Part 11 MOLTEN-SALT REACTORS 11. 12. 13. 14. 15. 16. 17. H. G. MacPrERsoN, Editor Oak Ridge N atvonal Laboratory Introduction Chemical Aspects of Molten-Fluoride-Salt Reactor Fuels Construction Materials for Molten-Salt Reactors Nuclear Aspects of Molten-Salt Reactors Equipment for Molten-Salt Reactor Heat-Transfer Systems Aircraft Reactor Experiment Conceptual Design of a Power Reactor CONTRIBUTORS L. G. ALEXANDER H. G. MAcPHERSON J. W. ALLEN W. D. MANLY E. S. BerTIs L. A. MANN F. F. BLANKENSHIP W. B. McDoNALD W. F. BoubprEAU H. J. METZ E. J. BREEDING P. PATRIARCA W. G. CoBB H. F. PoPPENDIEK W. H. Cooxk J. T. RoBERTS D. R. CuNEko M. T. RoBINSON J. H. DEVANN T. K. RocHE D. A. DougLas H. W. SAVAGE W. K. ERGEN G. M. SLAUGHTER W. R. GRIMES E. StorTO H. INoUYE A. TaBoADA D. H. JANSEN G. M. TorsoN G. W. KeLaOoLTZ F. C. VONDERLAGE B. W. Kinvon - G. D. WHITMAN M. E. LackEy J. ZASLER PAUL ERSKINE BROWN PREFACE The Oak Ridge National Laboratory, under the sponsorship of the U. S. Atomic Energy Commission, has engaged in research on molten salts as materials for use in high-temperature reactors for a number of years. The technology developed by this work was incorporated in the Aircraft Reactor Experiment and made available for purposes of civilian apphcatlon This earlier technology and the new information found in the civilian power reactor effort is summarized in this part. So many present and former members of the Laboratory staff have contributed directly or indirectly to the molten salt work that it should be regarded as a contribution from the entire Laboratory. The technical direction of the work was provided by A. M. Weinberg, R. C. Briant, W. H. Jordan, and S. J. Cromer. In addition to the contributors listed for the various chapters the editor would like to acknowledge the efforts of the following people who are currently engaged in the work reported: R. G. Affel, J. C. Amos, C. J. Barton, C. C. Beusman, W. E. Browning, S. Cantor, D O. Campbell G. L. Cathers B. H. Clampltt J. A. Conlin, M. H. Cooper, J. L. Crowley, J. Y. Estabrook, H. A. Friedman, P. A. Gnadt, A. G. Grindell, H. W. Hoffman, H. Insley, S. Langer, R. E. Mac- Pherson R. E. Moore, G. J. Nessle, R. F. Newton, W. R. Osborn, F. E. Romie, C. F. Sales, J. H. Shaffer, G. P. Smith, N. V. Smith, P. G. Smith, W. L. Snapp, W. K. Stair, R. A. Strehlow, C. D. Susano, R. E. Thoma, D. B. Trauger, J. J. Tudor, W. T. Ward, G. M. Watson, J. C. White, and H. C. Young. The technical reviews at Argonne National Laboratory and Westing- house Electric Corporation aided in achieving clarity. The editor and contributors of this part wish to express their apprecia- tion to A. W. Savolainen for her assistance in preparing the text in its final form. Oak Ridge, Tennessee H. G. MacPherson, Editor June 1958 CHAPTER 11 INTRODUCTION* The potential utility of a fluid-fueled reactor that can operate at a high temperature but with a low-pressure system has been recognized for a long time. Some years ago, R. C. Briant of the Oak Ridge National Lab- oratory suggested the use of the molten mixture of UF4 and ThF4, together with the fluorides of the alkali metals and beryllium or zirconium, as the fluid fuel. Laboratory work with such mixtures led to the operation, in 1954, of an experimental reactor, which was designated the Aircraft Reactor Experiment (ARE). Fluoride-salt mixtures suitable for use in power reactors have melting points in the temperature range 850 to 950°F and are sufficiently compatible with certain nickel-base alloys to assure long life for reactor components at temperatures up to 1300°F. Thus the natural, optimum operating tem- perature for a molten-salt-fueled reactor is such that the molten salt is a suitable heat source for a modern steam power plant. The principal advantages of the molten-salt system, other than high temperature, in comparison with one or more of the other fluid-fuel systems are (1) low- pressure operation, (2) stability of the liquid under radiation, (3) high solubility of uranium and thorium (as fluorides) in molten-salt mixtures, and (4) resistance to corrosion of the structural materials that does not depend on oxide or other film formation. The molten-salt system has the usual benefits attributed to fluid-fuel systems. The principal advantages over solid-fuel-element systems are (1) a high negative temperature coefficient of reactivity, (2) a lack of radia- tion damage that can limit fuel burnup, (3) the possibility of continuous fission-product removal, (4) the avoidance of the expense of fabricating new fuel elements, and (5) the possibility of adding makeup fuel as needed, which precludes the need for providing excess reactivity. The high negative temperature coeflicient and the lack of excess reactivity make possible a reactor, without control rods, which automatically adjusts its power in re- sponse to changes of the electrical load. The lack of excess reactivity also leads to a reactor that is not endangered by nuclear power excursions. One of the attractive features of the molten-salt system is the variety of reactor types that can be considered to cover a range of applications. The present state of the technology suggests that homogeneous reactors which use a molten salt composed of Bel's and either Li’F or NaF, with UF4 for fuel and ThF4 for a fertile material, are most suitable for early construction. *By H. G. MacPherson. 567 968 INTRODUCTION [cHAP. 11 These reactors can be either one or two region and, depending on the size of the reactor core and the thorium fluoride concentration, can cover a wide range of fuel inventories, breeding ratios, and fuel reprocessing sched- ules. The chief virtues of this class of molten-salt reactor are that the design is based on a well-developed technology and that the use of a simple fuel cycle contributes to reduced costs. With further development, the same base salt, that is, the mixture of BeF2 and Li’F, can be combined with a graphite moderator in a hetero- geneous arrangement to provide a self-contained Th-U233 system with a breeding ratio of one. The chief advantage of the molten-salt system over other liquid systems in pursuing this objective is that it is the only system in which a soluble thorium compound can be used, and thus the problem of slurry handling is avoided. The possibility of placing thorium in the core obviates the necessity of using graphite as a core-shell material. Plutonium is being investigated as an alternate fuel for the molten-salt reactor. Although it is too early to describe a plutonium-fueled reactor in detail, it is highly probable that a suitable PuFs-fueled reactor can be constructed and operated. The high melting temperature of the fluoride salts is the principal dif- ficulty in their use. Steps must be taken to preheat equipment and to keep the equipment above the melting point of the salt at all times. In addition, there is more parasitic neutron capture in the salts of the molten-salt reactor than there is in the heavy water of the heavy-water-moderated reactors, and thus the breeding ratios are lower. The poorer moderating ability of the salts requires larger critical masses for molten-salt reactors than for the aqueous systems. Finally, the molten-salt reactor shares with all fluid-fuel reactors the problems of certain containment of the fuel, the reliability of components, and the necessity for techniques of making repairs remotely. The low pressure of the molten-salt fuel system should be beneficial with regard to these engineering problems, but to evaluate them properly will require operating experience with experimental reactors. CHAPTER 12 CHEMICAL ASPECTS OF MOLTEN-FLUORIDE-SALT REACTOR FUELS* The search for a liquid for use at high temperatures and low pressures in a fluid-fueled reactor led to the choice of either fluorides or chlorides because of the requirements of radiation stability and solubility of appre- ciable quantities of uranium and thorium. The chlorides (based on the CI37 isotope) are most suitable for fast reactor use, but the low thermal-neutron absorption cross section of fluorine makes the fluorides a uniquely desirable choice for a high-temperature fluid-fueled reactor in the thermal or epi- thermal neutron region. Since for most molten-salt reactors considered to date the required con- centrations of UF4 and ThF4 have been moderately low, the molten-salt mixtures can be considered, to a first approximation, as base or solvent salt mixtures, to which the fissionable or fertile fluorides are added. For the fuel, the relatively small amounts of UF4 required make the correspond- ing binary or ternary mixtures of the diluents nearly controlling with regard to physical properties such as the melting point. | 12-1. CHOICE OF BASE OR SOLVENT SALTS The temperature dependence of the corrosion of nickel-base alloys by fluoride salts is described in Chapter 13. From the data given there, 1300°F (704°C) is taken as an upper limit for the molten-salt-to-metal interface temperature. To provide some leeway for radiation heating of the metal walls and to provide a safety margin, the maximum bulk temperature of the molten-salt fuel at the design condition will probably not exceed 1225°F. In a circulating-fuel reactor, in which heat is extracted from the fuel in an external heat exchanger, the temperature difference between the inlet and outlet of the reactor will be at least 100°F. The provision of a margin of safety of 100°F between minimum operating temperature and melting point makes salts with melting points above 1025°F of little interest at present, and therefore this discussion is limited largely to salt mixtures having melting points no higher than 1022°F (550°C). One of the basic features desired in the molten-salt reactor is a low pressure in the fuel system, so only fluorides with a low vapor pressure at the peak operating temperature (~700°C) are considered. *By W. R. Grimes, D. R. Cuneo, F. F. Blankenship, G. W. Keilholtz, H. F. Poppendiek, and M. T. Robinson. 569 570 CHEMICAL ASPECTS: MOLTEN-FLUORIDE-SALT FUELS [cHAP. 12 710°C KF —— LiF 454°c” 492°C F1c. 12-1. The system LiF-NaF-KF [A. G. Bergman and E. P. Dergunov, Compt. rend. acad. sci. U.R.S.S., 31, 7564 (1941)]. Of the pure fluorides of molten-salt reactor interest, only BeFs meets the melting-point requirement, and it is too viscous for use in the pure state. Thus only mixtures of two or more fluoride salts provide useful melting points and physical properties. The alkali-metal fluorides and the fluorides of beryllium and zirconium have been given the most serious attention for reactor use. Lead and bis- muth fluorides, which might otherwise be useful because of their low neutron absorption, have been eliminated because they are readily reduced to the metallic state by structural metals such as iron and chromium. Binary mixtures of alkali fluorides that have sufficiently low melting points are an equimolar mixture of KF and LiF, which has a melting point of 490°C, and a mixture of 60 mole 9, RbF with 40 mole 9, LiF, which has a melting point of 470°C. Up to 10 mole 9, UF4 can be added to these alkali fluoride systems without increasing the melting point above the 550°C limit. A melting-point diagram for the ternary system LiF-NaF-KF, Fig. 12-1, indicates a eutectic with a lower melting point than the melting points of the simple binary LiF-KF system. This eutectic has interesting properties as a heat-transfer fluid for molten-salt reactor systems, and data on its physical properties are given in Tables 12-1 and 12-2. The KF-LiF and RbF-LiF binaries and their ternary systems with NaF are the only available systems of the alkali-metal fluorides alone which have 571 CHOICE OF BASE OR SOLVENT SALTS 12-1] 8L 8ELy | 88€0°0 650 ¥ %% 8¢¢ (88-L2-5€) S Jog-ABN-ArT ¢L'¥ 0LT% | 00%0°0 ¢% 0 gL eS'% PS¥ (TS T1-G9%) AS-AN-A'] 78 80T | 60L0°0 8% 0 6 6L°¢ 01¢ (05-09) 7 A1Z-A8N 82l P91S | 9%60°0 2s'0 L8 L3¢ 09€ (6%-L9) g -A8N 2% pL19 | 68100 £9°0 0% 9% ' 0S¢ (05-0¢) sqed-Arl ) 709¢ 8I1°0 G9'0 0¥ 913 ¢0S (16-69) S qod-ArT ¢—0T X q v q v 06009 3V A Do 0/, ajoW GLoL/g?V = U D,00L 18 (D)L —V =9 yutod ‘worpsoduron) Aoeded 1eoy 00/8 SuneN o astodrjuan ‘AJIS00SIA ‘Aq1susp prnbr SHATIONT] NALIOJA TVOIdAJ, 40 XLISOOSIA ANV ALISN@(] ¥Od SNOILVADY ANV ‘SHLLIOVAV)) LVH] ‘SINIOJ ONILTAA 1-¢1 4714V ], 572 CHEMICAL ASPECTS: MOLTEN-FLUORIDE-SALT FUELS [CHAP. 12 Temperature, °C 2NaF-ZrFy e b ™ (T O Z ™ NaF 10 20 30 40 50 60 70 80 90 ZrFy ZrF4 ,mole % F1g. 12-2. The system NaF-ZrFy,. low melting points at low uranium concentrations. They would have utility as special purpose reactor fuel solvents if no mixtures with better properties were available. TABLE 12-2 THERMAL CoNDUCTIVITY OF TYPICAL FLUORIDE MIXTURES Thermal conductivity, Composition, Btu/ (hr) (ft) (°F) mole 9, Solid Liquid LiF-NaF-KF (46.5-11.5-42) 2.7 2.6 NaF-BeF2 (57-43) 2.4 Mixtures with melting points in the range of interest may be obtained over relatively wide limits of concentration if ZrF4 or BeF; is a component of the system. Phase relationships in the NaF-ZrF4 system are shown in Fig. 12-2. There is a broad region of low-melting-point compositions that have between 40 and 55 mole 9, ZrF 4. 12-1] °C - Temperature 700 (2 o @ 600 2 o @ E 50 8 0 800 400 300 200 L LiF CHOICE OF BASE OR Fia. 12-3. SOLVENT SALTS 573 BeFg + Liquid LigBeFy + Liquid BeFy mole % LiBeF3 + BeFg The system LiF-BeFs. 900 \ ¥ 800 \ NaF + Liquid 700 oe=0rnl Data a — Na,BeF, 4 LIQUID 2°°74 600 p /’ / ’ a — NagBeF 4 + NaF \ i - SSIB;E3 BeF, -+ LIQUID 400 BeFy + 8'— NaBeF3 , /Z " o ! a— NCII2BeF4+Bl—NOBeF,3\* \ l A ! NoBoF 8 — NaBeF4 + LIQUID 300 '~ NagBeF 4 +NaF B~ Nalefs | [BeF5+ 6 — NabeFg B8 — NaBeF;, +v — Na,BeF , < a’'— NanBeF BeF — NoBeF 1BeFy+7 - NagBefyrs 0284 | BeFp+ - Nabey 200 ¥ — NagBeF4 + NaF = 1 I NaF 10 20 30 40 50 60 70 80 90 BeF2 BeF,, mole % Fic. 12-4. The system NaF-BeF. 574 CHEMICAL ASPECTS: MOLTEN-FLUORIDE-SALT FUELS [cHAP. 12 BeF, Dotted Lines Represent 542 Incompletely Defined \ Phase Boundaries and Alkemade Lines The Symbol TC Represents All Temperatures Are in °C a Compound Whose Exact Composition Has Not 370 Been Determined NaF-BeF9 TC 356 P 345 < 350 (LiF-BeF9) 28/0 7& 50— 400 400~ 2 NaF (-2 BeFs 4 fl; ].. s 4555 BT “%fifiy’ =X/ /" 5 e 7 /"‘ 2 Na5F°BeF2 e, s ] P grum———= (- - BT ——— | mBS R . (/ / BT SN P 750; “ ‘ (NaF-LiF-BeFp)) ~Z X | 850 800 " 240 F 5";%336&) . 200 [ / \ LiF 800 750 700 649 700 750 800 850 900 950 NaF F16. 12-5. The system LiF-NaF-BeF's. The lowest melting binary systems are those containing BeFs and LiF or NaF. Since BeF'2 offers the best cross section of all the useful diluents, fuels based on these binary systems are likely to be of highest interest in thermal reactor designs. | The binary system LiF-BeF: has melting points below 500°C over the concentration range from 33 to 80 mole 9, BeF2. The presently accepted LiF-BeF 2 system diagram presented in Fig. 12-3 differs substantially from previously published diagrams [1-3]. It is characterized by a single eutectic between BeF2 and 2LiF - BeF2 that freezes at 356°C and contains 52 mole % BeFs. The compound 2LiF - BeF2 melts incongruently to LiF and liquid at 460°C; LiF - BeF'2 is formed by the reaction of solid BeFs and solid 2LiF - BeF2 below 274°C. The diagram of the NaF-BeF2 system (Fig. 12-4) is similar to that of the LiF'-BeF; system. The ternary system combining both NaF and LiF with Bel's, shown in Fig. 12-5, offers a wide variety of low-melting compo- sitions. Some of these are potentially useful as low-melting heat-transfer liquids, as well as for reactor fuels. 