - - 40 = — 15 S 30 : = \ — 10 b_)’ “ \ > _ — . o . — 5 {0 0 . - — o 800 - 900 | ‘ {000 1100 - 4200 - | - 4300 : Fig. 3. Viscosity of Li TEMPERATURE (°F) 2C03-Na2003—K2CO3(hl-36—23 mole %). 15 T (°C) N(poise) 183 (901°F) 0.0584 484 (903°F) 0.0547 539 (1000°F) 0.0323 598 (1110°F) - 0.0237 600 (1112°F) 0.0207 These‘values were reported subsequent to the ORNL values and were measured with an intrinsically more accurate and precise technique in e laborstory devoted to measurements on many carbonates. They.are much more favorable with respect to the use of carbonates as coolants. We are not aware of any reported measurements of the thermal con- ductivity of molten carbonates, but it is expected that the values will be near to those for molten nitrates and fluoroborates. | The consequences of an accidental introduction of molten carbonate into molten fluoride fuel are believed to be intolersable; it is expected that the carbonate would dissociate, with the carbon dioxide being re- leased and the residual oxide causing massive precipitation of insoluble oxides of uranium and thorium. It is likely also that the introduction ~of the foreign cations would be essentially irreversible. No direct tests have been performed to measure the results of mixing of carbonates and ‘fluorldes The possibility of using carbonates in proximity to metallic sodium raises less apprehension with respect to the consequences of a leak. Certainly it would be necessary to remove oxide from the sodium metal in order to control corrosion, but no dire consequences of a leak of sodium into the molten carbonate are foreseen. The possibility of a reaction of metallic sodium with sodium carbonate was examined briefly.ll The reaction 2Na + Na2C05(l) = 2Na20(s) + co(g) AF° -203 -131 =479 is unfavorable in free energy at 1000°K by about 24 kcal. A leak of carbonate into the steam generator (although unlikely because of the pressure differences) would require that the generator be thoroughly 16 flushed; this seems feasible since the carbonates are quite soluble in water. A leak of steam into carbonates would probably be reversible by side stream treatment with carbon dioxide. The effects of radiation on molten carbonates, particularly gamma radiation from a primary sodium coolant fluid, have not been determined. Since the effects of gamma radiation on molten fluorides and molten fluoroborates have been found to be negligible, and since carbonates are thermodynamically quite stable, it is not anticipated that radiation effects would be severe. . ‘ | If the liquidus temperature as high as 750°F (5996C) would be accept- able, carbonates would appear to merit seriofis additional consideratien as interfiediaté»coolants for IMFBR's. Nitrate-Nitrite Mixtures 'Mahy inorganic nitrate-nitrite mixtures have been used as heat transfer agents for high temperature industrial processes. Mixtures of cemmercial interest are illustrated by HTS (Heat Transfer Salt e'also DuPont Hitee),pa_eutectic mixture of NaNOB—KNOB--NaNO2 (7-53-40 wt %) which has a melting point of 288°F (1k2°C). HTS has been proposed for use in the temperature range 300 to 1000°F. Heat transfer and thermal property measurements with HTS were first reported in 1940.12 The authors: also investigated the corrosion, thermal stability, and handling of this salt mixture. Hoffman at ORNL has studied the heat transfer charecteristics of HTS flowing by forced cofivection %) through eifcular tubes and reported his results in 1960, The variations of density and viscosity'of-HTS with temperature are given by Figure 4. The heat capacity was reported as 0.373 Btu.lb_l(°F)'l for the liquid. ' Leeteemy L. com- parison of the effectiveness of several coolants was provided by means The thermal conductivity was reported as 0.35 Btu hr of the "cooling-work modulus" (the flow work per unit heat removal) de- rived by Rosenthal, Poppendiek, and Burnett.,lLL Hoffman has reported such a comparison of coolants in Figure 5, with the properties of HTS extra- polated to 1350°F for consistency. This comparison shows that HTS requires 10-20 times the pumping power required for sodium or FLINAK (NaF-LiF-KF eutectic). VISCOSITY (Ib-ft— - hr ) 17 ORNL—LR—DWG 44318 36 = 1110 32 28 n H n O o N AN T~ 0 300 400 500 600 TEMPERATURE (°F) 700 800 {30 100 Fig. 4. Density and Viscosity of HTS (nitrate-nitrite eutectic). DENSITY (1b/ ft2) 18 -2 ORNL-LR-DWG 44347 10 ' HTS 90 T, COOLING WORK MODULUS 10 FLINAK 1075 710" -2 -1 2 4 6 8 0 10 d, REACTOR TUBE DIAMETER ( ft) Fig. 5. Relative Heat Transfer Effectiveness of Reactor Coolants. 