575 CHOICE OF BASE OR SOLVENT SALTS 12-1] C'8 C6Se 18600 920 6 e6°¢ 028 (F-9%-09) v IN-YA1Z-18N G 0T 9% 0 eF 0S°Z 00¥ (8°'2-2%—5°69) *IN-*A9d-ABN 78 .80 0¥ 88°% ¥9% (8°3-S"08-L9) YIN- gV c—0T X q v g | v 0,009 ¥V weld /8o Do 0 ] J[ow c1o.4/g?V = U 0002 3% D)L —V =0 jutod ao%mhgaoo K10edBd 189 00/3 SuneN T astodIjued ‘AJ1S00ST A ‘Ay1suep pnbry SITVQ ONIIVAg Tanj 40 XLISOOSIA ANV ALISNA(] ¥OJd SNOILVADG ANV ‘SHILIOVAV)) LV ‘SINIOJ ONILTEA C—GI dT1dV], 576 CHEMICAL ASPECTS: MOLTEN-FLUORIDE-SALT FUELS [cHAP. 12 TABLE 124 TaERMAL ConbpucTIiviTY OF TYPICAL FLUORIDE FUELS Thermal conductivity, Composition, Btu/ (hr) (ft) °F) mole 9, Solid Liquid LiF-NaF-KF-UF,4 (44.5-10.9-43.5-1.1) NaF-ZrF4,~UF, (50-46-4) NaF-ZrF,~UF4 (53.5-40-6.5) NaF-KF-UF, (46.5-26-27.5) SN o O O = == DN O W W All Temperatures Are in °C 5NaF-2ZrF4 2NaF-ZrF4 3 NaF:2 ZrFy ZrFy 918 00 Fi1g. 12-6. The system NaF-ZrF,;-UF,. 12-2] FUEL AND BLANKET SOLUTIONS 577 UFy All Temperatures Are in °C E—Eutectic P—Peritectic LiF- 4UF, :Primary Phase Field \“\ EAN .\ PN T, MLF-4UR ALiF-UFy A AN D Q e, N\ 7LiF- 6UF, /\. o \O \ NS 2 e = |LiF.X¢ < \ 00\/%\ \7‘3‘0\\ 5 ‘}-:;T\‘ 7=\ E LiF 2LiF-BeF,” B 400”400 BeF, A‘\; 3 ] Fi1a. 12-7. The system LiF-BeF3-UF 4. 12-2. FuEL AND BLANKET SOLUTIONS 12-2.1 Choice of uranium fluoride. Uranium hexafluoride is a highly volatile compound, and it is obviously unsuitable as a component of a liquid for use at high temperatures. The compound UO2F3, which is rela- tively nonvolatile, is a strong oxidant that would be very difficult to con- tain. Fluorides of pentavalent uranium (UF5,UsFy, etc.) are not thermally stable [4] and would be prohibitively strong oxidants even if they could be stabilized in solution. Uranium trifluoride, when pure and under an inert atmosphere, is stable even at temperatures above 1000°C [4,5]; however, it is not so stable in molten fluoride solutions [6]. It disproportionates appreciably in such media by the reaction 4 UF; =3 UF4++4 U9, at temperatures below 800°C. Small amounts of UF'3 are permissible in the presence of relatively large concentrations of UF4 and may be beneficial insofar as corrosion is concerned. It is necessary, however, to use UF4 as the major uraniferous compound in the fuel. 578 CHEMICAL ASPECTS: MOLTEN-FLUORIDE-SALT FUELS [cHAP. 12 UF4 All Temperatures Are in °C E—Eutectic P—Peritectic NaU4F17 —Primary Phase Field 548° NGF‘2UF4 \ E N\ :\\\\ 5NaF-3UF4 4“@ INaF-UF.F \"‘ N PNl \650) A VN R e\ Nm X - NaF E 2NaF-BeF9 NaF-BeF, Fig. 12-8. The system NaF-BeF2-UF4. 12-2.2 Combination of UF,; with base salts. The fuel for the Aircraft Reactor Experiment (Chapter 16) was a mixture of UF4 with the NaF-ZrF4 base salt. The ternary diagram for this system is shown in Fig. 12-6. The compounds ZrF4 and UF4 have very similar unit cell parameters [4] and are isomorphous. They form a continuous series of solid solutions with a minimum melting point of 765°C for the solution containing 23 mole 9, UF4. This minimum is responsible for a broad shallow trough which pene- trates the ternary diagram to about the 45 mole 9, NaF composition. A continuous series of solid solutions without a maximum or a minimum exists between a—3NaF - UF4 and 3NaF - ZrF4; in this solution series the temperature drops sharply with decreasing ZrF4 concentration. A con- tinuous solid-solution series without a maximum or a minimum also exists between the isomorphous congruent compounds 7NaF - 6UF4+ and 7NaF - 6ZrF4; the liquidus decreases with increasing ZrF4 content. These two solid solutions share a boundary curve over a considerable composition range. The predominance of the primary phase fields of the three solid solutions presumably accounts for the complete absence of a ternary eutectic in this complex system. The liquidus surface over the area below 8 mole 9, UF4 and between 60 and 40 mole 9, NaF is relatively flat. All fuel compositions within this region have acceptable melting points. Minor 12-2] FUEL AND BLANKET SOLUTIONS 579 ThF,4 1080°C \060 A000" 950° 900° 850° LiF . 845°C ligBeF, LiBeFs (?) 543°C 475°C 360°C F1ac. 12-9. The system LiF-BeF2-ThFy. advantages in physical and thermal properties accrue from choosing mix- tures with minimum ZrF4 content in this composition range. Typical physical and thermal properties are given in Tables 12-3 and 12—4. The nuclear studies in Chapter 14 indicate that the combination of BeFs with NaF or with LiF (provided the separated Li? isotope can be used) are more suitable as reactor fuels. The diagram of Fig. 12-7 reveals that melting temperatures below 500°C can be obtained over wide com- position ranges in the three-component system LiF-BeFs-UF4. The lack of a low-melting eutectic in the NaF-UF4 binary system is responsible for melting points below 500°C being available over a considerably smaller concentration interval in the NaF-BeF2:-UF4 system (Fig. 12-8) than in its LiF-BeFo—UF4 counterpart. The four-component system LiF-NalF-BeF.-UF4 has not been com- pletely diagrammed. It is obvious, however, from examination of Fig. 12-5 that the ternary solvent LiF-NaF-BeF's offers a wide variety of low-melting compositions; it has been established that considerable quantities (up to at least 10 mole 9,) of UF4 can be added to this ternary system without elevation of the melting point to above 500°C. o980 CHEMICAL ASPECTS: MOLTEN-FLUORIDE-SALT FUELS [CHAP. 12 7 LiF-6 ThF4 45 LiF-2 ThF4 40 1 E—Eutectic \? P—Peritectic o 35 Liquidus s? P—59 Temperatures & Are in °C év 30 K E-560 3 LiF+ThF, E~-570 20 550 %50 15 - 650%00 10 700 0 750 \ 5 800 e T \ : 450 LI VoAV W NN NN N 5 10 15 20 25 30 35 40 45 2 LiF-BeFy P-490 BeF9 (mole %) Fia. 12-10. The system LiF-BeF3:-ThF4 in the concentration range 50 to 100 mole 9, LiF. 12-2.3 Systems containing thorium fluoride. All the normal compounds of thorilum are quadrivalent; accordingly, any use of thorium in molten fluoride melts must be as ThF4. A diagram of the LiF-BeFo—ThF4 ternary system, which is based solely on thermal data, is shown as Fig. 12-9. Recent studies in the 50 to 100 mole 9, LiF concentration range have demonstrated (Fig. 12-10) that the thermal data are qualitatively correct. Breeder reactor blanket or breeder reactor fuel solvent compositions in which the maximum ThF4 concentration is restricted to that available in salts having less than a 550°C liquidus may be chosen from an area of the phase diagram (Fig. 12-10) in which the upper limits of ThF4 concentra- tion are obtained in the composition 75 mole 9, LiF-16 mole 9, ThF+-9 mole % BeFq, 69.5 mole 9, LiF-21 mole 9%, ThF 4+9.5 mole 9, BeFq, 68 mole 9, LiF-22 mole 9, ThF4+~10 mole 9, BeFs. 12-2.4 Systems containing Thy and UFs;. The LiF-BeF>-UF4 and the LiF-BeF2-ThF, ternary systems are very similar; the two eutectics in the LiF-BeFs-ThF4 system are at temperatures and compositions virtually identical with those shown by the UF4-bearing system. The very great 12-3] PROPERTIES OF FLUORIDE MIXTURES 581 similarity of these two ternary systems and preliminary examination of the LiF-Bel's>-ThF4+~UF4 quaternary system suggests that fractional re- placement of UF4 by ThF4 will have little effect on the freezing tem- perature over the composition range of interest as reactor fuel. 12-2.5 Systems containing PuF3. The behavior of plutonium fluorides in molten fluoride mixtures has received considerably less study. Plu- tonium tetrafluoride will probably prove very soluble, as have UF4 and ThF4, in suitable fluoride-salt diluents, but is likely to prove too strong an oxidant to be compatible with presently available structural alloys. The trifluoride of plutonium dissolves to the extent of 0.25 to 0.45 mole 9 in LiF-Bel's mixtures containing 25 to 50 mole 9, BeF2. As indicated in Chapter 14, it is believed that such concentrations are in excess of those required to fuel a high-temperature plutonium burner. 12-3. PHYSICAL AND THERMAL PROPERTIES OF FLUORIDE MIXTURES The melting points, heat capacities, and equations for density and vis- cosity of a range of molten mixtures of possible interest as reactor fuels are presented above in Tables 12-1 and 12-3, and thermal-conductivity values are given in Tables 12-2 and 12-4; the methods by which the data were ob- tained are described here. The temperatures above which the materials are completely 1in the liquid state were determined in phase equilibrium studies. The methods used included (1) thermal analysis, (2) differential- thermal analysis, (3) quenching from high-temperature equilibrium states, (4) visual observation of the melting process, and (5) phase separation by filtration at high temperatures. Measurements of density were made by welghing, with an analytical balance, a plummet suspended in the molten mixture. Enthalpies, heats of fusion, and heat capacities were determined from measurements of heat liberated when samples in capsules of Ni or Inconel were dropped from various temperatures into calorimeters; both ice calorimeters and large copper-block calorimeters were used. Measure- ments of the viscosities of the molten salts were made with the use of a capillary eflux apparatus and a modified Brookfield rotating-cylinder device; agreement between the measurements made by the two methods indicated that the numbers obtained were within 4- 109. Thermal conductivities of the molten mixtures were measured in an apparatus similar to that described by Lucks and Deem [7], in which the heating plate is movable so that the thickness of the liquid specimen can be varied. The uncertainty in these values is probably less than 4 25%. The variation of the thermal conductivity of a molten fluoride salt with temperature is relatively small. The conductivities of solid fluoride mix- tures were measured by use of a steady-state technique in which heat was passed through a solid slab. 582 CHEMICAL ASPECTS: MOLTEN-FLUORIDE-SALT FUELS [cHAP. 12 The vapor pressures of PuFg [8], UF4 [9], and ThF 4 are negligibly small at temperatures that are likely to be practical for reactor operations. Of the fluoride mixtures likely to be of interest as diluents for high-temperature reactor fuels, only AlF3, BeF: [9], and ZrF4 [10-12] have appreciable vapor pressures below 700°C. Measurements of total pressure in equilibrium with NaF-ZrF4-UF, melts between 800 and 1000°C with the use of an apparatus similar to that described by Rodebush and Dixon [13] yielded the data shown in Table 12-5. Sense et al. [14], who used a transport method to evaluate partial TABLE 12-5 VAPOR PRESSURES OF FLUORIDE MIXTURES CONTAINING ZRF4 Composition, Vapor pressure constants™® Vapor pressure mole 9, o at 900°C, mm Hg NaF ZrF, UF, A B X103 100 7.792 9.171 0.9 100 12.542 11.360 617 57 43 7.340 7.289 14 50 50 7.635 7.213 32 50 46 4 7.888 7.551 28 53 43 4 7.37 7.105 21 *For the eqfiation log P (mm Hg) = A — (B/T), where T is in °K. pressures in the NaF-ZrF4 system, obtained slightly different values for the vapor pressures and showed that the vapor phase above these liquids is quite complex. The vapor-pressure values obtained from both investi- gations are less than 2 mm Hg for the equimolar NaF-ZrF4 mixture at 700°C. However, since the vapor is nearly pure ZrF4, and since ZrF4 does not melt under low pressures of its vapor, even this modest vapor pressure leads to engineering difficulties; all lines, equipment, and connections ex- posed to the vapor must be protected from sublimed ZrF4 “‘snow.” Measurements made with the Rodebush apparatus have shown that the vapor pressure above liquids of analogous composition decreases with in- creasing size of the alkali cation. All these systems show large negative deviations from Raoult’s law, which are a consequence of the large, posi- tive, excess, partial-molal entropies of solution of ZrF4. This phenomenon has been interpreted qualitatively as an effect of substituting nonbridging 583 PROPERTIES OF FLUORIDE MIXTURES 12-3] "M, Ul st 7 oxoym ‘(/g) — v = (8H wu) J 3o uoryenbs oy 1044 ._”mfl I8 39 OSuU9Y %Q paure}qo sjep WOIJ .@omQEOO* ¢0'0 ¢l 1 ¢'8 GLIC' T LEG 6 GEOT-LG8 ¢c Gl 60 0 €901 I 080 6 L9911 G6¢ 6 GC0I—948 0¥ 09 Iy 0 L8T 1 86 IL0°T ¢4 6 966964 0¢ 0¢ ¥6°0 L8111 6.6 980 I 90 01 886—¢08 64 v 69 1 90 1 LL6 960 I &y 01 LL6—G8L VL 9¢ 70T X 701 X 70T X q V q vV q V ¢ 19d JEN wm I Oo 06008 7% %100 - AN “10d AN ‘Teatoqul 0/, sfow arnssoxd rode A arnjeradwaJ, . uorjisoduwo)) 1syue)suoo aunssard rodep «STINILXIN eHAG-JVN 40 SHINSSHYJ HOdVA 9-gI ATaV], 584 CHEMICAL ASPECTS: MOLTEN-FLUORIDE-SALT FUELS [CHAP. 12 fluoride ions for fluoride bridges between zirconium ions as the alkali fluoride concentration is increased in the melt [12]. Vapor pressure data obtained by the transport method for NaF-BeF; mixtures [15] are shown in Table 12-6, which indicates that the vapor phases are not pure BeF2. While pressures above LiF-BeF2 must be ex- pected to be higher than those shown for NaF-BeF2 mixtures, the values of Table 12-6 suggest that the “‘snow’” problem with BeFs mixtures is much less severe than with ZrF4 melts. Physical property values indicate that the molten fluoride salts are, in general, adequate heat-transfer media. It is apparent, however, from vapor pressure measurements and from spectrophotometric examination of analogous chloride systems that such melts have complex structures and are far from ideal solutions. 12—4. PRoDUCTION AND PURIFICATION OF FLUORIDE MIXTURES Since commercial fluorides that have a low concentration of the usual nuclear poisons are available, the production of fluoride mixtures is largely a purification process designed to minimize corrosion and to ensure the removal of oxides, oxyfluorides, and sulfur, rather than to improve the neutron economy. The fluorides are purified by high-temperature treat- ment with anhydrous HF and Hs gases, and are subsequently stored in sealed nickel containers under an atmosphere of helium. 12—4.1 Purification equipment. A schematic diagram of the purification and storage vessels used for preparation of fuel for the Aircraft Reactor Experiment (Chapter 16) is shown in Fig. 12-11. The reaction vessel in which the chemical processing is accomplished and the receiver vessel into which the purified mixture is ultimately transferred are vertical cylindrical containers of high-purity low-carbon nickel. The top of the reactor vessel is pierced by a charging port which is capped well above the heated zone by a Teflon-gasketed flange. The tops of both the receiver and the reaction vessels are pierced by short risers which terminate in Swagelok fittings, through which gas lines, thermowells, etc., can be introduced. A transfer line terminates near the bottom of the reactor vessel and near the top of the receiver; entry of this tube is effected through copper-gasketed flanges on l-in.-diameter tubes which pierce the tops of both vessels. This transfer line contains a filter of micrometallic sintered nickel and a sampler which collects a specimen of liquid during transfer. Through one of the risers in the receiver a tube extends to the receiver bottom; this tube, which is sealed outside the vessel, serves as a means for transfer of the purified mixture to other equipment. This assembly is connected to a manifold through which He, H2, HF, or vacuum can be supplied to either vessel. By a combination of large tube 12-4] PURIFICATION OF FLUORIDE MIXTURES 985 Teflon-Gasketed Flange 4-in. Dia Charging Port I ) \—D 3/8-in. Nickel Transfer Line w Filter Reaction Vessel | . Receiver Vessel/ p Fic. 12-11. Diagram of purification and storage system. furnaces, resistance heaters, and lagging, sections of the apparatus can be brought independently to controlled temperatures in excess of 800°C. 12-4.2 Purification processing. The raw materials, in batches of proper composition, are blended and charged into the reaction vessel. The material is melted and heated to 700°C under an atmosphere of anhydrous HF to remove HoO with a minimum of hydrolysis. The HF is replaced with Ho for a period of 1 hr, during which the temperature is raised to 800°C, to reduce U5+ and Ut to U%+ (in the case of simulated fuel mixtures), and sulfur compounds to S~ and extraneous oxidants (Fe** ™, for example) to 586 CHEMICAL ASPECTS: MOLTEN-FLUORIDE-SALT FUELS [cHAP. 12 lower valence states. The hydrogen, as well as all subsequent reagent gases, is fed at a rate of about 3 liters/min to the reaction vessel through the re- ceiver and transfer line and, accordingly, it bubbles up through the molten charge. The hydrogen is then replaced by anhydrous HF, which serves, during a 2- to 3-hr period at 800°C, to volatilize H2S and HCI and to con- vert oxides and oxyfluorides of uranium and zirconium to tetrafluorides at the expense of dissolution of considerable NiF; into the melt through re- action of HF with the container. A final 24- to 30-hr treatment at 800°C with Ho suffices to reduce this NiFs and the contained FeF2 to soluble metals. At the conclusion of the purification treatment a pressure of helium above the salt in the reactor vessel is used to force the melt through the transfer line with its filter and sampler into the receiver. The metallic iron and nickel are left in the reactor vessel or on the sintered nickel filter. The purified melt is permitted to freeze under an atmosphere of helium in the receiver vessel. 12-5. RADIATION STABILITY OF FLUORIDE MIXTURES When fission of an active constituent occurs in a molten fluoride solu- tion, both electromagnetic radiations and particles of very high energy and intensity originate within the fluid. Local overheating as a consequence of rapid slowing down of fission fragments by the fluid is probably of little consequence in a reactor where the liquid is forced to flow turbulently and where rapid and intimate mixing occurs. Moreover, the bonding in such liquids is essentially completely ionic. Such a solution, which has neither covalent bonds to sever nor a lattice to disrupt, should be quite resistant to damage by particulate or electromagnetic radiation. More than 100 exposures to reactor radiation of various fluoride mix- tures containing UF4 in capsules of Inconel have been conducted; in these tests the fluid was not deliberately agitated. The power level of each test was fixed by selecting the U235 content of the test mixture. Thermal neu- tron fluxes have ranged from 10! to 10!* neutrons/(cm?)(sec) and power levels have varied from 80 to 8000 w/cm3. The capsules have, in general, been exposed at 1500°F for 300 hr, although several tests have been con- ducted for 600 to 800 hr. A list of the materials that have been studied is presented in Table 12-7. Methods of examination of the fuels after irra- diation have included (1) freezing-point determinations, (2) chemical analysis, (3) examination with a shielded petrographic microscope, (4) as- say by mass spectrography, and (5) examination by a gamma-ray spectro- scope. The condition of the container was checked with a shielded metal- lograph. No changes in the fuel, except for the expected burnup of U235 have been observed as a consequence of irradiation. Corrosion of the Inconel 12-5] RADIATION STABILITY OF FLUORIDE MIXTURES TaBLE 12-7 MoLTEN SALTS WHIicH HAvE BEEN STUDIED IN IN-PiLE CaprsULE TEsTS 587 Composition, System mole % NaF-KF-UF4 46.5-26-27.5 NaF-BeF-UF, 25-60-15 NaF-BeFo-UF4 47-51-2 NaF-BeFs-UF4 50-46-4 NaF—ZrF4—UF4 63—-25—-12 NaF-ZrF,~UF, 53.5-40-6.5 NaF-ZrF4-UF, 50-48-2 NaF-ZrFUF; 50-48-2 TABLE 12-8 DESCRIPTIONS OF INCONEL ForcED-CircuraTioN Loors OrERATED IN THE LITR aAnp THE MTR Loop designation LITR LITR MTR Horizontal Vertical Horizontal NaF-ZrF ,~UF 4 composition, mole 9 62.5-12.5-25 | 63-25-12 | 53.5-40-6.5 Maximum fission power, w/cm3 400 500 800 Total power, kw - 2.8 10 20 Dilution factor* 180 7.3 5 Maximum fuel temperature, °F 1500 1600 1500 Fuel temperature differential, °F 30 250 155 Fuel Reynolds number 6000 3000 5000 Operating time, hr 645 332 467 Time at full power, hr 475 235 271 *Ratio of volume of fuel in system to volume of fuel in reactor core. 588 CHEMICAL ASPECTS: MOLTEN-FLUORIDE-SALT FUELS [CHAP. 12 capsules to a depth of less than 4 mils in 300 hr was found; such corrosion is comparable to that found in unirradiated control specimens [16]. In capsules which suffered accidental excursions in temperatures to above 2000°F, grain growth of the Inconel occurred and corrosion to a depth of 12 mils was found. Such Increases in corrosion were almost certainly the result of the serious overheating rather than a consequence of the radiation field. Tests have also been made in which the fissioning fuel is pumped through a system in which a thermal gradient is maintained in the fluid. These tests included the Aircraft Reactor Experiment (described in Chapter 16) and three types of forced-circulation loop tests. A large loop, in which the pump was outside the reactor shield, was operated in a horizontal beam hole of the LITR.* A smaller loop was operated in a vertical position in the LITR lattice with the pump just outside the lattice. A third loop was operated completely within a beam-hole of the MTR.t The operating con- ditions for these three loops are given in Table 12-8. The corrosion that occurred in these loop tests, which were of short duration and which provided relatively small temperature gradients, was to a depth of less than 4 mils and, as in the capsule tests, was comparable to that found in similar tests outside the radiation field [16]. Therefore it is concluded that within the obvious limitations of the experience up to the present time there is no effect of radiation on the fuel and no accelera- tion of corrosion by the radiation field. 12—-6. BEaAvior oF FissioN Propucts When fission of an active metal occurs in a molten solution of its fluo- ride, the fission fragments must originate in energy states and ionization levels very far from those normally encountered. These fragments, how- ever, quickly lose energy through collisions in the melt and come to equi- librium as common chemical entities. The valence states which they ulti- mately assume are determined by the necessity for cation-anion equivalence in the melt and the requirement that redox equilibrium be established among components of the melt and constituents of the metallic container. Structural metals such as Inconel in contact with a molten fluoride so- lution are not stable to F2, UF5, or UFs. It is clear, therefore, that when fission of uranium as UF, takes place, the ultimate equilibrium must be such that four cation equivalents are furnished to satisfy the fluoride ions released. Thermochemical data, from which the stability of fission-product fluorides in complex dilute solution could be predicted, are lacking in *Low Intensity Test Reactor, a tank type research reactor located at Qak Ridge, Tennessee. TMaterials Testing Reactor, a tank type research reactor located at Arco, Idaho. 12-6] BEHAVIOR OF FISSION PRODUCTS 589 many cases. No precise definition of the valence state of all fission-product fluorides can be given; it is, accordingly, not certain whether the fission process results in oxidation of the container metal as a consequence of de- positing the more noble fission products in the metallic state. 12-6.1 Fission products of well-defined valence. The noble gases. The fission products krypton and xenon can exist only as elements. The solubilities of the noble gases in NaF-ZrFs (53-47 mole %) [17], NaF-ZrF4+-UF4 (50-46—4 mole %) [17], and LiF-NaF-KF (46.5-11.5-42 mole %) obey Henry’s law, increase with increasing temperature, decrease with increasing atomic weight of the solute, and vary appreciably with composition of the solute. The Henry’s law constants and the heats of solution for the noble gases in the NaF-ZrF4 and LiF-NaF-KF mixtures are given in Table 12-9. The solubility of krypton in the NaF-ZrF 4 mix- ture appears to be about 3 X1078 moles/(cm3)(atm). TABLE 12-9 SoLUBILITIES AT 600°C AND HEATS oF SOLUTION FOR NOBLE GASES IN MoLTeEN FrLuoriDE MIXTURES In NaF-ZrF4 In LiF-NaF-KF (53—-47 mole %) (46.5-11.5-42 mole 9) Gas Heat of Heat of K* solution, K* solution, kcal/mole kcal/mole X108 X108 Helium 21.6 +1 6.2 11.34+0.7 8.0 Neon 11.3 +£0.3 7.8 4.440.2 8.9 Argon 5.1 £0.15 8.2 - Xenon 1.9440.2 11.1 *Henry’s law constant in moles of gas per cubic centimeter of solvent per atmosphere. The positive heat of solution ensures that blanketing or sparging of the fuel with helium or argon in a low-temperature region of the reactor cannot lead to difficulty due to decreased solubility and bubble formation in higher temperature regions of the system. Small-scale in-pile tests have revealed that, as these solubility data suggest, xenon at low concentration is re- tained in a stagnant melt but is readily removed by sparging with helium. Only a very small fraction of the anticipated xenon poisoning was observed 590 CHEMICAL ASPECTS: MOLTEN-FLUORIDE-SALT FUELS [CcHAP. 12 during operation of the Aircraft Reactor Experiment, even though the sys- tem contained no special apparatus for xenon removal [18]. It seems certain that krypton and xenon isotopes of reasonable half-life can be readily removed from all practical molten-salt reactors. Elements of Groups I-A, II-A, II11-B, and IV-B. The fission products Rb, Cs, Sr, Ba, Zr, Y, and the lanthanides form very stable fluorides; they should, accordingly, exist in the molten fluoride fuel in their ordinary valence states. High concentrations of ZrF4 and the alkali and alkaline earth fluorides can be dissolved in LiF-NaF-KF, LiFs—BeF3, or NaF-ZrF4 mixtures at 600°C. The solubilitids at 600°C of YF3 and of selected rare- earth fluorides in NaF-ZrF4 (53-47 mole %) and LiF-BeF2 (65-35 mole %) are shown in Table 12-10. For these materials the solubility increases TaBLE 12-10 SoLUBILITY OF YF3 AND oF SOME RARE-EARTH FLUORIDES IN NaAF-ZrF4 anp 1N LiIF-BgFs aT 600°C Solubility, mole 9, MF3 Fluoride ) In NaF-ZrF, In LiF-BekFs (57-43 mole 9) (62—-38 mole %)) YF; 3.6 LaFg 2.1 CelF'3 2.3 0.48 about 0.5%/°C and increases slightly with increasing atomic number in the lanthanide series; the saturating phase is the simple trifluoride. For solutions containing more than one rare earth the primary phase is a solid solution of the rare-earth trifluorides; the ratio of rare-earth cations in the molten solution is virtually identical with the ratio in the precipitated solid solution. Quite high burnups would be required before a molten fluoride reactor could saturate its fuel with any of these fission products. 12-6.2 Fission products of uncertain valence. The valence states as- sumed by the nonmetallic elements Se, Te, Br, and I must depend strongly on the oxidation potential defined by the container and the fluoride melt, and the states are not at present well defined. The sparse thermochemical data suggest that if they were in the pure state the fluorides of Ge, As, Nb, Mo, Ru, Rh, Pd, Ag, Cd, Sn, and Sb would be reduced to the cor- responding metal by the chromium in Inconel. While fluorides of some of 12-7] FUEL REPROCESSING 091 these elements may be stabilized in dilute molten solution in the melt, it is possible that none of this group exists as a compound in the equilibrium mixture. An appreciable, and probably large, fraction of the niobium and ruthenium produced in the Aircraft Reactor Experiment was deposited in or on the Inconel walls of the fluid circuit; a detectable, but probably small, fraction of the ruthenium was volatilized, presumably as RuF's, from the melt. - 12-6.3 Oxidizing nature of the fission process. The fission of a mole of UF4 would yield more equivalents of cation than of anion if the noble gas 1sotopes of half-life greater than 10 min were lost and if all other elements formed fluorides of their lowest reported valence state. If this were the case .the system would, presumably, retain cation-anion equivalence by reduction of fluorides of the most noble fission products to metal and perhaps by reduction of some U4t to U3*. If, however, all the elements of uncertain valence state listed in Article 12-6.2 deposit as metals, the balance would be in the opposite direction. Only about 3.2 equivalents of combined cations result, and since the number of active anion equivalents is a minimum of 4 (from the four fluorines of UF4), the deficiency must be alleviated by oxidation of the container. The evidence from the Aircraft Reactor Experiment, the in-pile loops, and the in-pile capsules has not shown the fission process to cause serious oxidation of the container; it is possible that these experiments burned too little uranium to yield significant results. If fission of UF4 is shown to be oxidizing, the detrimental effect could be overcome by deliberate and occasional addition of a reducing agent to create a small and stable concentration of soluble UF3 in the fuel mixture. 12-7. FuEL REPROCESSING Numerous conventional processes such as solvent extraction, selective precipitation, and preferential ion exchange could be readily applied to molten fluoride fuels after solution in water. However, these liquids are readily amenable to remote handling and serve as media in which chemical reactions can be conducted. Most development efforts have, accordingly, been concerned with direct and nonaqueous reprocessing methods. Recovery of uranium from solid fuel elements by dissolution of the element in a fluoride bath followed by application of anhydrous HF and subsequent volatilization of the uranium as UFg has been described [19,20]. The volatilization step accomplishes a good separation from Cs, Sr, and the rare earths, fair separation from Zr, and-poor separation from Nb and Ru. The fission products I, Te, and Mo volatilize completely from the melt. The nonvolatile fission products are discarded in the fluoride solvent. Further decontamination of the UFg is effected by selective ab- 592 CHEMICAL ASPECTS: MOLTEN-FLUORIDE-SALT FUELS [cHAP. 12 sorption and desorption on beds of NaF. At 100°C, UF¢ is absorbed on the bed by the reversible reaction UFe¢ (g) + 3NaF—=3NaF - UF, which was first reported by Martin, Albers, and Dust [21]. Niobium ac- tivity, along with activity attributable to particulate matter, is also absorbed; ruthenium activity, however, largely passes through the bed. Subsequent desorption of the UF¢ at temperatures up to 400°C is accom- plished without desorption of the niobium. The desorbed UFg is passed through a second NaF bed held at 400°C as a final step and is subsequently recovered in refrigerated traps. The decontaminations obtained are greater than 10° for gross beta and gamma emitters, greater than 107 for Cs, Sr, and lanthanides, greater than 10° for Nb, and about 10% for Ru. Uranium was recovered from the molten-salt fuel of the Aircraft Reactor Experiment by this method, and its utility for molten-fluoride fuel systems or breeder blankets was demonstrated. Recovery of plutonium or thorium, however, is not possible with this process. There are numerous possible methods for reprocessing molten-salt fuels. The behavior of the rare-earth fluorides indicates that some decontamina- tion of molten-fluoride fuels may be obtained by substitution of CeFs or LaF's, in a sidestream circuit, for rare earths of higher cross section. It seems likely that Pul's can be recovered with the rare-earth fluorides and subsequently separated from them after oxidation to Pul4. Further, it appears that both selective precipitation of various fission-product ele- ments and active constituents as oxides, and selective chemisorption of these materials on solid oxide beds are capable of development into valu- able separation procedures. Only preliminary studies of these and other possible processes have been made. 993 REFERENCES 1. D. N. Roy et al., Fluoride Model Systems: IV. The Systems LiF—BeF; and RbFs—BeF;, J. Am. Ceram. Soc. 37, 300 (1954). 