19 Irradiation of HTS to a dose of 3.3 x 1018 thermal neutrons/cm? and an accompanying epithermal dose of somevhat less than half.the thermal - dose was reported by Hoffman to have been performed by O. Sisman of ORNL. The irradiated samples were said to have become more hygroscopic, and some breakdown to gaseous products was reported. As a consequence of high- temperature thermal breakdown or radiation-thermal breakdown it would -seem appropriate to arrange treatment of a.bypass stream with NEO5 or Né03 to regenerate the desired composition, but this was considered to ‘pose no more difficult engineering problems than those involved in the use of organic coolants. A leak of sodium into an HTS salt mixture would cause an exothermic ~reaction to form sodium oxide and liberate nitrogen or nitrogen oxides. ‘The heat liberated would be of the same ordér of magnitude of that in- volved in the. sodium-water reaction. The chemical consequences in‘the salt would be reversible by treatment with nitrogen oxides. If salt leaked into the primary sodium system of an IMFBR, a similar reaction would occur and the resulting sodium oxide would have to be removed by appropriate traps. . The nitrate-nitrite salts appear to present no insurmountable dif- ficulties, but their use would involve a number of disadvantages in the trade-off area. Whether their compatibility with structural materials and the large industrial use which they have enjoyed for heat transfer purposes is sufficient to offset these disadvantages is a question to be resolved by more detailed engineering evaluation. Fluoroborates After a survey of the materials considered available, itAappeared that fluoroborates, especially a mixture of sodium fluoroborate and sodium fluoride, offered the greatest promise for development as intermediate coolants for molten salt reactors. Liquidus temperatures as low as 380°C ' (716°F) were available. The cost of materials is known to be very low (less than $0.50/1b for material of high purity). The vapor pressure of BF, above the melts Has been found to be relatively low (less than one 3 atmosphere). The corrosiveness of the material to Hastelloy;N appears to 20 be low. No violent exothermic reactions occur when fluoroborates are mixed with steam or with fluoride fuel salts. In fact, it has been dis- covered that fluoroborates are essentially immiscible with molten mixtures - of lithium and beryllium fluorides. Uranium and other tri- and tetravalent elements were not extracted into fluorcborates, and no high-melting com- pounds were found when sodium fluoroborate was equilibrated with a fluoride salt mixture of LiF-BeFE-UFu-ThFu. Operation of a test loop (containing residues of this fluoride -salt) with a flushing charge of NaF-NaBF) did, however, reveal the deposition. of a green salt in the upper region of the pump bowl. The composition of the salt was essentially 7NaF°6(Th,U)Fu, suggesting that either entrainment of the residue or solution-deposition of it had oecurred, along with some replacement of Li by Na; although more study of the immiscibility phenomenon is indicated, there:is -no informa- tion available to cause alarm over the possibility of accidental mixing of fluoroborates with fluoride salts. For MSBR use, moreover, the ac- cidental introduction of fluoroborates into the circulating fuel would cause..a large reactivity decrease because of the boron, and thus even a small leak would be quickly detected. The boron could be -easily removed from the fuel salt by treatment with HF. .For LMFBR use, a leak of sodium into the fluoroborate would be ex- pected to result in immediate and complete reaction to produce sodium fluoride and elemental.boron.15 In early work, boron trifluoride was . ‘reported to have been passed over red-hot potassium in a gun barrel to produce boron and KF, and metallic sodium or potassium heated in boron trifluoride were reported to react with the production of fire to give boron and the metallic fluoride. Calculations suggest that the heat liberated'when sodium metal reacts with sodium fluoroborate will be about 105 kecal per mole of sodium fluoroborate, or 1.5 kcal per gram of sodium | introduced. Introduction of one gram of sodium into 100 grams of the NaF-NaBF; mixture would be expected to raise its temperature by 43°C (109°F). The amount of heat involved in injecting sodium into sodium fluoroborate is almost the same as the amount involved in adding sodium to water (1..48 kcal per gram of sodium); thus the magnitude of this problem would be about the same but the location would be shifted from the steam generator 21 to the intermediate heat exchanger; the consequencés would not involve the liberation of hydrogen but would.involfie the addition of radiocactive sodium to the intermediate coolant. The consequences of injecting sodium fluoro- borate into the pfimary sodium stream would be similar to the reverse; the ‘removal of the boron might be difficult if it were produced in a finely divided form and dispersed throughout the coolant. Evaluation of Fluoroborates to Illustrate use of Molten Salts for IMFBR's As indicated in the introduction, the original plan for the assessment of molten salts as intermediate coolants for ILMFBR use was based on the as- sumption that fluoroborates might be the most worthy candidates. Conse- quently; an attempt was made to evaluate fluorbborates_from an engineering and design point of view. Although it now appears that carbonates may possibly prove to be more favorable than fluoroborates for ILMFBR use, the evalvation of fluoroborates is presented to illustrate the type of problems which are sure to be encountered and the design and engineering factors which will have to be taken into account in using any molten salt for the intermediate coolant of an IMFBR. Sultability of Coolant Salt for Operating Conditions of LMFBR Cycle and Engineering Design Changes Required for Its Use General -‘If all the components of an MSBR intermediate system using sodium fluoroborate as the circulated coolant were fully developed, this same secondary system would be applicable to an LMFBR power plant without significant changes in the operating conditions or without raising major new development problems in the fluid dynamics or heat transfer. | As will be explained below, substitution of the MSBR coolant salt for sodium as the secondary coolant for the IMFBR could narrow the freedom of choice of steam conditions and feedwater temperatures, and could affect thé circulation'rates in both the secondary and steam circuits, but these changes should not impose particular operating difficulties nor should they be detrimental to the overall plant thermal efficiency. 22 To judge the effects of using sodium fluoroborate rather than sodium, the performances of the two systems can be roughly compared.: For the pur- poses. of this survey, the properties of the two coolants were assumed to be as shown in Table 2. . The values for the salt will need later adjust- ment as more physical property data become available, but it is believed that the values shown are sufficiently reliable for some generalized conclusions to be drawn. Table 2. Physical Properties Assumed for this otudy Sodium _ Fluoroborate Composition ' . 'Na : NaBF) -NaF (92-8 mole %) Specific heat, Btu/lb-°F 0.% 0.%6 Volumetric heat capacity, 15.6 2.1 Btu/ft2-°F _ | Viscosity, lb/hr £t 0.6 6" at 800-900°F Density, w/et0 52 117 at 870°F Thermal conduct1v1ty, 41 0.2 Btu/hr-ft-°F | Liquidus temperature, °F 207 716 Vapor pressure, nm Hg 39 at 1160°F 270 at 1160°F Recent measurements of NaBF viscosity by Mound Laboratory gave values of 2.4 to 4.8 lb/hr ft in the temperature range of interest. The mixture with NaF should also be at least this low in viscosity. Even though there is uncertainty in the properties of the sodium fluoroborate, it is nevertheless clear that the differences between the properties of the salt and those of sodium are sufficient to cause design optimization‘stfidies to yield different numbers of.modules, arrangement of heat transfer surfaces, and circulation rates in both the secondary and steam Systems. A numerical éomparison of equipment costs and operating performances would therefore be dependent upon optimization studies which are beyond the scope of this survey. There are, however, some trends in the design aspects which can be briefly mentioned. 23 Operating Temperatures - The use of fluoroborate instead of sodium would not impose any limitation on the high temperature siQe of the sec- ondary system; this temperature would only be limited by thé allowable temperature in the primary system, which, in turn, might be set by the compatibility of the fuel element with the primary sodium. The properties of fluoroborate salt, however, affect the choice of the loWer_temperature erating part of the system. Figure 6 illustrates an assumed LMFBR flow- sheet based on sodiumAand Figure:T,illustrates an assumed MSBR secondary system.l6 The operating temperatures of these two flowsheets are not the same, but they are shown to illustrate the additional complexity which might be introduced by the use of salt. d A property of the salt which may impose lower temperature limitations is its liquidus temperature (716°F, 380°C). This is actually below the lowest sodium temperature shown in Figure 6 (725°F), but there would always be some risk of freezing salt in the tubes if the inlet water or inlet - steam were much below the liquidus temperature. TFigure 7 illustrates how this contingency could be avoided by degrading some of the 1000°F steam from the boiler-superheater to reheat steam in the preheater and to mix with entering water ahead of the pump; the entering watér would thereby be raised to 7O0°F, and the entering steam to 650°F. These temperatures are less than 70°F below the liquidus temperature, and it is believed that the inevitable temperature gradient across the tube wall would pre- vent the occurrence of temperatures low enough to freeze any salt. Accordingly, the cycle shown in Figure 7, perhaps with some reduction in the throttle temperature, could probably be used in an LMFBR plant, particularly since at least one preliminary study has indicated that super- critical pressure steam may be economically desirable for a high tempera- ture IMFBR in any event.>! | Additional study and testing will be required to decide just how low in temperature the inlet steam and water may be permitted to go without freezing salt in the tubes. p = psia 85 x 10° Ib/hr 535p, 1005°F ORNL~-DWG 68-6326 -;-—4i(::f' 1050°F PUMP REHEATER 397 Mwt 1{00°F ' PRIMARY HEAT EXCHANGER 2444 Mwt - 825°F A 725°F S 3515 p, 1000°F | 7x108 1b/hr _'_*f/\/\/\/\/—‘—‘_f'l Coolant. § SUPERHEATER 560 Mwt 922°F | |725°F | — - EVAPORATOR 1486 Mwt | b -| | ars°F . L : ' 633°F,558p 6.5 x 10% 1b/hr Fig. 6. Assumed LMFBR Secondary System Flowsheet Using Sodium as he ORNL—-DWG 68-6325 p=psia _ : 540 p, 1000°F, 1518.5h - = | | - h = ENTHALPY, Btu/ib . 5 1 x 106 1b/hr 3600 p, 1000°F, 1424.0h “250[: i-_- 10)(106 |b/hl’ I PUMP < Y - |1000°F REHEATER + REHEAT STEAM 293.5 Mwt | _ - PREHEATER — , 850°F r 100.5 Mwt 1300 °F PRIMARY HEAT | EXCHANGER 13235r1__j 600 p, 552°F 2225 Mwt —_ v e ] 570 p, 650°F 1256.7h - | {000°F - 10 x 10® Ib/hr - | SU?’