2. A. V. NovoseLova et al.,, Thermal and X-ray Analysis of the Lithium- Beryllium Fluoride System, J. Phys. Chem. USSR 26, 1244 (1952). 3. W. R. GriMEs et al., Chemical Aspects of Molten Fluoride Reactors, paper to be presented at Second International Conference on Peaceful Uses of Atomic Energy, Geneva, 1958. 4. J. J. Katz and E. RaBiNnowircH, The Chemastry of Uranitum, National Nuclear Energy Series, Division VIII, Volume 5. New York: McGraw-Hill Book Co., Inc., 1951. 5. W. R. GriMEs et al., Oak Ridge National Laboratory. Unpublished. 6. B. H. CLampiTT et al., Oak Ridge National Laboratory, 1957. Unpublished. 7. C. F. Lucks and H. W. DeEM, Apparatus for Measuring the Thermal Conductwity of Liquid at Elevated Temperatures; Thermal Conductivity of Fused NaOH to 600°C, Am. Soc. of Mech. Eng. Meeting, June 1956. (Preprint 56SA31) 8. G. T. SEaBorG and J. J. KaTrz (Eds.), The Actinide Elements, National Nuclear Energy Series, Division IV, Volume 14A. New York: McGraw-Hill Book Co., Inc., 1953. 9. W. R. GrimMEs et al.,, Fused-salt Systems, Sec. 6 in Reactor Handbook, Vol. 2, Engineering, USAEC Report AECD-3646, 1955. (pp. 799-850) 10. K. A. SENsE et al., The Vapor Pressure of Zirconium Fluoride, J. Phys. Chem. 58, 995 (1954). 11. W. FiscHER, Institut fiir Anorganische Chemie, Technicshe Hochschule, Hannover, personal communication; [Data for equation taken from S. LAUTER, Dissertation, Institut fiir Anorganische Chemie, Technische Hochschule, Hannover (1948).] 12. S. CanToR et al., Vapor Pressures and Derived Thermodynamic Informa- tion for the System RbF—ZrF4, J. Phys. Chem. 62, 96 (1958). 13. W. H. RopeBusH and A. L. DixonN, The Vapor Pressures of Metals; A New Experimental Method, Phys. Rev. 26, 851 (1925). 14. K. A. SENSE et al., Vapor Pressure and Derived Information of the Sodium Fluoride-Zirconium Fluoride System. Description of a Method for the De- termination of Molecular Complexes Present in the Vapor Phase, J. Phys. Chem. 61, 337 (1957). 15. K. A. SENSE et al., Vapor Pressure and Equilibrium Studies of the Sodium Fluoride-Beryllium Fluoride System, USAEC Report BMI-1186, Battelle Me- morial Institute, May 27, 1957. 16. W. D. MANLY et al., Metallurgical Problems in Molten Fluoride Systems, paper to be presented at Second International Conference on the Peaceful Uses of Atomic Energy, Geneva, 1958. | 17. W. R. GriMmEs et al., Solubility of Noble Gases in Molten Fluorides. I. In Mixtures of NaF—ZrF4 (53-47 mole 9;) and NaF—ZrF4+—UF4 (50-46-4 mole %), J. Phys. Chem. (in press). 594 CHEMICAL ASPECTS: MOLTEN-FLUORIDE-SALT FUELS [CHAP. 12 18. E. S. BeTtis et al., The Aircraft Reactor Experiment—Operation, Nuclear Sct. and Eng. 2, 841 (1957). 19. G. I. CatuEers, Uranium Recovery for Spent Fuel by Dissolution in Fused Salt and Fluorination, Nuclear Sct. and Eng. 2, 768 (1957). 20. F. R. Bruck et al., in Progress tn Nuclear Energy, Series I1I, Process Chemistry, Vol. I. New York: McGraw-Hill Book Co., Inc., 1956. 21. Von H. MaRrTIN et al., Double Fluorides of Uranium Hexafluoride, Z. anorg. u. allgem. Chem. 265, 128 (1951). CHAPTER 13 CONSTRUCTION MATERIALS FOR MOLTEN-SALT REACTORS* 13—-1. SURVEY OF SUITABLE MATERIALS A molten-salt reactor system requires structural materials which will effectively resist corrosion by the fluoride salt mixtures utilized in the core and blanket regions. Evaluation tests of various materials in fluoride salt systems have indicated that nickel-base alloys are, in general, superior to other commercial alloys for the containment of these salts under dynamic flow conditions. In order to select the alloy best suited to this application, an extensive program of corrosion tests was carried out on the available commercial nickel-base alloys, particularly Inconel, which typifies the chromium-containing alloys, and Hastelloy B, which is representative of the molybdenum-containing alloys. Alloys contalning appreciable quantities of chromium are attacked by molten salts, mainly by the removal of chromium from hot-leg sections through reaction with UFy, if present, and with other oxidizing impurities in the salt. The removal of chromium is accompanied by the formation of subsurface voids in the metal. The depth of void formation depends strongly on the operating temperatures of the system and on the com- position of the salt mixture. On the other hand, Hastelloy B, in which the chromium is replaced with molybdenum, shows excellent compatibility with fluoride salts at tempera- tures in excess of 1600°F. Unfortunately, Hastelloy B cannot be used as a structural material in high-temperature systems because of its age- hardening characteristics, poor fabricability, and oxidation resistance. The information gained in the testing of Hastelloy B and Inconel led to the development of an alloy, designated INOR-8, which combines the better properties of both alloys for molten-salt reactor construction. The approximate compositions of the three alloys, Inconel, Hastelloy B, and INOR-S8, are given in Table 13-1. INOR-8 has excellent corrosion resistance to molten fluoride salts at temperatures considerably above those expected in molten-salt reactor service; further, no measurable attack has been observed thus far in tests at reactor operating temperatures of 1200 to 1300°F. The mechanical properties of INOR-8 at operating temperatures are superior to those of many stainless steels and are virtually unaffected by long-time exposure *By W. D. Manly, J. W. Allen, W. H. Cook, J. H. DeVan, D. A. Douglas, H. Inouye, D. H. Jansen, P. Patriarca, T. K. Roche, G. M. Slaughter, A. Taboada, and G. M. Tolson. 295 l 596 MOLTEN-SALT REACTOR CONSTRUCTION MATERIALS [cHAP. 13 TaBLE 13-1 CoMPOSITIONS OF POTENTIAL STRUCTURAL MATERIALS Quantity in alloy, w/o Components Inconel INOR-8 Hastelloy B Chromium 14-17 6-8 1 (max) Iron 6-10 5 (max) 4-7 Molybdenum 15-18 26-30 Manganese 1 (max) 0.8 (max) 1.0 (max) Carbon 0.15 (max) 0.04-0.08 0.05 (max) Silicon 0.5 0.35 (max) 1.0 (max) Sulfur 0.01 0.01 (max) 0.03 (max) Copper 0.5 0.35 (max) Cobalt 0.2 (max) 2.5 (max) Nickel 72 (min) Balance Balance _»g _ | Hot-Leg Samples Clam- % o shell Heaters fig 1) Cold-Leg Samples Fig. 13-1. Diagram of a standard thermal-convection loop, showing locations at which metallographic sections are taken after operation. 13-1] SURVEY OF SUITABLE MATERIALS 597 D Cold Sectio \ / / § & / FI/ Heated Sections 5 1 | 1 | ]° = h— /Freeze Valve Heater Connecting Lugs / . \/ Fill Tank Fic. 13-2. Diagram of forced-circulation loop for corrosion testing. to salts. The material is structurally stable in the operating temperature range, and the oxidation rate is less than 2 mils in 100,000 hr. No difficulty is encountered in fabricating standard shapes when the commercial prac- tices established for nickel-base alloys are used. Tubing, plates, bars, forgings, and castings of INOR-8 have been made successfully by several major metal manufacturing companies, and some of these companies are prepared to supply it on a commercial basis. Welding procedures have been established, and a good history of reliability of welds exists. The material has been found to be easily weldable with rod of the same com- position. Inconel is, of course, an alternate choice for the primary-circuit struc- tural material, and much information is available on its compatibility with molten salts and sodium. Although probably adequate, Inconel does not have the degrée of flexibility that INOR-8 has in corrosion resistance to different salt systems, and its lower strength at reactor operating tempera- tures would require heavier structural components. A considerable nuclear advantage would exist in a reactor with an uncanned graphite moderator exposed to the molten salts. Long-time exposure of graphite to a molten salt results in the salt penetrating the available pores, but it is probable, with the “impermeable’ types of 098 MOLTEN-SALT REACTOR CONSTRUCTION MATERIALS [cHAP. 13 graphite now being developed, that the degree of salt penetration en- countered can be tolerated. The attack of the graphite by the salt and the carburization of the metal container seem to be negligible if the temperature is kept below 1300°F. More tests are needed to finally establish the com- patibility of graphite-salt-alloy systems. Finally, a survey has been made of materials suitable for bearings and valve seats in molten salts. Cermets, ceramics, and refractory metals appear to be promising for this application and are presently being in- vestigated. 13-2. CoRrrOSION OF NICKEL-BASE ALLoYS BY MOLTEN SALTS 13-2.1 Apparatus used for corrosion tests. Nickel-base alloys have been exposed to flowing molten salts in both thermal-convection loops and in loops containing pumps for forced circulation of the salts. The thermal- convection loops are designed as shown in Fig. 13-1. When the bottom and an adjacent side of the loop are heated, usually with clamshell heaters, convection forces in the contained fluid establish flow rates of up to 8 ft/min, depending on the temperature difference between the heated and unheated portions of the loop. The forced-circulation loops are designed as shown in Fig. 13-2. Heat is applied to the hot leg of this type of loop by direct resistance heating of the tubing. Large temperature differences (up te 300°F) are obtained by air-cooling of the cold leg. Reynolds numbers of up to 10,000 are attainable with 1/2-in.-ID tubing, and somewhat higher values can be obtained with smaller tubing. 13-2.2 Mechanism of corrosion. Most of the data on corrosion have been obtained with Inconel, and the theory of the corrosive mechanism was worked out for this alloy. The corrosion of INOR-8 occurs to a lesser degree but follows a pattern similar to that observed for Inconel and pre- sumably the same theory applies. The formation of subsurface voids is initiated by the oxidation of chro- mium along exposed surfaces through oxidation-reduction reactions with impurities or constituents of the molten fluoride-salt mixture. As the sur- face is depleted in chromium, chromium from the interior diffuses down the concentration gradient toward the surface. Since diffusion occurs by a vacancy process and in this particular situation is essentially monodirec- tional, it is possible to build up an excess number of vacancies in the metal. These precipitate in areas of disregistry, principally at grain boundaries and impurities, to form voids. These voids tend to agglomerate and grow in size with increasing time and temperature. Examinations have demon- strated that the subsurface voids are not interconnected with each other or with the surface. Voids of the same type have been found in Inconel 13-2] MOLTEN-SALT CORROSION OF NICKEL-BASE ALLOYS 999 after high-temperature oxidation tests and high-temperature vacuum tests in which chromium was selectively removed. The selective removal of chromium by a fluoride-salt mixture depends on various chemical reactions, for example: 1. Impurities in the melt: FeFs 4+ Cr = CrFa2+ Fe. (13-1) 2. Oxide films on the metal surface: 2Fe-03 4+ 3CrF4 === 3CrO2 + 4FeFs. (13-2) 3. Co‘nstituents of the fuel: Cr+ 2UF4 == 2UF3 4 CrFo. (13-3) The ferric fluoride formed by the reaction of Eq. (13-2) dissolves in the melt and further attacks the chromium by the reaction of Eq. (13-1). The time-dependence of void formation in Inconel, as observed both in thermal-convection and forced-circulation systems, indicates that the at- tack is initially quite rapid but that it then decreases until a straight-line relationship exists between depth of void formation and time. This effect can be explained in terms of the corrosion reactions discussed above. The initial rapid attack found for both types of loops stems from the reaction of chromium with impurities in the melt [reactions (13-1) and (13-2)] and with the UF4 constituent of the salt [reaction (13-3)] to establish a quasi- equilibrium amount of CrF3: in the salt. At this point attack proceeds linearly with time and occurs by a mass-transfer mechanism which, al- though it arises from a different cause, is similar to the phenomenon of temperature-gradient mass transfer observed in liquid metal corrosion. In molten fluoride-salt systems, the driving force for mass transfer is a result of a temperature dependence of the equilibrium constant for the reaction between chromium and UFs (Eq. 13-3). If nickel and iron are considered inert diluents for chromium in Inconel, the process can be simply described. Under rapid circulation, a uniform concentration of UF4, UF3, and CrFs is maintained throughout the fluid; the concentrations must satisfy the equilibrium constant ' _ YOiF; ° ’Y urs Ncer, - N2ur, K,=K, Ky . 134 Y Yor - Y2ur, Ner - N2uF, (134) where N represents the mole fraction and 7 the activity coefficient of the indicated component. 600 MOLTEN-SALT REACTOR CONSTRUCTION MATERIALS [cHAP. 13 Fia. 13-3. Hot-leg section from an Inconel thermal-convection loop which cir- culated the fuel mixture NaF-ZrF4+UF4 (50-46-4 mole 97) for 1000 hr at 1500°F. (250%) Under these steady-state conditions, there exists a temperature 7', inter- mediate between the maximum and minimum temperatures of the loop, at which the initial composition of the structural metal is at equilibrium with the fused salt. Since Ky increases with increasing temperature, the chromium concentration in the alloy surface is diminished at temperatures higher than T and is augmented at temperatures lower than 7. In some melts, NaF-LiF-KF-UF4, for example, the equilibrium constant of reac- tion (13-3) changes sufficiently with temperature under extreme tempera- ture conditions to cause precipitation of pure chromium crystals in the cold zone. In other melts, for example NaF-ZrF,—UF4, the temperature- dependence of the corrosion equilibrium is small, and the equilibrium is satisfied at all useful temperatures without the formation of crystalline chromium. In the latter systems the rate of chromium removal from the salt stream at cold-leg regions is dependent on the rate at which chromium can diffuse into the cold-leg wall. If the chromium concentration gradient tends to be small, or if the bulk of the cold-leg surface is held at a relatively low temperature, the corrosion rate in such systems is almost negligible. It is -obvious that addition of the equilibrium concentrations of UF3 and CrF2 to molten fluorides prior to circulation in Inconel equipment would minimize the initial removal of chromium from the alloy by reac- 13-2] MOLTEN-SALT CORROSION OF NICKEL-BASE ALLOYS 601 Fia. 13-4. Hot-leg section of Inconel thermal-convection loop which circulated the fuel mixture NaF-ZrF+UF, (55.3-40.7-4 mole 9,) for 1000 hr at 1250°F. (250 %) tion (13-3). (It would not, of course, affect the mass-transfer process which arises as a consequence of the temperature-dependence of this reaction.) Deliberate additions of these materials have not been practiced in routine corrosion tests because (1) the effect at the uranium concentrations nor- mally employed is small, and (2) the experimental and analytical difficul- ties are considerable. Addition of more than the equilibrium quantity of UF5 may lead to deposition of some uranium metal in the equipment walls through the reaction 4UF4s == 3UF4++ U°. (13-5) For ultimate use in reactor systems, however, it may be possible to treat the fuel material with calculated quantities of metallic chromium to pro- vide the proper UF3 and CrF concentrations at startup. According to the theory described above, there should be no great dif- ference in the corrosion found in thermal-convection loops and in forced- circulation loops. The data are in general agreement with this conclusion so long as the same maximum metal-salt interface temperature is present in both types of loop. The results of many tests with both types of loop are summarized in Table 13-2 without distinguishing between the two types of loop. The maximum bulk temperature of the salt as it left the heated section of the loop is given. It is known that the actual metal-salt interface temperature was not greater than 1300°F in the loops with a 602 MOLTEN-SALT REACTOR CONSTRUCTION MATERIALS [CHAP. 13 TaBLE 13-2 SuMMARY OF CoORROSION DaTA OBTAINED IN THERMAL-CONVECTION AND ForcED-CircuLATION Loor TEsts oF INCONEL aND INOR-8 ExposED TO VARIOUS CIRCULATING SALT MIXTURES Maxi 1t Ti ¢ Depth of subsurface Constituents of UF4 or ThF, Loop ¢ ax1muri1 S8 1m(z'o void formation at base salts content material emp eor ature, °p erl?r ton, hottest part of loop, in. NaF-ZrF, 1 mole 9, UF4 Inconel 1250 1000 <0.001 1 mole 9, UF, Inconel 1270 6300 0-0.0025 4 mole 9, UF, Inconel 1250 1000 0.002 4 mole 9, UF4 Inconel 1500 1000 0.007-0.010 ) 4 mole 9, UF, INOR-8 1500 1000 . 0.002-0.003 0 Inconel 1500 1000 0.002-0.003 NaF-BeF; 1 mole 9, UF, Inconel 1250 1000 0.001 0 Inconel 1500 500 0.004-0.010 3 mole 9, UF, Inconel 1500 500 0.008-0.014 1 mole 9, UF4 . INOR-8 1250 6300 0.001 LiF-BeF; 1 mole 9, UF, Inconel 1250 1000 0.001-0.002 3 mole %, UF4 Inconel 1500 500 0.012-0.020 1 mole 9, UF, INOR-8 1250 1000 0 NaF-LiF-BeF; 0 Inconel 1125 1000 0.002 0 Inconel 1500 500 0.003-0.005 3 mole 9, UF, Inconel 1500 500 0.008-0.013 NaF-LiF-KF 0 Inconel 1125 1000 0.001 2.5 mole 9, UF, Inconel 1500 500 0.017 0 INOR-8 1250 1340 0 2.5 mole 9, UF, INOR-8 1500 1000 0.001-0.003 LiF 29 mole 9, ThF, Inconel 1250 1000 0-0.0015 NaF-BeF, 7 mole 9, ThF, INOR-8 1250 1000 0 maximum salt temperature of 1250°F, and was between 1600 and 1650°F for the loop with a maximum salt temperature of 1500°F, The data in Table 13-2 are grouped by types of base salt because the salt has a definite effect on the measured attack of Inconel at 1500°F. The salts that contain BeF2 are somewhat more corrosive than those containing ZrF4, and the presence of LiF, except in combination with NaF, seems to accelerate corrosion. At the temperature of interest in molten-salt reactors, that is, 1250°F, the same trend of relative corrosiveness of the different salts may exist for Inconel, but the low rates of attack observed in tests preclude a conclusive decision on this point. Similarly, if there is any preferential effect of the base salts on INOR-8, the small amounts of attack tend to hide it. As expected from the theory, the corrosion depends sharply on the UF4 concentration. Studies of the nuclear properties of molten-salt power reactors have indicated (see Chapter 14) that the UF4 content of the fuel will usually be less than 1 mole %, and therefore the corrosiveness of salts 13-2] MOLTEN-SALT CORROSION OF NICKEL-BASE ALLOYS 603 Fig. 13-5. Hot-leg section of Inconel thermal-convection loop which circulated the fuel mixture LiF-BeF3s-UF4 (62-37-1 mole 9,) for 1000 hr at 1250°F. (250X) with higher UF4 concentrations, such as those described in Table 13-2, will be avoided. The extreme effect of temperature is also clearly indicated in Table 13-2. In general, the corrosion rates are three to six times higher at 1500°F than at 1250°F. This effect is further emphasized in the photomicrographs presented in Figs. 13-3 and 13-4, which offer a comparison of metallo- graphic specimens of Inconel that were exposed to similar salts of the NaF- 7rF~UF4 system at 1500°F and at 1250°F. A metallographic specimen of Inconel that was exposed at 1250°F to the salt proposed for fueling of the molten-salt power reactor is shown in Fig. 13-5. 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 corrosion by the salt mixture, the mass transfer in sodium-containing systems 1s extremely dependent on the maximum system operating temperature. The results of 604 MOLTEN-SALT REACTOR CONSTRUCTION MATERIALS [cHAP. 13 numerous tests indicate that the nickel-base alloys, such as Inconel and INOR-S8, are satisfactory containers for sodium at temperatures below 1300°F, and that above 1300°F the austenitic stainless steels are preferable. 13-3. FaBricaTioN orF INOR-8 13-3.1 Casting. Normal melting procedures, such as induction or elec- tric furnace melting, are suitable for preparing INOR-8. Specialized tech- niques, such as melting under vacuum or consumable-electrode melting, have also been used without difficulty. Since the major alloying constitu- ents do not have high vapor pressures and are relatively inert, melting losses are negligible, and thus the specified chemical composition can be obtained through the use of standard melting techniques. Preliminary studies indi- cate that intricately shaped components can be cast from this material. 13-3.2 Hot forging. The temperature range of forgeability of INOR-8 1s 1800 to 2250°F. This wide range permits operations such as hammer and press forging with a minimum number of reheats between passes and substantial reductions without cracking. The production of hollow shells for the manufacture of tubing has been accomplished by extruding forged and drilled billets at 2150°F with glass as a lubricant. Successful extru- sions have been made on commercial presses at extrusion ratios of up to 14:1. Forging recoveries of up to 90% of the ingot weight have been re- ported by one vendor. 13-3.3 Cold-forming. In the fully annealed condition, the ductility of the alloy ranges between 40 and 50% elongation for a 2-in. gage length. Thus, cold-forming operations, such as tube reducing, rolling, and wire drawing, can be accomplished with normal production schedules. The ef- fects of cold-forming on the ultimate tensile strength, yield strength, and elongation are shown in Fig. 13-6. Forgeability studies have shown that variations in the carbon content have an effect on the cold-forming of the alloy. Slight variations of other components, in general, have no significant effects. The solid solubility of carbon in the alloy is about 0.01%. Carbon present in excess of this amount precipitates as discrete particles of (Ni,Mo)sC throughout the matrix; the particles dissolve sparingly even at the high annealing temperature of 2150°F. Thus cold-working of the alloy causes these particles to align in the direction of elongation and, if they are present in sufficient quantity, they form continuous stringers of carbides. The lines of weakness caused by the stringers are sufficient to propagate longitudinal fractures in tubular products during fabrication. The upper limit of the carbon content for tubing is about 0.10%, and for other products it appears to be greater than 0.20%. The carbon content of the alloy is controllable to about 0.02% in the range below 0.10%. | 13-3] FABRICATION OF INOR-8 605 (x103) 25o1||[||||1oo 200 |— Ultimate Tensile Strength 80 60 0.2% Offset Yield Strength Strength (psi) 100 Elongation 50 Per Cent Elongation (2-in. Gage Length) | | l L1 L1 | 0 0 10 20 30 40 50 60 70 80 90 Reduction in thickness (%) Fig. 13-6. Work hardening curves for INOR-8 annealed 1 hr at 2150°F before reduction. 13-3.4 Welding. The parts of the reactor system are joined by welding, and therefore the integrity of the system is in large measure dependent on the reliability of the welds. During the welding of thick sections, the material will be subjected to a high degree of restraint, and consequently both the base metal and the weld metal must not be susceptible to cracking, embrittlement, or other undesirable features. Extensive tests of weld specimens have been made. The circular-groove test, which accurately predicted the weldability of conventional materials with known welding characteristics, was found to give reliable results for nickel-base alloys. In the circular-groove test, an inert-gas-shielded tung- sten-arc weld pass is made by fusion welding (i.e., the weld metal contains no filler metal) in a circular groove machined into a plate of the base metal. The presence or absence of cracks in the weld metal is then observed. Test samples of two heats of INOR-8 alloys, together with samples of four other alloys for comparison, are shown in Fig. 13-7. As may be seen, the restraint of the weld metal caused complete circumferential cracking in INOR-8 heat 8284, which contained 0.04%, B, whereas there are no cracks in INOR-8 heat 30-38, which differed from heat 8284 primarily in the 606 MOLTEN-SALT REACTOR CONSTRUCTION MATERIALS [CHAP. 13 NOR 8 HEAT 30-38 STAINLESS STEEL F1a. 13-7. Circular-groove tests of weld metal cracking. WELD DEPOSIT 7/ T~ s 12 in o // N 777 . ’ /o in. 117 2! [ I/\4 . BACKING STRIP —~ /o-in-THICK PLATE F1a. 13-8. Weld test plate design showing method of obtaining specimen. 13-3] FABRICATION OF INOR-8 607 Fic. 13-9. Weld in slot of vacuum-melted ingot. absence of boron. Two other INOR-8 heats that did not contain boron similarly did not crack when subjected to the circular-groove test. In order to further study the effect of boron in INOR-8 heats, several 3-b vacuum-melted ingots with nominal boron contents of up to 0.10% were prepared, slotted, and welded as shown in Fig. 13-9. All ingots with 0.029, or more boron cracked in this test. A procedure specification for the welding of INOR-8 tubing is available that is based on the results of these cracking tests and examinations of numerous successful welds. The integrity of a joint, which is a measure of the quality of a weld, is determined through visual, radiographic, and metallographic examinations and mechanical tests at room and service temperatures. It has been established through such examinations and tests that sound joints can be made in INOR-8 tubing that contains less than 0.029, boron. Weld test plates of the type shown in Fig. 13-8 have also been used for studying the mechanical properties of welded joints. Such test plates were side-bend tested in the apparatus illustrated in Fig. 13-10. The results of the tests, presented in Table 13-3, indicate excellent weld metal ductility. For example, the ductility of heat M—5 material is greater than 409, at temperatures up to and including 1500°F. 608 MOLTEN-SALT REACTOR CONSTRUCTION MATERIALS [CHAP. 13 Hydraulic Cylinder Specimen Thermocouple —~—— Bend Quartz Rod e Specimen Control Thermocouple ¥ a - e e ne Ay el ¥ dnfonv e : T S P > TR ° ) }:;A 1\74 Heating Element— Connections (220v or 110v) £ - Quartz Mirror 10123435 Scale in Inches F1a. 13-10. Apparatus for bend tests at high temperatures. 13-3.5 Brazing. Welded and back-brazed tube-to-tube sheet joints are normally used in the fabrication of heat exchangers for molten salt service. The back-brazing operation serves to remove the notch inherent in con- ventional tube-to-tube sheet joints, and the braze material minimizes the possibility of leakage through a weld failure that might be created by ther- mal stresses in service. The nickel-base brazing alloys listed in Table 13—4 have been shown to be satisfactory in contact with the salt mixture LiF-KF-NaF-UF. in tests conducted at 1500°F for 100 hr. Further, two precious metal-base brazing alloys, 829, Au-189, Ni and 809, Au-209, Cu, were unattacked in the LiF-KF-NaF-UF4 salt after 2000 hr at 1200°F. These two precious 609 FABRICATION OF INOR-8 13-3] -porgadde oBID 9S1Y OWI} je I9qY I9INO0 38 JBY} ST PAPIOIAI uoryesuo[y | ‘poaeadde YoBId JSIY YOTYM 9B 38} ST PIPI0ISL 9[due pudd, 8 CT or< 06 < 8 T 8 CT 0r< 06 <. 00¢T or< 06< 8 CT or< 06 < or< 06 < CT 0¢ v < 06< 00€T or< 06< or< 06 < o< 06 < 00ct or< 06 < v < 06 < v < 06 < 00TT or< 06 < v < 06< o< 06 < wooy % “ar/1 dap % “ury/1 3ap % “ar /1 dap ul uorye3uo[yq ‘o[3ue puag Ul uorye3uory ‘o[3ue puag ul fuor)e3uo[y £ 9[8ue pusg 1 ampspfimmaop [ouoouy (61-dS ¥¢°H) 8-4ONI (S-IN ¥¢9H) 8-4ONI 401 [830W JO[[LA SATINVS TANOON] ANV 8- ONI QEATI \\-SY 40 SIS ], ANA{-EAIg 40 SIINSAY C—QT WIdV], 610 MOLTEN-SALT REACTOR CONSTRUCTION MATERIALS TABLE 134 NickEL-BASE BraziNnGg ArrLoys ror USE IN HeAaT EXCHANGER FABRICATION Brazing alloy content, w/o Components Alloy 52 Alloy 91 Alloy 93 Nickel 91.2 91.3 93.3 Silicon 4.5 4.5 3.5 Boron 2.9 2.9 1.9 Iron and carbon Balance Balance Balance [cHAP. 13 metal alloys were also tested in the LiF-BeFs—UF4 mixture and again were not attacked. 13-3.6 Nondestructive testing. An ultrasonic inspection technique is available for the detection of flaws in plate, piping, and tubing. The water- immersed pulse-echo ultrasound equipment has been adapted to high- speed use. KEddy-current, dye-penetrant, and radiographic inspection methods are also used as required. The inspected materials have included Inconel, austenitic stainless steel, INOR-8, and the Hastelloy and other nickel-molybdenum-base alloys. Methods are being developed for the nondestructive testing of weld- ments during initial construction and after replacement by remote means in a high-intensity radiation field, such as that which will be present if malntenance work is required after operation of a molten-salt reactor. The ultrasonic technique appears to be best suited to semiautomatic and remote operation and of any of the applicable methods, it will probably be the least affected by radiation. Studies have indicated that the diffi- culties encountered due to the high ultrasonic attenuation of the weld structures in the ultrasonic inspection of Inconel welds and welds of some of the austenitic stainless steels are not present in the inspection of INOR-8 welds. In addition, the troublesome large variations in ultrasonic attenua- tion common to Inconel and austenitic stainless steel welds are less severe in INOR-8 welds. The mechanical equipment designed for the remote welding operation will be useful for the inspection operation. In the routine inspection of reactor-grade construction materials, a tube, pipe, plate, or rod is rejected if a void is detected that is larger than 59 of the thickness of the part being inspected. In the inspection of a weld, the integrity of the weld must be better than 959, of that of the base metal. 13-4] MECHANICAL AND THERMAL PROPERTIES OF INOR-S8 611 Typical rejection rates for Inconel and INOR-8 are given below: Rejection rate (9)) Item Inconel INOR-8 Tubing 17 20 Pipe 12 14 Plate 8 8 Rod 5 5 Welds 14 14 The rejection rates for INOR-8 are expected to decline as more experience is gained in fabrication. 13—4. M ECHANICAL AND THERMAL ProrPeErRTIES OF INOR-8 13—4.1 Elasticity. A typical stress-strain curve for INOR-8 at 1200°F is shown in Fig. 13-11. Data from similar curves obtained from tests at room temperature up to 1400°F are summarized in Fig. 13-12 to show changes in tensile strength, yield strength, and ductility as a function of temperature. The temperature dependence of the Young’s modulus of this material is illustrated in Fig. 13-13. 