(E):?LIEEA_TER _ | 3500p, 866°F .8 h 1931.5 Mwt 1307.8 75x10°% Ib/ hr Y I 850°F 850 °F 3800 p 700°F 3500 p, 551°F 769.2 h PUMP Fig. 7. Assumed MSBR Secondary System Flowsheet Using Salt as Coolant. ' 546.3 h ¢e 26 - Primary Heat Exchanger - Even though the heat transfer properties of sodium fluoroborate are generally considered good as compared with many heat transport fluids, they are not so good as sodium, and the effect of the change on the heat transfer coefficients and area requirements-could be marked. | | A primary exchanger designed for sodium-to-salt heat transfer could require about 50% more surface than a sodium-to-sodium unit, aséuming that about the same velocity of 10 ft/éec were used inside the tubes in each case. To achieve essentially the same velocity requires a different de- sign, however, in that the volumetric flow requirement of the salt is only about 37% of that needed for sodium for the same heat transport capacity. In order to maintain the velocity with a smaller volumetric flow rate, fewer tubes or ones of a smaller diameter must be used, either of which would increase the pressure drop through the exchanger.. For example, if l/2-in. 0D tubes were used instead of the B/M-in. tubes used in the sodium system, the salt velocity would be about 10 ft/sec compared to 11 ft/sec for the sodium in the larger tubes, but the loss of head due to flow of the'salt would be about 2.7 times that for sodium. The above-mentioned effects of using sodium fluoroborate rather than sodium in thé IMFBR primary heat exchanger would not be as pronounced if enhanced heaf-transfer tubing were used. When using helically grooved type tubing, for example, the heat transfer film coefficient on the inside might be approximately doubled (enhanced heat transfer has been demonstrated with water but not yet with salt for such tubing), although for a given length of tubing the Ap would also be about doubled. If this type of tubing were used, the primary heat exchanger surface area requirements might be about the same in both the sodium and salt systems. 1In this case the pres- sure drop of the salt system would be more than twice as great as that in the sodium loop. This situation represents a logical application of en- hanced type of surfaces and would merit serious study and development. If the primary heat exchanger is made appreciably larger due'to the effects mentioned above,. there would be some increase in the inventory of primary sodium. 27 Design of a sodium-to-fluoroborate IMFBR primary heat exchanger pro- bably would require that surfaces in contact with the fluoroborate be fabricated of Hastelloy-N, some other high-nickel alloy, or nickel metal. - The material cost of Hastelloy-N could be about fwice as much as that of the stainless steel used in sodium-to-sodium exchangers, although labor and other fabrication costs would not be greatly different. If, as sug- gested above, the surface requirements for the sodium-to-salt exchanger are also significantly greater, then the cost difference between the two types of units would be accentuated. It must be emphasized that all the above comments regarding the rela- tive performances of the primary heat exchangers are based on assumed pro- perties for the sodium fluoroborate salt. The viscosity of the sodium fluoroborate over the temperature range of interest is particularly un- certain and could have a major effect on the estimates. As noted above, our most recent information suggests that the viscosity of pure sodium fluoroborate may be between 1 and 2 centipoise (2.4 to 4.8 1b/hr-ft) in the temperature range of interest. Additional study will be required to confirm these results and to extend them to cover mixtures of sodium fluoride with sodium fluoroborate. _ The overall result of the above-mentioned factors is that if salt rather than sodium is used as the secondary coolant, the cost of the primary salt exchangers could be greater, both through the need for more surface and from more expensive type of tubing and tubing material. Pri- mary sodium inventory could be increased, and cell dimensions may neéd to be expanded to accommodate larger units. The pumping power requirements, but not the volumetric flow rate, might also be increased. Steam Generator and Reheater - In the evaporator portion of a sub-. critical pressure steam generator the outside film coefficient of heat transfer when fising sodium fluoroborate might be only about one-eighth of that which could be obtained by using sodium. The overall coefficient of heat transfer could be in the order of 60% of that obtained in a sodium- to-water evaporator. While not investigated in this survey, it can be presumed that the same difference in performance would exist if the steam pressure were supercritical. 