13—4.2 Plasticity. A series of relaxation tests of INOR-8 at 1200 and 1300°F has indicated that creep will be an important design consideration for reactors operating in this temperature range. The rate at which the stress must be relaxed in order to maintain a constant elastic strain at 1300°F is shown in Fig. 13-14, and similar data.for 1200°F are presented in Fig. 13-15. The time lapse before the material becomes plastic 1s about 1 hr at 1300°F and about 10 hr at 1200°F. The time period during which the material behaves elastically becomes much longer at lower tempera- tures, and below some temperature, as yet undetermined, the metal will continue to behave elastically indefinitely. It is possible to summarize the creep data by comparing the times to 1.09, total strain, as a function of stress, in the data shown in Fig. 13-16. The reproducibility of creep data for this material is indicated by the separate curves shown in Fig. 13-17. It may be seen that quite good corre- lation between the creep curves is obtained at the lower stress values. Some scatter in time to rupture occurs at 25,000 psi, a stress which corre- sponds to the 0.29, offset yield strength at this temperature. Such scatter is to be expected at this high stress level. 612 MOLTEN-SALT REACTOR CONSTRUCTION MATERIALS 35,000 | 7 30,000 / 25,000 |— 20,000 — Stress (psi) 15,000 — 10,000 — 5,000 — o | | / 1 | | | 0.0006 0.0012 0.0018 0.0024 0.0030 0.0036 Elongation (in./in.) Fia. 13-11. Stress-strain relationships for INOR-8 at 1200°F. Initial slope (rep- resented by dashed line at left) is equivalent to a static modulus of elasticity in tension of 25,200,000 psi. The dashed line at right is the curve for plastic deforma- tion of 0.002 in/in; its intersection with the stress-strain curve indicates a yield strength of 25,800 psi for 0.29, offset. Ultimate tensile strength, 73,895 psi; gage length, 3.25 in.; material used was from heat 3038. [cHAP. 13 The tensile strengths of several metals are compared with the tensile strength of INOR-8 at 1300°F in the following tabulation, and the creep properties of the several alloys at 1.09, strain are compared in Fig. 13—18. Material Tensile strength at 1300°F, psi 18-8 stainless steel Cr-Mo steel (69, Cr) Hastelloy B Hastelloy C Inconel INOR-8 40,000 20,000 70,000 100,000 60,000 65,000 The test results indicate that the elastic and plastic strengths of INOR-8 are near the top of the range of strength properties of the several alloys 13-4] MECHANICAL AND THERMAL PROPERTIES OF INOR-8 613 (x103) 110 e | I I [ | I | | - Tensile Strength 100 — 3 90 — & 80 — 8 g 70 —R . o Elongation, % e @ £ g 5 A 60 g & 50 —H g 40 — < Yield Strength 30 }— — 9 20 | | | I | | ] | & N 0 2 4 6 8 10 12 14 16 18x102) Temperature, °F F1gc. 13-12. Tensile properties of INOR-8 as a function of temperature. (x109) 34 30 28 |— 26 — 24 |- Young’s Modulus (psi) 22 — ol L 1 1 1 1 0 200 400 600 800 1000 1200 1400 Temperature (°F) F1a. 13-13. Young’s modulus for INOR-8. 614 MOLTEN-SALT REACTOR CONSTRUCTION MATERIALS [cuAP. 13 (x103) 28 T T TTTTT T TTTTTT T T 171111 0.2% Constant Strain 16 — 0.1% Constant Strain — 0.05% Constant Strain Stress (psi) Lt e EER Lt 0.1 0.2 0.5 1.0 2 5 10 20 50 100 Time (hr) Fic. 13-14. Relaxation of INOR-8 at 1300°F at various constant strains. (x103) 28 l T T T 11711 l T T T 171711 l T T 1111 24 — 0.1% Constant Strain 20 — o [ 0.05% Constant Strain Stress (psi) ~ I I * Discontinued Test 0 | L | I EEEE | L1t 1 2 5 10 20 50 100 200 500 1000 Time (hr) Fic. 13-15. Relaxation of INOR-8 at 1200°F at various constant strains. 13-4] MECHANICAL AND THERMAL PROPERTIES OF INOR-8 615 109 - orreir o rreer v P eyt v b T ri=g 5 /1100°F = 2 T TN ]04 — ~ = — — ~ = -~ . FE 1300°F/ ~o = ‘% 5 — ~ ~ - § e —— s 2 — W Extrapolated ~ ]03 — :~ 5 |— = | 1yr 10 yr L | " 102 L e b b Preree L Eriime b el 1 10 100 1000 10,000 100,000 Time (hr) to 1% Strain Fic. 13-16. Creep data for INOR-S8. 100 = T T T TTTTT T T T 1T - - ¥* —] 50 - — — Results for Three Samples - Stressed at 25,000 psi 20— — 2 L 1.0 — o — — & |- _] 0.5 [ — B Results for Two Samples N - Stressed at 20,000 psi — 0.2 — o A 11 L L L Ll 10 20 50 100 200 500 1000 2000 Time (hr) F1g. 13-17. Creep-rupture data for INOR-S8. 616 MOLTEN-SALT REACTOR CONSTRUCTION MATERIALS [cHAP. 13 R b DT I EEE — — — ed — — Stress (psi) IRERERN L 1i1ll] p— () I 2 L1 it [ L 111l L 1 111 10 20 50 100 200 500 1000 2000 5000 10,000 Time to 1% Strain at 1300°F (hr) F1c. 13-18. Comparison of the creep properties of several alloys. commonly considered for high-temperature use. Since INOR-8 was de- signed to avoid the defects inherent in these other metals, it 1s apparent that the undesirable aspects have been eliminated without any serious loss in strength. 13—4.3 Aging characteristics. Numerous secondary phases that are ca- pable of embrittling a nickel-base alloy can exist in the Ni-Mo-Cr-Fe-C system, but no brittle phase exists if the alloy contains less than 209, Mo, 89, Cr, and 59, Fe. INOR-8, which contains only 15 to 189, Mo, consists principally of two phases: the nickel-rich solid solution and a complex car- bide with the approximate composition (Ni, Mo)¢C. Studies of the effect of the carbides on creep strength have shown that the highest strength exists when a continuous network of carbides surrounds the grains. Tests have shown that carbide precipitation does not cause significant embrittle- ment at temperatures up to 1480°F. Aging for 500 hr at various tempera- tures, as shown in Fig. 13—-19, improves the tensile properties of the alloy. The tensile properties at room temperature, as shown in Table 13-5, are virtually unaffected by aging. INOR-8 13-4] MECHANICAL AND THERMAL PROPERTIES OF (x103) 110 | ! | 100 |— - = 90 & = 80 @ g 70 A 2 60 § Annealed 1 hr at 2100°F =0 - Aged 500 hr at Test Temperature o 40 — — 0.2% Yield Point 30 w————__—_——_-fi 60 —_— I | 50 — 2 < 40 Elongati — ® 30— o c 2 20 - — Wl 10 — — 0 l . | 1000 1100 1200 1300 1400 Test Temperature (°F) 617 F1c. 13-19. Effect of aging on high-temperature tensile properties of INOR-8. TABLE 13-5 REesuLts oF RooM-TEMPERATURE EMBRITTLEMENT TESTS oF INOR-8 Ultimate tensile | Yield point at | Elongation, Heat treatment strength, psi 0.29, offset, psi % Annealed* 114,400 44,700 50 Annealed and aged 500 hr at 1000°F 112,000 42 500 53 Annealed and aged 500 hr at 1100°F 112,600 44,000 51 Annealed and aged 500 hr at 1200°F 112,300 44,700 51 Annealed and aged 500 hr at 1300°F 112,000 44 500 49 Annealed and aged 500 hr at 1400°F 112,400 43,900 50 *0.045-in. sheet, annealed 1 hr at 2100°F and tested at a strain rate of 0.05 in/min. 6018 MOLTEN-SALT REACTOR CONSTRUCTION MATERIALS [cHAP. 13 13—4.4 Thermal conductivity and coefficient of linear thermal expansion. Values of the thermal conductivity and coefficient of linear thermal ex- pansion are given in Tables 13—6 and 13-7. TABLE 13-6 CoMPARISON OF THERMAL Conpuctivity VALUES ForR INOR-8 AND INCONEL AT SEVERAL TEMPERATURES Thermal conductivity, Btu/(ft2)(sec) (°F/ft) Temperature, °F INOR-S8 Inconel 212 5.56 9.44 392 6.77 9.92 572 " 11.16 10.40 752 12.10 10.89 933 14.27 11.61 1112 16.21 12.10 1292 18.15 12.58 TaBLE 13-7 CoErFICIENT OF LINEAR ExpansioN or INOR-8 FOR SEVERAL TEMPERATURE RANGES Temperature range, °F Coeffimenii;n(;f(ilrllx;??;‘ )expansmn, X106 70-400 5 76 70-600 6.93 70-800 6. 58 70-1000 6.89 70-1600 8 10 70-1800 8 39 13-5] OXIDATION RESISTANCE 619 13-5. OXIDATION RESISTANCE The oxidation resistance of nickel-molybdenum alloys depends on the service temperature, the temperature cycle, the molybdenum content, and the chromium content. The oxidation rate of the binary nickel-molybdenum alloy passes through a maximum for the alloy containing 159, Mo, and the scale formed by the oxidation is NiMoO4 and NiO. Upon thermal cycling from above 1400°F to below 660°F, the NiMoO4 undergoes a phase trans- formation which causes the protective scale on the oxidized metal to spall. Subsequent temperature cycles then result in an accelerated oxidation rate. Similarly, the oxidation rate of nickel-molybdenum alloys containing chromium passes through a maximum for alloys containing between 2 and 69, Cr. Alloys containing more than 69, Cr are insensitive to thermal cycling and the molybdenum content because the oxide scale is pre- dominantly stable CroOs. An abrupt decrease, by a factor of about 40, in the oxidation rate at 1800°F is observed when the chromium content is increased from 5.9 to 6.29. The oxidation resistance of INOR-8 is excellent, and continuous opera- tion at temperatures up to 1800°F is feasible. Intermittent use at tempera- tures as high as 1900°F could be tolerated. For temperatures up to 1200°F, the oxidation rate is not measurable; it is essentially nonexistent after 1000 hr of exposure in static air. It is estimated that oxidation of 0.001 to 0.002 in. would occur in 100,000 hr of operation at 1200°F. The effect of temperature on the oxidation rate of the alloy is shown in Table 13-8. TABLE 13-8 OxIDATION RATE oF INOR-8 AT VarioUs TEMPERATURES* . : o Test temperature, Weight gain, mg/cm Shape of °F rate curve In 100 hr In 1000 hr 1200 0.00 0.00 Cubic or logarithmic 1600 0.25 0.67¢ Cubic 1800 0.48 1.5% Parabolic 1900 0.52 2.07 Parabolic 2000 2.70 28.27 Linear *3.7 mg/cm2 = 0.001 in. of oxidation. tExtrapolated from data obtained after 170 hr at temperature. 620 MOLTEN-SALT REACTOR CONSTRUCTION MATERIALS [cHAP. 13 Fia. 13-20. Components of a duplex heat exchanger fabricated of Inconel clad with type-316 stainless steel. 13-6. FABRICATION OF A DUPLEX TuBING HEAT EXCHANGER The compatibility of INOR-8 and sodium is adequate in the temperature range presently contemplated for molten-salt reactor heat-exchanger oper- ation. At higher temperatures, mass transfer could become a problem, and therefore the fabrication of duplex tubing has been investigated. Satis- factory duplex tubing has been made that consists of Inconel clad with type—316 stainless steel, and components for a duplex heat exchanger have been fabricated, as shown in Fig. 13-20. The fabrication of duplex tubing is accomplished by coextrusion of billets of the two alloys. The high temperature and pressure used result in the formation of a metallurgical bond between the two alloys. In sub- sequent reduction steps the bonded composite behaves as one material. The ratios of the alloys that comprise the composite are controllable to within 39,,. The uniformity and bond integrity obtained in this process are illustrated in Fig. 13-21. The problem of welding INOR-8-stainless steel duplex tubing is being studied. Experiments have indicated that proper selection of alloy ratios and weld design will assure welds that will be satisfactory in high-tempera- ture service. To determine whether interdiffusion of the alloys would result in a con- tinuous brittle layer at the interface, tests were made in the temperature range 1300 to 1800°F. As expected, a new phase appeared at the interface between INOR—-8 and the stainless steel which increased in depth along the grain boundaries with increases in the temperature. The interface of a duplex sheet held at 1300°F for 500 hr is shown in Fig. 13-22. Tests of this sheet showed an ultimate tensile strength of 94,400 psi, a 0.29, offset yield strength of 36,800 psi, and an elongation of 519,. Creep tests of the 13-6] FABRICATION OF A DUPLEX TUBING HEAT EXCHANGER 621 Ly R F1c. 13-21. Duplex tubing consisting of Inconel over type-316 stainless steel. Etchant: glyceria regia. sheet showed that the diffusion resulted in an increase in the creep re- sistance with no significant loss of ductility. Thus no major difficulties would be expected in the construction of an INOR-8-stainless steel heat exchanger. The construction experience thus far has involved only the 20-tube heat exchanger shown in Fig. 13-20. 622 MOLTEN-SALT REACTOR CONSTRUCTION MATERIALS [cEAP. 13 Unstressed Stainless Steel Fre. 13-22. Unstressed and stressed specimens of INOR-8 clad with type-316 stainless steel after 500 hr at 1300°F. Etchant: electrolytic H2SO4 (29 solution). (100X%) A o1 NN g F1a. 13-23. CCN graphite (A) before and (B) after exposure for 1000 hr to NaF- ZrF4+-UF4 (50-46-4 mole 9) at 1300°F as an insert in the hot leg of a thermal- convection loop. Nominal bulk density of graphite specimen: 1.9 g/cm3. 13-8] GRAPHITE WITH MOLTEN SALTS AND NICKEL-BASE ALLOYS 623 13—-7. AvaiLaBiLiTy oF INOR-8 Two production heats of INOR-8 of 10,000 1b each and numerous smaller heats of up to 5000 Ib have been melted and fabricated into various shapes by normal production methods. Evaluation of these commercial products has shown them to have properties similar to those of the laboratory heats prepared for material selection. Purchase orders are filled by the vendors in one to six months, and the costs range from $2.00 per pound in ingot form to $10.00 per pound for cold-drawn welding wire. The costs of tubing, plate, and bar products depend to a large extent on the specifications of the finished products. 13-8. CoMPATIBILITY OF GRAPHITE WITH MOLTEN SALTS AND NickeL-BAsE ALLOYS If graphite could be used as a moderator in direct contact with a molten salt, it would make possible a molten-salt reactor with a breeding ratio in excess of one (see Chapter 14). Problems that might restrict the usefulness of this approach are possible reactions of graphite and the fuel salt, pene- tration of the pores of the graphite by the fuel, and carburization of the nickel-alloy container. Many molten fluoride salts have been melted and handled in graphite crucibles, and in these short-term uses the graphite is inert to the salt. Tests at temperatures up to 1800°F with the ternary salt mixture NaF-ZrF4+UF4 gave no indication of the decomposition of the fluoride and no gas evolution so long as the graphite was free from a silicon im- purity. Longer-time tests of graphite immersed in fluoride salts have shown greater indications of penetration of the graphite by salts, and it must be assumed that the salt will eventually penetrate the available pores in the graphite. The “impermeable’’ grades of graphite available experimentally show greatly reduced penetration, and a sample of high-density, bonded, natural graphite (Degussa) showed very little penetration. Although quantitative figures are not available, it is likely that the extent of pene- tration of “‘impermeable’ graphite grades can be tolerated. Although these penetration tests showed no visible effects of attack of the graphite by the salt, analyses of the salt for carbon showed that at 1500°F more than 1% carbon may be picked up in 100 hr. The carbon pickup appears to be sensitive to temperature, however, inasmuch as only 0.025% carbon was found in the salt after a 1000-hr exposure at 1300°F. In some instances coatings have been found on the graphite after ex- posure to the salt in Inconel containers, as illustrated in Fig. 13-23. A cross section through the coating is shown in Fig. 13-24. The coating was 624 MOLTEN-SALT REACTOR CONSTRUCTION MATERIALS [CHAP. 13 F1c. 13-24. Cross sections of samples shown in Fig. 13-23. (A) Before exposure; (B) after exposure. Note the thin film of Cr3sCz on the surface in (B). The black areas in (A) are pores. In (B) the pores are filled with salt. (100X) found to be nearly pure chromium carbide, CrsCs. The source of the chromium was the Inconel container. - In the tests run thus far, no positive indication has been found of car- burization of the nickel-alloy containers exposed to molten salts and graphite at the temperatures at present contemplated for power reactors (<1300°F). The carburization effect seems to be quite temperature sensi- tive, however, since tests at 1500°F showed carburization of Hastelloy B to a depth of 0.003 in. in 500 hr of exposure to NaF-ZrF+UF4 containing graphite. A test of Inconel and graphite in a thermal-convection loop in which the maximum bulk temperature of the fluoride salt was 1500°F gave a maximum carburization depth of 0.05 in. in 500 hr. In this case, however, the temperature of the metal-salt interface where the carburization oc- curred was considerably higher than 1500°F, probably about 1650°F. A mixture of sodium and graphite is known to be a good carburizing agent, and tests with it have confirmed the large effect of temperature on the carburization of both Inconel and INOR-8, as shown in Table 13-9. 13-10] SUMMARY OF MATERIAL PROBLEMS 625 TABLE 13-9 ErrFEcT OF TEMPERATURE ON CARBURIZATION OF INcoNEL AND INOR-8 18y 100 HR Alloy Temperature, °F Depth of carburization, in. Inconel 1500 0.009 1200 0 INOR-8 1500 0.010 1200 0 Many additional tests are being performed with a variety of molten fluoride salts to measure both penetration of the graphite and carburization of INOR-8. The effects of carburization on the mechanical properties will be determined. 13-9. MATERIALS FOR VALVE SEATS AND BEARING SURFACES Nearly all metals, alloys, and hard-facing materials tend to undergo solid-phase bonding when held together under pressure in molten fluoride salts at temperatures above 1000°F. Such bonding tends to make the startup of hydrodynamic bearings difficult or impossible, and it reduces the chance of opening a valve that has been closed for any length of time. Screening tests in a search for nonbonding materials that will stand up under the molten salt environment have indicated that the most promising materials are TiC-Ni and WC—-Co types of cermets with nickel or cobalt contents of less than 35 w/o, tungsten, and molybdenum. The tests, in general, have been of less than 1000-hr duration, so the useful lives of these materials have not yet been determined. 13-10. SUMMARY OF MATERIAL PROBLEMS Although much experimental work remains to be done before the con- struction of a complete power reactor system can begin, it is apparent that considerable progress has been achieved in solving the material problems of the reactor core. A strong, stable, and corrosion-resistant alloy with good welding and forming characteristics is available. Production tech- niques have been developed, and the alloy has been produced in com- mercial quantities by several alloy vendors. Finally, it appears that even at the peak operating temperature, no serious effect on the alloy occurs when the molten salt it contains is in direct contact with graphite. CHAPTER 14 NUCLEAR ASPECTS OF MOLTEN-SALT REACTORS* The ability of certain molten salts to dissolve uranium and thorium. salts in quantities of reactor interest made possible the consideration of fluid- fueled reactors with thorium in the fuel, without the danger of nuclear ac- cidents as a result of the settling of a slurry. This additional degree .of freedom has been exploited in the study of molten-salt reactors. Mixtures of the fluorides of alkali metals and zirconium or beryllium, as discussed in Chapter 12, possess the most desirable combination of low neutron absorption, high solubility of uranium and thorium compounds, chemical inertness at high temperatures, and thermal and radiation sta- bility. The following comparison of the capture cross sections of the alkali metals reveals that Li? containing 0.019], Li® has a cross section at 0.0795 ev and 1150°F that is a factor of 4 lower than that of sodium, which also has a relatively low cross section: Element Cross section, barns Li? (containing 0.019%, Li®) 0.073 Sodium 0.290 Potassium 1.13 Rubidium 0.401 Cesium 29 The capture cross section of beryllium is also satisfactorily low at all neutron energies, and therefore mixtures of LiF and BeF2, which have satisfactory melting points, viscosities, and solubilities for UF4 and ThF 4, were selected for investigation in the reactor physics study. Mixtures of NaF, ZrF4, and UF4 were studied previously, and such a fuel was successfully used in the Aircraft Reactor Experiment (see Chap- ters 12 and 16). Inconel was shown to be reasonably resistant to corrosion by this mixture at- 1500°F, and there is reason to expect that Inconel equipment would have a life of at least several years at 1200°F. As a fuel for a central-station power reactor, however, the NaF-ZrF4 system has several serious disadvantages. The sodium capture cross section is less favorable than that of Li17?. More important, recent data [1] indicate that the capture cross section of zirconium is quite high in the epithermal and intermediate neutron energy ranges. In comparison with the LiF-BeF2 system, the NaF-ZrF4 system has inferior heat-transfer characteristics. *By L. G. Alexander. 626 NUCLEAR ASPECTS OF MOLTEN-SALT REACTORS 027 Finally, the INOR alloys (see Chapter 13) show promise of being as resistant to the beryllium salts as to the zirconium salts, and therefore there is no compelling reason for selecting the NaF-ZrF4 system. Reactor calculations were performed by means of the Univac* program Ocusol [2], a modification of the Eyewash program [3], and the Oraclef program Sorghum. Ocusol is a 31-group, multiregion, spherically symmet- ric, age-diffusion code. The group-averaged cross sections for the various elements of interest that were used were based on the latest available data [4]. Where data were lacking, reasonable interpolations based on resonance theory were made. The estimated cross sections were made to agree with measured resonance integrals where available. Saturation and Doppler broadening of the resonances in thorium as a function of concentration were estimated. Inelastic scattering in thorium and fluorine was taken into account crudely by adjusting the value of £o;; however, the Ocusol code does not provide for group skipping or anisotropy of scattering. Sorghum is a 31-group, two-region, zero-dimensional, burnout code. The group-diffusion equations were integrated over the core to remove the spatial dependency. The spectrum was computed, in terms of a space-averaged group flux, from group scattering and leakage parameters taken from an Ocusol calculation. A critical calculation requires about 1 min on the Oracle; changes in concentration of 14 elements during a specified time can then be computed in about 1 sec. The major assumption involved is that the group scattering and leakage probabilities do not change appreciably with changes in core composition as burnup progresses. This assumption has been verified to a satisfactory degree of approximation. The molten salts may be used as homogeneous moderators or simply as fuel carriers in heterogeneous reactors. Although, as discussed below, graphite-moderated heterogeneous reactors have certain potential advan- tages, their technical feasibility depends upon the compatibility of fuel, graphite, and metal, which has not as yet been established. For this rea- son, the homogeneous reactors, although inferior in nuclear performance, have been given greatest attention. A preliminary study indicated that if the integrity of the core vessel could be guaranteed, the nuclear economy of two-region reactors would probably be superior to that of bare and reflected one-region reactors. The two-region reactors were, accordingly, studied in detail. Although entrance and exit conditions dictate other than a spherical shape, it was necessary, for the calculations, to use a model comprising the following concentric *Universal Automatic Computer at New York University, Institute of Mathe- matics. 1Oak Ridge Automatic Computer and Logical Engine at Oak Ridge National Laboratory. 628 NUCLEAR ASPECTS OF MOLTEN-SALT REACTORS [cHAP. 14 spherical regions: (1) the core, (2) an INOR-8 core vessel 1/3 in. thick, (3) a blanket approximately 2 ft thick, and (4) an INOR-8 reactor vessel 2/3 in. thick. The diameter of the core and the concentration of thorium in the core were selected as independent variables. The primary dependent variables were the critical concentration of the fuel (U235 U233 or Pu239), and the distribution of the neutron absorptions among the various atomic species in the reactor. From these, the critical mass, critical inventory, regeneration ratio, burnup rate, etc. can be readily calculated, as described in the following section. 14-1. HoMmoGENEOUS REACTORS FUELED wiTe U235 While the isotope U233 would be a superior fuel in molten fluoride-salt reactors (see Section 14-2), it is unfortunately not available in quantity. Any realistic appraisal of the immediate capabilities of these reactors must be based on the use of U235, The study of homogeneous reactors was divided into two phases: (1) the mapping of the nuclear characteristics of the initial (i.e., “clean’) states as a function of core diameter and thorium concentration, and (2) the analysis of the subsequent performance of selected initial states with various processing schemes and rates. The detailed results of these studies are given in the following paragraphs. Briefly, it was found that regenera- tion ratios of up to 0.65 can be obtained with moderate investment in U235 (less than 1000 kg) and that, if the fission products are removed (Article 14-1.2) at a rate such that the equilibrium inventory is equal to one year’s production, the regeneration ratio can be maintained above 0.5 for at least 20 years. 14-1.1 Initial states. A complete parametric study of molten fluoride- salt reactors having diameters in the range of 4 to 10 ft and thorium con- centrations in the fuel ranging from O to 1 mole 9, ThF4 was performed. In these reactors, the basic fuel salt (fuel salt No. 1) was a mixture of 31 mole 9, BeF2 and 69 mole 9, LiF, which has a density of about 2.0 g/cc at 1150°F. The core vessel was composed of INOR-8. The blanket fluid (blanket salt No. 1) was a mixture of 25 mole 9, ThF4 and 75 mole 9}, LiF, which has a density of about 4.3 g/cc at 1150°F. In order to shorten the calculations in this series, the reactor vessel was neglected, since the re- sultant error was small. These reactors contained no fission products or nonfissionable isotopes of uranium other than U238, A summary of the results is presented in Table 14-1, in which the neutron balance is presented in terms of neutrons absorbed in a given element per neutron absorbed in U235 (both by fission and the n—y reaction). The sum of the absorptions is therefore equal to 7, the number of neutrons produced by fission per neutron absorbed in fuel. Further, the sum of the 14-1] HOMOGENEOUS REACTORS FUELED WITH U235 629 (x1019) 40 l X\ l 30 |— ‘ \: mole % ThF, In .\ vel Salt 20 -n 1] o _— " P ° 4 O :I U233 Concentration atoms/cm3 e Calculated Y Values / 2 — Interpolations of Data / No ThF4 In Fuel Salt 1 I l 1l 2 4 6 8 10 Core Diameter, ft ) Fig. 14-1. Initial critical concentration of U235 in two-region, homogeneous, molten fluoride-salt reactors. absorptions in U238 and thorium in the fuel, and in thorium in the blanket salt gives directly the regeneration ratio. The losses to other elements are penalties imposed on the regeneration ratio by these poisons; 1.e., if the core vessel could be constructed of some material with a negligible cross sec- tion, the regeneration ratio could be increased by the amount listed for capture in the core vessel. The inventories in these reactors depend in part on the volume of the fuel in the pipes, pumps, and heat exchangers in the external portion of the fuel circuit. The inventories listed in Table 14-1 are for systems having a volume of 339 ft3 external to the core, which corresponds approximately to a power level of 600 Mw of heat. In these calculations it was assumed that the heat was transferred to an intermediate coolant composed of the fluorides of Li, Be, and Na before being transferred to sodium metal. In more recent designs (see Chapter 17), this intermediate salt loop has been replaced by a sodium loop, and the external volumes are somewhat less because of the improved equipment design and layout. Critical concentration, mass, inventory, and regeneration ratio. The data in Table 14—1 are more easily comprehended in the form of graphs, such as Fig. 14-1, which presents the critical concentration in these reactors as a function of core diameter and thorium concentration in the fuel salt. The data points represent calculated values, and the lines are reasonable interpolations. The maximum concentration calculated, about 35 X 10'? [cHAP. 14 NUCLEAR ASPECTS OF MOLTEN-SALT REACTORS 630 ponujuo9 9¢°0 290 19°0 09°0 8G 0 L8°0 6S°0 OI}EB1 UOI}BISUSSY 820 g0%¥ 0 €70 €% 0 &% 0 6¢°0 gy 0 © ‘01981 UOISSY-03-armyden A-u Ge 0%0°0 L8°0 gg'0 L8°0 g9 2S00 % ‘suolssy [ewIdy, 810 Sg¥ 043 8GT S0T L ST 043 A5 ‘ABISUD UOISSY UBIPIJA 26°1 PLT gL' 1 gL' 1 gL' 1 LLT gL' 1 L ‘pre1d wonnaN g18S "0 S06¢ 0 1€0% 0 11370 915% 0 608S 0 8¥¥S 0 Jes 9quB[q Ul 4, 8L81°0 P191°0 68210 2€80°0 J[es [ong Ul Y, 6$20°0 g1%0°'0 1€%0°0 1S%0°0 £9%0° 0 92700 08500 }[eS [0 UL geg(] . L€10°0 0%10°0 I710°0 ev10°0 L¥10°0 9S10°0 62200 a3ey8o] 79100 06000 €600 0 86000 8010 0 8€10°0 8210°0 J[es jexue[q Ul J-I] 8Se1°0 g0v0°0 9€%0 0 ¢8%0 0 L1S0°0 8%80°0 09500 [9S89A 910)) 18610 89500 I18S0°0 70900 £890°0 1280°0 19500 Jes [eng W J-IT-og 63220 6S6% 0 G86% 0 700 0 9663 0 G185 0 L1630 (A-1) ¢z} ILLL°0 1700 ST0L°0 9669 0 $00L°0 G8IL 0 €20L°0 (SUOISSY) gez(] isorger uorpdiosqe uoIjnoN 88T 08¥1 0€€T 0811 168 109 08ET | geall JO 3 ‘AIojuaAur [BONLID) 0°L¥ erd Q13 e8I 44! 0'18 4! ggg[]l JO 3 ‘sseur [BOILI) 08 '€ g eg 0°08 9°63 103 g 11 8°6¢ +KYISUSD WOYE gez() L01°0 8€6 "0 6¥3°0 13L°0 1950 81€°0 g6 0 % O0W ‘}[BS [0 UI gez() 0 I GL 0 g0 6% 0 0 0 % erou ‘s [ony ut ¥ 4T, 9 g G ¢ G G b 1] ‘I830WBIP 910)) L 9 g v e 4 I IoqUINU 98B)) "¢} 6EE 19WN[OA [9N] [BUINIXH cez[) HIIM QITINJ SHOIOVAY LTVS-HAIY0ONT] NILTOA "(389Y) MIA 009 :Iomod [BIO, ‘AT % orouw G2, + Y iYL, % 9[0W GZ : 1 "ON 3[®S 19yus[g YAUL + YA0 + VT % oo 69 + e g9g %, o[owr 1¢ i1 "ON BS [on] ‘SNOANTDONO ‘NOIDIAY-OMJ, 40 SOIILSIHELOVEVH) GVITOAN ALVLG-TVILIN] I-$1 @1av], 631 HOMOGENEOUS REACTORS FUELED WITH U232 14-1] 09/ (g1-0T X) SWOIVy %@3@8@60 ‘gez[] UI fioQgOde uoJjnau Jod peqaosqe mQO.S.DmZ._.. 