28 In the superheating.regioh of the steam generator the outside film coefficient for sodium fluoroborate might be about one-seventh that in a sodium-to-steam exchanger, and the overall heat transfer coefficient again about 60% of that in a sodium exchanger. In the reheater, the salt film coefficient might be only about‘one- eighth the sodium film coefficient in a sodium-to-steam unit. In this case the steam-side film coefficient should be controlling and the effect on the overall heat transfer coefficient would not be as great. It was estimated that the overall coefficient of the salt unit would be about T5% of that for sodium. In this case there would be little or no incentive for use of enhanced type of heat transfer tubing. The temperature difference between the incoming feedwater to the evaporator section of the steam generator and the salt or sodiufi tempera- ture on the shell side is important in the design of the equipment, from the standpoint of the thermal stress induced by the thermal gradient across the tube wall. This difference must also be controlled in order to avoid -freezing the coolant salt but the allowable value has not been established. This problem has beefi briefly discussed above. The situation is hot unique to the salt coolant, however, in that, as may be noted in the sodium system shown in Figure 6, the evaporator is supplied with 478°F feedwater and has an exit sodium temperature of 725°F, suggesting either that this particular set of conditions will not be satisfactory from a thermal stress stand- point in the present IMFBR design or that our MSBR designs are too | conservative. As with the primary heat exchangers, if it is determined that Hastelloy N is required for compatibility with the sodium fluoroborate, the manu- facturing cost of the steam generators and reheaters could be significantly greater, both from the materials cost and the requirement for somewhat more heat transfer surface. On the other hand, there could be substantial sav- ings in the complexity and cost of the units‘thfough elimination of the speclal arrangements to accommodaté a sodium-water reaction. While some pressure-relief system is needed on the intermediate circulating system in any event to prevent pressures in the steam system from being trans- mitted to the primary heat exchanger, fhis system might be less elaborate for the salt loop since no chemical release of energy would be involved 29 and it would not be necessary to dispose of hydrogen gas. These cost effects are offsetting, and it is possible that the overall difference in equipment costs between the two coolant choices are not so important as other considerations in comparing the two. Piping - The volumetric flow rate of the salt need be onlyv57% of that of sodium to obtain the same heat transport capacity. If the same pipe sizes were used in both systems, the salt velocity would be corre- spondingly lower and the pumping effort would perhaps be about 439 that in a sodium system. If the salt system were designed for about the same velocity as uséd in sodium piping, typically a 15-in. pipe would be used instead of a 24-in. sodium pipe, but in this case the loss of head due to friction in the piping might be about 1.3 times as great. The optimum design for a salt system would likely fall somewhere in between these extremes. These effects are probably not of great éonsequence in com- paring the two coolants. Circulating-Pump Power Requirements - If optimization studies were made of an IMFBR system using sodium fluoroborate in the intermediate system, it would probably be determined that the velocities of the salt should be as high or higher than those of sodium. If this is true, the total head requirements on the pump could be twice as great for the salt system. Even though the mass flow rate of the salt need be only 80% of that required for sodium, the difference in the pumping effort is sub- stantial. While this is undesirable, it is probably not of over-riding importance in considering the suitability of the salt as a coolant. Heat Systems - Both the sodium and fluoroborate intermediate cooling systems would require provisions for maintaining the coolant above the freezing point. The melting temperature of about 200°F (93°C) for sodium would be easier to maintain by electrically heated tracers than the 716°F (380°C) needed for salt systems, but the difference is one of the amount of heat rather than the presence or absence of a tracer heat system. The heat loss from either system would be about the same. (In this connection, a heated cell has been considered in some MSBR design studies in contrast to the use of tracers on piping and equipment.) 50 Compatibility with LMFBR Materials Including Effects of Radiation The tests now in progress suggest that fluoroborates are compatible with Hastelloy—N.* If Hastelloy-N were chosen as the material for the secondary coolant circuit of an IMFBR, two problems would be presented: it is compatible with steam and is it compatible with sodium? With re- spect to the first, compatibility with steam, a test program is underway in which ORNL and TVA are cooperating to test Hastelloy-N with supercritical steam in the Bull Run Steam Plant. One advantage of Hastelloy-N for steam service is that the alloy is not susceptible to stress-corrosion cracking under conditions where serious problems have resulted in stainless steel. With respect to compatibility with sodium,.extensive studies were made in the early stages of the molten salt development program. The results were summarized in the MSBR Status Report of 1958 (Ref. 18) as follows: "The effect of sodium on the structural materials of interest has also been extensively -studied, since sodium is proposed for use as the intermediate heat transfer medium. Corrosion problems inherent in the utilization of sodium for heat transfer purposes do not involve so much the deterioration of the metal surfaces as the tendency for components of the container material to be transported from hot to cold regions and to form plugs of deposited material in the cold region. As -in the case of the cor- rosion by the salt mixture, the mass transfer in sodium- containing systems is extremely dependent on the maximum system operating temperature. The results of numerous tests indicate that the nickel-base alloys such as Inconel and INOR-8, are satisfactory containers for sodium at temperatures below 1300°F and that above 1300°F the aus- tenitic stainless steels are preferable.'” * Recent experiments indicate that the specific corrosion rate depends upon the water content of the salt. ' For example, the corrosion rate in a ‘thermal convection loop at a peak temperature of 1125°F (607°C) was about 0.35 mpy when the salt contained LOO ppm water and 0.75 mpy when 1000 ppm water was present. Although these corrosion rates are not very high, they do indicate that a process will have to be developed for on-stream re- moval of moisture from the sodium fluoroborate coolant circuit. This process, which will likely consist of bubbling HF or BF3 through a side ‘stream of the salt, is currently under development. 31 This early assessment of nickel-base alloy performance in sodium has now 19 been confirmed in recent British = and Russian20 studies. Also Atomics International and ORNL have completed extensive tests of Hastelloy-N in NaK9Na-K, 30-70 wt %) extending to lower temperature which show that mass transfer rate of the alloy at 350°F is only slightly greater than that of type 316 stainless steel.21 Thus, since the secondary coolant temper- atures of the IMFBR are limited to a maximum of about 1100°F, it is likely that Inconel or Hastelloy-N would be compatible with sodium. The effects of radiation in -an IMFBR are likely to be much less serious than those in a molten salt reactor, since the fluoroborate would be exposed only to the gamma flux from the radiocactive sodium in the pri- mary circuit. ‘Experiments at ORNL have exposed the eutectic NaBFu-NaF mixture at 600°C to the gamma radiation from decaying HFIR fuel elements to a total dose of 7.7 x 10707 with an average intensity of 0.15 w/g and a maximum intensity of 0.5 w/g. (The average intensity of absorbed gamma radiation estimated for the MSBR heat exchangers is around 0.25 w/g). The salt was contained in a Hastelloy-N capsule. Ixamination following the irradiation revealed no evidence‘of radiation decomposition of fluoroborate and no incompatibility with the Hastelloy-N.22 Effects of Safety and Accident Conditions As noted earlier, sodium is expected to react quite exothermically with fluoroborates to producé boron, sodium fluoride, and approximately 1.5 kilocalories per gram of sodium. No gaseous products of the reaction are foreseen, but the liberated heat could cause the vapor pressure of BF5 to rise locally, the amount depending on the ratio of sodium to fluoroborate and the rate of'the reaction. No experimental studies have yet been made of this reaction. | A leak of fluoroborate from the intermediate coolant system to the ocoutisde would release some BF3, a toxic gas which is essentially equivalent to HF in its properties. As the salt cooled it would solidify, trapping any contained radiocactive impurities. The fluoroborate salts are very soluble in water and can easily be dissolved and flushed away from the area of a spill. 52 A leak of steam into the fluoroborate would cause some hydrolysis of the material with the release of HF. This reaction should be reversible by side-stream treatment with BF, or HF-BFE; studies of this reaction will be part of the MSBE development grogram. . A leak of fluoroborate into the steam generator is unlikely because of the pressure differential which is expected. However, if one occurred, the solubility of the fluoroborate would permit its easy removal. The effect of fluoride on corrosion of the steam generator has not yet been assessed; fluorides do contribute to stress corrosion cracking, but their effect is much less serious than that of chlorides. Leakage of fluoroborate into the primary system of an IMFBR would be expected to give rise to circulating, elemental boron which should result in detectable changes in reactivity. Boron is said not to dissolve .in boiling sodium (b.p. 880°C), but the reference is quite 01d.%’ Tt would - be important to study the production of boron and its distribution within the primary system, and to develop techniques for preventing the reaction or for the removal of the boron. (The difference in density, 2.3 for crystalline or l.75 for amorphous boron, vs 0.97 for sodium might facili- tate separation or removal.) Availability and Cost of Fluoroborate Fluoroborate production technology is sufficiently well-developed that incentives to assure its availability are not needed. Several producers of industrial chemicals can now supply sufficient quantities to fill foresee- able needs of the U.S. breeder reactor program. Sodium fluoroborate of high qfiality is currently available at low prices as a consequence of its use in industrial electroplating processes. We have recently obtained 2400 1b of NaBFu, of greater than 99% purity at a cost of $0.50 per pound. That cost included a $0.05 per pound charge for a minor process modifica- tion which was introduced in order to obtain a product which contained <200 ppm oxygen. Chemical analysis of the material showed that it ex- ceeded the following specifications: 55 NaBF) 99.08% 0, 0.025% Pb 0.004% Si 0.01% Ca 0.01% Fe 0.023%% HQO insol. <0.01% H,0 0.01% The compound was synthesized from borax and hydrogen fluoride as starting materials. As anhydrous technical grade chemicals, these materials cost $88 per ton and $0.18 per pound, respéctively. Since these costs appear to be the principal variables in the final cost of the coolant cost, it is anticipated that the cost of coolant salt relative to overall reactor costs will remain very low and essentially constant in the future. Development Requirements for IMFBR use on Molten Salts Cycle Choice Additional study and development would be required to select the best steam cycle. Factors favoring the supercritical cycle include the higher thermal efficieficy which may be achieved, the raising of the feed- water temperature which lowers the risk of freezing salt and reduces the thermal stresses on the tube wall, and the possibility that the steam generator could be designed with more confidence in the prediction of the internal operating conditions. Development Program for the MSBR The work planned for the MSBR illustrates the type of study which would be necessary for the LMFBR. It includes study of the following items as noted. Steam Generator Tests - To obtain information necessary to design equipment for generating superheated steam at subcritical and super- critical pressures and for reheating steam by use of molten salt coolants. 34 Heat Transfer Enhancement - To study the effects of changes in the tube shapes as a means for improving the efficiency of heat transfer and the predictability of heat transfer as well as lowering the capital cost and salt inventory. Pressure Relief System - To determine how much simplification can be achieved as a result of the absence of the possibility of a sodium-steam reaction and the necessity of disposing of hydrogen. Fundamental Molten Salt Heat Transfer - To improve and extend our knowledge of the technology, particularly with respect to engineering factors such as transfer coefficients, pressure drop, effects of possible corrosion and scale deposits on .the tubes, and effects of wetting and interfacial deposits. Chemical Factors - Including choice of compositiorn for minimum liquidus temperature, procurement or purification of commercial material, study of steam and sodium reactions with fluoroborates, study of on-site repurification by batch or side-stream stream, and more precise determina- tion of all relevant physical properties. Evaluation of Carbonates for ILMFBR Use -Msny of the Questions which would need attention are”the same as ‘those now under study as part of the ORNL development of fluoroborates. The outstanding needs seem to be for an evaluation of the heat transfer and physical properties of carbonate mixtures, for a more accurate deter- mination of the liquidus temperature in the three component system, lithium carbonate-sodium carbonate-potassium carbonate, which is of basic interest, and for study of the possibility of further lowering the liquidus temperatures by the addition of fourth components to the system. Of sec- ondary importance would appear to be such chemical factors as purification, results of mixing with sodium or steam, reactions with container materials, and effects of gamma radiation. . -t l. lO. 11. 12. 13. 1L, 15. 16. 17. ~Salt Mixture NaNO,‘NaNO,-KNO 55 References R. E. Thoma, Phase Diagrams of Nuclear Reactor Materials, USAEC Report ORNL-2548, November 6, 1959, p. 3k. Loc. cit., p. L2, MSRP Quarterly Progreés Report January 31, 1958, USAEC Report ORNL-2474, pp. 90-91. R. F. Newton, Thermodynamic Stability of Metals and Their Chlorides in the Presence of UCl5 and PuClB, personal communication, October 31, 1965 L H. F. Bauman and E. S. Bettis, private communication. G. J. Janz and M. R. Lorenz, J. Chem. Eng. Data 6(3): 321-23 (1961). Rolin, Maurice and Recapet, The Thermodynamic Properties of the Alkali Metal Carbonates, I. The Ternary Diagram NaECOB-KéCOB-LiQCOB, Bull. Soc. Chim France, 1964, No. 9, p. 210L. G. J. Janz and Saegusa, Molten Carbonates as Electrolytes: Viscosity and Transport Properties, J. Electrochem. Soc., pp. 452-56 (May 1963). E. S. Bettis, private communication. MSRP Semiannual Progress Report August 31, 1962, USAEC Report ORNL-3369, pp. 135-36. S. Cantor, personal communication. W. E., Kirst, W. M. Nagle, and J. B. Castner, Trans. Am. Inst. Chemn. Eng., 36, 371 (1940). H. W. Hoffman, S. I. Cohen, Fused Salt Heat Transfer - Part III: Forced-Convection Heat Transfer in Circular Tubes Containing the 5 3 37 USAEC Report ORNL-243%, March 1, 1960. M. W. Rosenthal, H. F. Poppendiek, and R. M. Burnett, A Method for Evaluating the Heat Transfer Effectiveness of Reactor Coolants, ORNL Report ORNL-CF-54-11-63, November 4, 195k. H. S. Booth and D. R, Martin, Boron Trifluoride and Its Derivatives, John Wiley and Sons, Inc., New York, 1949, pp. 31. P. R. Kasten, E. S. Bettis, and R. C. Robertson, Design Studies of a 1000 Mw(e) Molten-Salt Breeder Reactor, USAEC Report ORNL-3996, August 1966. H. B. Holz et al., 1000 Mwe FBR Parametric Study of Secondary Sodium-Steam System, NAA-SR-12060 (August 1966). 18. 19. 20. 21. 22. 25. 36 MSRP Status Report, ORNL-263k4 (November 12, 1958) pp. 119-123. A. M. Thorley and C. Tyzack, Alkali Metal Coolants, IAEA; 97 (1967). B. A. Nevsorov et al., Procl Interim. Conf. Peaceful Uses Atomlc Energy, Third Geneva, l96h 9 (1965) pp. 561-69, H. W. Savage et et al., SNAP-8 Corrosion Program Summary Report, USAEC Report ORNL-3898 (1965). E. L. Compere, H. C. Savage, and J. M. Baker, Gamma Irradiation of Fluoroborate, MSRP Semiannual Report for Period Ending Aug. 31, 1968, USAEC Report ORNL-L3LkL. H. Moissan, Compt. Rend. 11k, 319 (1892). " “ 131. 132, 133. 13k, 135. 136, O @M= O\ =W PHOSTNHgENYNQEHUOEZ O AQ G - F. E. Se S. G. J. E. B. . B. NMEWEEGIENYS2HOQIRE YR E Y G H L. Anderson Baes Beall Bender Bettis Billington Bohlmann Borkowski Boyd Briggs Cantor Cottrell Crowley Culler Ditto Eatherly . Ferguson Ferris Fraas Frye Furlong Grimes . Grindell Harms . Helms Haubenreich Hoffman . Huntley . Jordan . Kasten Kedl . Kelley . Koger Korsmeyer . Kress 37 Internal Distribution 560 5T 38. 39. Lo. L1, 50. 51-100. 101. 102. 103. 104, 105. 106. 107. 108. 109. 110. 111. 112. 113. 114, 115. 116-117. 118-119. 120-129. 130. ORNL=-TM-2696 J. A. Lane A, P, Litman M. I. Lundin H. G. MacPherson R. E. MacPherson H. BE. McCoy H. F. McDuffie H. A. McLain L. E. McNeese J. R. McWherter A. 5. Meyer R. L. Moore E. L. Nicholson A. M. Perry R. C. Robertson M. W. Rosenthal J. Roth Dunlap Scott M. J. Skinner A. N. Smith I. Spiewak J. R. Tallackson R. E. Thoma D. B. Trauvger G. M. Watson A. M. Weinberg J. R. Weir M. E. Whatley J. C. White L. V. Wilson Gale Young Central Research Library Document Reference Section Laboratory Records (LRD) Laboratory Records (LRD-~RC) External Distribution F. W. Albaugh, Pacific Northwest Laboratory, 703 Building Richland, Washington 99352 A. Amorosi, IMFBR Program Office, Argonne National Laboratory, 9700 S. Cass Avenue, Argonne, Illinois 60439 Arthur Bunke, Byron Jackson Pump Company, P. 0. Box 2017, Terminal Annex, Los Angeles, California 90054 Karl Cohen, Westinghouse Electric Co., P. 0. Box 158, Madison, Pennsylvania 15663 D. F. Cope, USAEC Site Representative, ORNL E. L. Daman, Foster Wheeler Co., 110 S. Orange, Livingston, Nl YI‘ 07039 137. 138. 139. 140. 141, 142, 143, Ll 145, 146. 147. 148. 149. 150-151., 152, 153, 15k, 155. 156. 157« 158. 159.. 160. 161. 162, 163, 164, - 165. 38 R. W. Dickinson, Liquid Metals Engineering Center, P. O. Box 1449, Canoga Park, California 91304 Joseph Fltzgerald Cambrldge Nuclear Corp., 131 Portland Street, Cambridge, Mass. 2139 A. Giambusso, DRDT U. 5. Atamic Energy Commission, Washington, D. C. 20545 _ Nicholas Grossman, DRDT, U. S. Atomic Energy Commission, Washington, D. C. 20545 D. B. Hall, LASL, P. O. Box 1663, Los Alamos, N. M. 8754k R. E. Hosklns, TVA, 503 Power Building, Chattanooga, Tenn. 37401 G. M. Kavanagh, USAEC Washington, D. C. 20545 . J. C. R. Kelly, Jr., Westlnghouse Electric Corporation, Atomic: Power Division, P. O. Box 158, Madison, Pa. 15663 B. W. Kinyon, Combustion Engineering, 911 E. Main Street, Chattanooga, Tenn. 37401 Stephen Lawroski, Argonne National Laboratory, 9700 S. Cass, Argonne, Illinois 60439 _ George Lewis, Charles T. Main Co., 441 Stuart, Boston, Mass. 02116 C. Rogers McCullough, Southern Nuclear Engineers, 5401 Westbard Ave., Bethesda, Maryland 20016 ' W. B. McDonald, Pacific Northwest Laboratory, 705 Building, 700 Area, Richland, Washington 99352 T. W, McIntosh DRDT, U. S. Atomic Energy Commission, Washlngton, D. C. 20545 ‘ ‘ F. T. Miles, Brookhaven National Laboratory, Associates Universities, Inc., Upton3 L. I., New York 11973 A. J. Pressesky, DRDT, U. S. Atomic Energy Commission, Washington, D. C. 20545 - P. B. Probert, Babcock and Wllcox, Power Generation D1v181on, 20 S. Van Buren Avenue, Barberton, OChio 44203 C. J. Raseman, Brookhaven National Laboratory, Upton, L. I., New York 119753 Leonard F. C. Reichle, Ebasco Services, Inc., 2 Rector Street, New York, N. Y. 10006 ' B. T. Resnlck DRDT, U. S. Atomic Energy Commission, Washlngton, D. C. 20545 M. A. Rosen, DRDT, U. S. Atomic Energy Commission, Washlngton D. C. 20545 F. A. Ross, DRDT, U. S. Atomic Energy Commission, Washington, D. C. 20545 | M. Shaw, Director, DRDT, U. S. Atomic Energy Commission, Washlngton, D. C. 205&5 Sidney Siegel, Atomics International, P. 0. Box 309, Canoga Park, California 9310k E. BE. Sinclair, DRDT, U. S. Atomic Energy Commission, Washington, D. C. 20545 Bernard Singer, DRDT, U. S. Atomlc Energy Comm1851on Washington, D. C. 20545 Earl O. Smith, Black and Veatch, P. O. Box 8L05, Kansas City, Missour 6hllh A. N, Tardlff DRDT, U. S. Atomic Energy Commission, Washington, D. C. 20545 166. 167. 168. 169. 170. 171. 172. 173-187. 188. 189-308. 59 Barry L. Tarmy, Esso Research and Engineering Co., Esso Engineering Technology Department, P. 0. Box 101, Florham Park, N. J. 07932 C. H. Waugaman, Power Research Staff, Tennessee Valley Authority, 30% Power Building, Chattanooga, Tennessee 37401 Ralph A. Webb, Babcock and Wilcox, P. 0. Box 1260, Lynchburg, Virginia 2&505 G. W. Wensch, DRDT, U. S. Atomic Energy Comm1851on, Washington, D. C. 205&5 M. J. Whitman, DRDT, U. S. Atomic Energy Commission, Washington, D. C. 20545 Warren Winsche, Brookhaven National Laboratory, Associated Universities, Inc., Upton, L. I., New York 11973 Walter H. Zinn, Combustion Engineering, Nuclear Division, P. O. Box 500, Windsor, Conn. 06095 Division of Technical Information Extension (DIIE) Laboratory and University Division (ORO) FREP Distribution.