2670 Z¥ 0 19°0 19°0 09°0 09°0 19°0 O}l UOTJBISUSSY 920 S% 0 620 gv 0 gy 0 170 9¢°0 0 ‘oryer uotssy-03-ornjded A-u b 66 &g 870 9670 8 el Yo 'sUOISSY [BUIOY], 01°0 [BULIOY, 910 02T 00T 8¢ 9°G A9 ‘A319U0 UOISSY UBIPOIN 961 003G 161 eL'1 eL'1 QL1 28’1 L ‘pretd uornoN £09€ 0 €L0¥ 0 860% 0 20ge 0 8LE€ 0 £89¢ 0 818¥% 0 J[es Jo¥{uB[q Ul YT, S%03 0 192170 8€¥C 0 g¥1% 0 I#81°0 13810 J[es [ong Ul YL, 9610 0 ev10°0 #5200 2s%0 0 L9%0°0 LL%0°0 G280 0 1[8S [0 UI geg() 800 0 #8000 8900 0 91100 6110 0 23100 62100 . ogexea] g110°0 ev10°0 3100 #2000 18000 16000 91100 J[es j9quB(q Ul J-T] geo1 0 16%1°0 19600 8820 0 Ge70 0 g¥30 0 660 0 [9sS9A 2100 9195 0 0¢1¥ 0 08810 1€90°0 6990 0 0LL0°0 2801 0 J[es o0y ut J-1T-°dg 0L0% 0 €661 0 288% 0 9662 0 000€ 0 8165 0 L8930 (A-1) ¢ez) 0€6L "0 L0080 8¥LL°0 700L 0 000Z°0 g80L 0 gveL 0 (SUOISSY) ggz() 1soryer uorydiosqe UOINON P81 01T 0€G 091T 3L6 912 VOb | cee(l JO 39 'AIOJUSAUL [BIMLLT g'18 L'8¥ 962 162 eV 6L1 101 ggz[] JO 3Y ‘sseul [BOILI) LL'G 99T 0¥ G €2 961 e ¢I'8 xK3ISUBD WOIE gez[) 8L0°0 Lv0°0 PIT°0 2990 2550 80% 0 665 0 % 10w “4[88 91y UI geg() Al 0 6% 0 I gL 0 g0 A % 910w ‘}[es [any ul ¥ Y], 8 8 L 9 9 9 9 1} ‘19}9WEIP AI0)) 4! el Gl 11 01T 6 Q I9quUINU 988 (ponunuod) -1 @IV, [cHAP. 14 NUCLEAR ASPECTS OF MOLTEN-SALT REACTORS 632 ‘ggg() UT POQI0sqB uoIinau 19d Poquosqs SUOIINAN | *20/(g1—0T X) SWOIY 29°0 L9°0 29'0 gs'0 2e'0 €9°0 79°0 79°0 OI}8.I UOIJBIoUaSaY ¢e'0 0g°'0 920 %30 12°0 0%'0 9¢°0 0¢°0 P ‘01981 UOISSY-0-01n3de0 A-u 91 0g g 9¢ 99 g g1 63 % ‘suorssy [euay, 981 9S1°0 00I°0 | TeuWmdYy, | [BULISYL, L3 g'g LT°0 A9 ‘A318U9 UOTSSY UBIPIA] €81 06T 61 00'% €0°'% 0L'1 28'1 681 b ‘prard woxynaN Ge8T°0 | €906°0 | 28€C°0 | L192°0 | 1€08°0 |9%e2'0 | #992°0 | 9%0¢ 0 s JoxUBIq Ul YT, 90970 | ¥e€F 0 | 1698°0 | 60%Z 0 GISE°0 | L688°0 | 8F0€°0 }[BS [oN} UI Y, e¥€0°0 |1 9920°0 | 6020°0 | €ST0°0 | 0ZI0'0 | STFO'0 | 8980°0 | Z220°0 }BS [oN] UI gez() ge00°'0 | LS00°0 | 6S00°0 | 09000 [ T1900°0 |%L00°0 | 220000 | 08000 a3exyBo] 6¥00°0 | 0900°0 | €000 |6800°0 |%ITO0 | 22000 | 12000 | 68000 [es JoxuB[q Ul J-T] €960°0 | L6%0°0 | 6990°0 | SI60°0 | I6ZI°0 | €£280°0 | 00S0°0 | 2ZL0°0 [9889A 010)) 90cI°0 | SELT'0 | 98%C°0 | 92280 | €IL8°0 |9¥80°0 | Z0IT°0 | 891 0 Jes [ong ur J-r7-og ¢lS¢'0 | L0€C°0 | 860C°0 | G850 | TLZLT'0 | $983°0 | 829Z2°0 | 62870 (A-0) ¢gz) 8GVL°0 | €69L°0 | G060 | 8TPLO0 | 62280 | 9PIL0 | 298L°0 | 129L°0 (SuoIssy) ¢ez) Jsorjes uorydiosqe uoayna N L89 4 gLe 9.1 11 ¥Z8 ceg 01€ | ¢ez) JO 31 ‘Arojuaaur [Bonyur) L1 863 G91 L01 L9 $9¢ 083 L81 ggz() JO 3 ‘sseul [BALYLID) 8% "L 0S¥ 88 ' 981 QLT T &4 €0°8 L9°¥ xAJISUSD WOIE gez() ¢0% 0 L3170 1800 2S00 €800 6%€° 0 9320 ge1'0 % S10W “)BS [aN] UL gep() I ¢L0 ¢ 0 ¢z'0 0 I ¢L0 ¢'0 % orowt ‘yyes [any ur ¥ gy, 01 01 01 0T 01 8 8 8 3} ‘I9jowBIp 910 2% 12 03 61 81 LT o1 Sl IaquInu 9sB)) (ponunuod) [—HI TIAV], 14-1] HOMOGENEOUS REACTORS FUELED WITH U235 633 0T T1T 1 | l mole % ThF4 In Fuel Salt @ 1 400 |— e Calculated values /— — Interpoly ° a Data o~ \ = 6 300 | — -e @ - / 075 ¢ " .—\—_ /__-— o @ = v T 200 |— ly ] £ o 0.50 ¢ O ~~ o\\ I\ 0.25 100 . ) N ¢ ) ~———No ThF4 In Fuel Salt ol | | | 04 5 6 7 8 9 10 Core Diameter, ft Fic. 14-2. Initial critical masses of U235 in two-region, homogeneous, molten fluoride-salt reactors. atoms of U235 per cubic centimeter of fuel salt, or about 1 mole 9, UFy, is an order of magnitude smaller than the maximum permissible concentra- tion (about 10 mole 9). The corresponding critical masses are graphed in Fig. 14-2. As may be seen, the critical mass is a rather complex function of the diameter and the thorium concentration. The calculated points are shown here also, and the solid lines represent, it is felt, reliable interpolations. The dashed lines were drawn where insufficient numbers of points were calculated to define the curves precisely; however, they are thought to be qualitatively correct. Since reactors having diameters less than 6 ft are not economically attrac- tive, only one case with a 4-ft-diameter core was computed. The critical masses obtained in this study ranged from 40 to 400 kg of U235, However, the critical inventory in the entire fuel cirecuit is of more interest to the reactor designer than is the critical mass. The critical in- ventories corresponding to an external fuel volume of 339 ft3 are therefore shown in Fig. 14-3. Inventories for other external volumes may be com- puted from the relation I=M<1+§DIZ;>’ where D is the core diameter in feet, M is the critical mass taken from Fig. 14-2, V., is the volume of the external system in cubic feet, and I is the inventory in kilograms of U235, The inventories plotted in Fig. 14-3 634 2000 NUCLEAR ASPECTS OF MOLTEN-SALT REACTORS 1000 b l Critical Inventory, kg of U235 W S | l ! mole % ThFy4 In Fuel Salt No ThF4 In Fuel Salt I T 6 8 10 Core Diameter, ft [cHAP. 14 F1g. 14-3. Initial critical inventories of U235 in two-region, homogeneous, molten fluoride-salt reactors. External fuel volume, 339 ft3. 0.8 | | | | | Numbers On Data Points Are Core Diameters In Feet ( ) mole % ThF4 In Fuel Salt 0.7 — — 0 W10 98 N 2 pro | e (o.n)\fi né 0.6 — 70’ Te————— 06 b —] o 1 5 (0.25) 3 5p ¢ .- & 7.,/\No ThF4 In Fuel Salt > [ %10 & 0.5 |— 8/0 — %0 0.4 |— — 0.3 100 | | | | l 0 200 400 600 800 1000 1200 Critical Inventory, kg U235 F1a. 14-4. Initial fuel regeneration in two-region, homogeneous, molten fluoride- salt reactors fueled with U235, Total power, 600 Mw (heat); external fuel volume, 339 ft3; core and blanket salts No. 1. 14-1] HOMOGENEOUS REACTORS FUELED WITH U?3% 635 0.8 | l | ( ) mole % ThF4 In Fuel Salt 0.7 — nEeEy — - 10 — o— = ° 9./(0.75) m s 6 7 (0.50) ' % 0.6 |— ®(0.25) ] .2 B c ® o No ThF 4 In Fuel Salt 205 | — 0.4 |— Numbers On Data Points ] ’ Are Core Diameters In Feet 0.3 | | | 0 200 400 4600 800 Critical Inventory, kg of 235 Fre. 14-5. Maximum initial regeneration ratios in two-region, homogeneous, molten fluoride-salt reactors fueled with U235, Total power, 600 Mw (heat); ex- ternal fuel volume, 339 ft3. | range from slightly above 100 kg in an 8-ft-diameter core with no thorium present to 1500 kg in a 5-ft-diameter core with 1 mole 9, ThF4 present. The optimum combination of core diameter and thorium concentration is, qualitatively, that which minimizes the sum of inventory charges (in- cluding charges on Li?, Be, and Th) and fuel reprocessing costs. The fuel costs are directly related to the regeneration ratio, and this varies in a complex manner with inventory of U235 and thorium concentration, as shown in Fig. 14—4. It may be seen that at a given thorium concentration, the regeneration ratio (with one exception) passes through a maximum as the core diameter is varied between 5 and 10 ft. These maxima increase with increasing thorium concentration, but the inventory values at which they occur also Increase. Plotting the maximum regeneration ratio versus critical inventory generates the curve shown in Fig. 14-9. It may be seen that a small in- vestment in U235 (200 kg) will give a regeneration ratio of 0.58, that 400 kg will give a ratio of 0.66, and that further increases in fuel inventory have little effect. The effects of changes in the compositions of the fuel and blanket salts are indicated in the following description of the results of a series of calcu- lations for which salts with more favorable melting points and viscosities were assumed. The BeFs content was raised to 37 mole % in the fuel salt 636 NUCLEAR ASPECTS OF MOLTEN-SALT REACTORS [cHAP. 14 | | ! mole % ThFy in Fuel Salt Inventory , kg of U235 Core Diameter, f Fia. 14-6. Initial critical inventories of U235 in two-region, homogeneous, molten fluoride-salt reactors. Total power, 600 Mw (heat); external fuel volume, 339 ft3; core and blanket salts No. 2. 0.7 | | l Numbers On Data Points are Core Diameters In Feet 8. o \, \8\ B 7 S 8 .1.\. \. 7 . = \ ® (1) O 0.6 — ® 1 ® ] e o \ 6 \ \‘ > o« ° ° S & 7o (0.50) (0.75) [ 8e (0.25) () mole % ThF4 In Fuel Salt 0.5 | | | | | o 200 400 600 800 1000 1200 Critical Inventory, U235, kg F1G. 14-7. Initial fuel regeneration in two-reg salt reactors fueled with U235, Total 339 ft3; core and blanket salts No. 2 ion, homogeneous, molten fluoride- power, 600 Mw (heat); external fuel volume, 14-1] HOMOGENEOUS REACTORS FUELED WITH U230 637 (fuel salt No. 2), and the blanket composition (blanket salt No. 2) was fixed at 13 mole % ThF4, 16 mole % BeFs, and 71 mole % LiF. Blanket salt No. 2 is a somewhat better reflector than No. 1, and fuel salt No. 2 a somewhat better moderator. As a result, at a given core diameter and thorium concentration in the fuel salt, both the critical concentration and the regeneration ratio are somewhat lower for the No. 2 salts. Reservations concerning the feasibility of constructing and guaranteeing the integrity of core vessels in large sizes (10 ft and over), together with preliminary consideration of inventory charges for large systems, led to the conclusion that a feasible reactor would probably have a core diameter lying in the range between 6 and 8 ft. Accordingly, a parametric study in this range with the No. 2 fuel and blanket salts was performed. In this study the presence of an outer reactor vessel consisting of 2/3 in. of INOR-8 was taken into account. The results are presented in Table 14-2 and Figs. 14-6 and 14-7. In general, the nuclear performance is somewhat better with the No. 2 salt than with the No. 1 salt. Neutron balances and miscellaneous details. The distributions of the neutron captures are given in Tables 14-1 and 14-2, where the relative hardness of the neutron spectrum is indicated by the median fission energies and the percentages of thermal fissions. It may be seen that losses to Li, Be, and F in the fuel salt and to the core vessel are substantial, especially in the more thermal reactors (e.g., Case No. 18). However, in the thermal reactors, losses by radiative capture in U?3° are relatively low. Increasing the hardness decreases losses to salt and core vessel sharply (Case No. 5), but increases the loss to the n—y reaction. It is these opposing trends which account for the complicated relation between regeneration ratio and critical inventory exhibited in Figs. 14—4 and 14-7. The numbers given for capture in the Li and F in the blanket show that these elements are well shielded by the thorium in the blanket, and the leakage values show that leakage from the reactor is less than 0.01 neutron per neutron absorbed in U235 in reactors over 6 ft in diameter. The blanket contributes sub- stantially to the regeneration of fuel, accounting for not less than one-third of the total even in the 10-ft-diameter core containing 1 mole % ThF4. Effect of substitution of sodium for L17. In the event that Li7 should prove not to be available in quantity, it would be possible to operate the reactor with mixtures of sodium and beryllium fluorides as the basic fuel salt. The penalty imposed by sodium in terms of critical inventory and regeneration ratio is shown in Fig. 14-8, where typical Na—Be systems are compared with the corresponding Li-Be systems. With no thorium in the core, the use of sodium increases the critical inventory by a factor of 1.5 (to about 300 kg) and lowers the regeneration ratio by a factor of 2. The regeneration penalty is less severe, percentagewise, with 1 mole % ThF4 in the fuel salt; in an 8-ft-diameter core, the inventory rises from 800 kg to 1100 kg [cHAP. 14 NUCLEAR ASPECTS OF MOLTEN-SALT REACTORS 638 PINULIU 09 ¢9'0 LG°0 830 8¢ 0 8¢ 0 6S°0 , OI}BI UOI}BISUITIY 280 Lg°0 70 40 6€°0 €80 0 ‘o131 UOISSY-03-aInjdes A-u Ve ey 80 8¢ L I¢ 9% ‘SuoISSy [BULIAYT, ¢Iv°0 €281 °0 194 0T '8¢ LY 01 08% 0 A9 ‘A319Ud UOISSY UBIPATA] L8'1 961 GL'1 /7! LL°T 98°1 L ‘pretf woana N 762E 0 0LL8°0 2962 0 8LIE 0 124€ 0 G91% 0 s JoxuB[q UT Y, G9GZ 0 68L1°0 L8€T 0 21330 30610 09€1°0 I[es [0y Ul Yy, 20€0°0 $120°0 L3070 2s%0°0 92700 ¢z€0°0 [®S [ON] UL ger() 0S00°0 ¢¢00°0 19000 ¥900°0 ¢900°0 02000 9383 8o'| 81000 61000 61000 120070 12000 ¥300 0 [9SS9A I9IN() L9T0°0 ¢rco o LITO0 GE10°0 6G10°0 ¥1¢0 0 e Jeyur[q ur gy G060 0 8GET 0 eLv0°0 GLG0°0 9¢.0°0 86010 [9889A 810 L6ET 0 LY1IG 0 ¢690°0 €900 00600 LOET 0 HES [ong Ut J-TT-og 8¢¥a 0 Gl1C 0 ¢v0€ 0 966¢ 0 9¢8¢ 0 8% 0 (A-1) ¢ez) LS. 0 88820 86690 7%0L 0 PLIL 0 91GL "0 (SUOTSSY) ¢z soryer uorydiosqe uosjnapN 90¢ 8LT 0101 06. 0%S 162 ogz[] JO 3 ‘ArojuaAut eorjray) 901 ¢'19 ¥$% 861 ¢el L'3L ¢zl JO 8Y ‘ssewt [BO1ILI)) 8¢'G &L ¢ 6% 0¢ G6°GT 16 01 L8 G xA3ISUID WOIE ¢ez() Ge1°0 7800 08%°0 ey 0 0I€ 0 691°0 % °10W ‘41BS [90] UL ¢¢,) G0 GZ'0 I GL 0 G0 620 % o10W “yfes [ony ur ¥ qyf, L L 9 9 9 Q 3} ‘1930WBIP 910)) 82 L3 92 GZ G €3 IoquINU 9sBY) ‘e 6EE -OWN[OA [N [BUINXT *(989Y) M\ 009 :Iomod [810], AT % °1ow 12 + 2q9g %, opowt 91 + ¥ Y, % O10W €T :g "ON 9[es pequB[g TIUL + 70 + AV % P[0W €9 + 240G % o[out £¢ iz “ON 3[eS [oN] gez(l HLIM QITIN] SYOLOVAY LIVS-AATHONT] NALTOTN ‘SNOEANTDONOY ‘NOIDTY-OMJ, JO SOILSTMALOVAVH]) YVATONN ELVIQ-TVLLINT &1 a1avy, 639 HOMOGENEOUS REACTORS FUELED WITH U?3° 14-1] ‘ggg[] Ul PqIosqe uoIinsu 1od paqIosqe SuOIINaN{ '00/(1—0T X) SWOIVy €9°0 $9°0 290 i) 09°0 19°0 01981 UOI}BISUATY L8°0 €e 0 820 Gz 0 I¥°0 L8°0 0 ‘orpel UOISSY-03-01n3deo A-U I1 €¢ 8¢ Ig €V IT 0/, ‘SUOISST [BULIAY ], GL L 8150 9810 reuwioq} %1¢ g9 GG 19°2 A9 ‘A819Ud UOISSY UBIPIN 081 081 €6°'1 L6°T GL'1 08'1 L ‘pperd uoxnaN 083220 19920 2683 0 0%3¢ 0 99620 9982 " 0 }[es J9xuB[q Ul Y, L8980 109€°0 ev0€° 0 8L61°0 2g0g 0 08820 J[Bs [oN) Ul YT, 26800 G180 0 £820°0 LLT0°0 L¥%0°0 26800 [BS [90] UT gez[) 0700 0 Z¥00"0 £%00° 0 ¢%00 0 ¢%00 0 8700 0 9388o] €100 0 $100°0 97000 LT00°0 G100°0 91000 [ossoA 1IN0 ¢010'0 0810°0 29100 86100 8010 0 1€10°0 J[es joxuB[q Ul J-1T 81500 96900 29600 80810 L6%0°0 ¥%90°0 [9sS94 9100) 9701 0 $GE1 0 GH61°0 69.%°0 $280°0 01010 jres [Ny ul J-I7-og gIL% 0 $9%G 0 98150 98610 9065 "0 81.%°0 (A1) ¢ez) 883L 0 98SL°0 yI8L0 71080 ¥60L°0 ¢8%L 0 (SUOISSY) ¢ez() 1 so13e1 uor)dIosqe UOINBN 709 LE €% 6¥1 €8. 767 cez(] JO 39 ‘AI0JUAAUL [BORLLT 29 Go1 €01 "9 123 IL1 gez() JO BY ‘SSBUL [BOTYLL) 606 29°¢ 1g°¢ V25 6L €1 0.8 xAHISUSD WOJE g7 () $9% 0 €910 6600 7900 99¢°0 $9%°0 Yo 10U “TES [91] UL () I GL 0 G0 6% 0 I cL 0 0/, aJour ‘yes [ony ut ¥ gy, 8 8 8 8 L L 3J ‘19joWEIP 910)) e ee ze 1€ 0¢ 6% I9QUINU 3SE)) (penunuod) g-HI aIgv], 640 NUCLEAR ASPECTS OF MOLTEN-SALT REACTORS [cHAP. 14 0.8 l l | Li—-Be Salt with 1 mole ]0. % ThF4 . . 06— ¢ Fuel Salt A 8 \z-— — ® o £ kNo ThF4 8e-—e o Fuel Salt A 10 c 2 0.4 __.8. Fuel Salt B/ s — g ’ | 67 Na-Be Salt g 100 d No ThF4 With 1 mole % ThF4 5 £ - 8 Fuel Salt B | Numbers On Data Points Are Core Diameters In Feet | l | 0 400 800 1200 1600 Critical Inventory, Kg of y23s5 F1a. 14-8. Comparison of regeneration ratio and critical inventory in two-region, homogeneous, molten fluoride-salt reactors fueled with U235, Fuel salt A: 37 mole % BeF2 plus 63 mole %, Li’F. Fuel salt B: 46 mole 9, BeF; plus 54 mole 9, NaF. and the regeneration ratio falls from 0.62 to 0.50. Details of the neutron balances are given in Table 14-3. Reactwity coefficients. By means of a series of calculations in which the thermal base, the core radius, and the density of the fuel salt are varied independently, the components of the temperature coefficient of reactivity of a reactor can be estimated as illustrated below for a core 8 ft in diam- eter and a thorium concentration of 0.75 mole 9} in the fuel salt at 1150°F. From the expression k=f(T, p, R), where k is the multiplication constant, 7' is the mean temperature in the core, p is the mean density of the fuel salt in the core, and R is the core radius, it follows that 1dk 1[0k 1/0k\ dR , 1/0k\ dp Efi—i(5_7_’>p,ze+70<573>p,1’3—7—’+%<55>3,Td_T’ (14-1) where the term (1/k)(0k/3T),, r represents the fractional change in k due to a change in the thermal base for slowing down of neutrons, the term (1/k)(0k/0p)r, r represents the change due to expulsion of fuel from the core by thermal expansion of the fluid, and the term (1/k)(dk/ OR),, 7 represents the change due to an increase in core volume and fuel holding 641 HOMOGENEOUS REACTORS FUELED WITH U?23° 14-1] ‘agz(] UI POQI0sqe uorneu 1ad pagIosqe suopnaN | "09/(g1-0T X) SWOIV 2g 0 810 19°0 720 050 7¢°0 OI}®1 UOI}BISULZYY 170 L3°0 g% 0 620 %0 A 0 ‘O1J81 UOISSY-0-01n3ded A-U 'y vl 149 v 0 LT 9/, ‘SUOISSY [BULISY,], L80°0 9¢ 020 06T e'1 Ad ‘£319U9 UOISSY UBIPIIN gL' 1 61 gL'l 16°1 eL'1 €81 L ‘ppatd wonmaN 8¥01 0 09ST1°0 0S%1°0 €912 0 0313 0 $00€ 0 J[es JoxUB[q UL YL, 0L9¢°0 0S1€°0 813¢ 0 J[es [ong Ul Y, L9%0°0 €200 #8700 £920°0 L1300 09€0°0 188 [9N] Ul gez[) 9083 0 G010°0 9110°0 S¥10°0 Z810°0 23200 938O 90€3% 0 2670 0 G1€0°0 09900 1€%0°0 1280°0 168 jo3uE[q Ul J-Oog-8N 90€% 0 L160°0 2620°0 GgI1 0 9.%0°0 ISIT°0 [98894 BI00) 90€3% "0 6119°0 IT%1°0 GGLY 0 eSI1"0 1€.2°0 J[es [ong Ul JOg-eN 6165 0 8€15 0 6865 0 €925 0 ¥10€ 0 €895 0 (A-1) ggz) 180L°0 g98L°0 1104°0 LELL O 9869 0 LIBL 0 (SUOISSY) gez() Jsoryer uorpdiosqe UOIININ 0LTT ¥eC 0011 Q13 0€31 90¢ ggz() JO 33 'AIOJUAAUL [BOYLLY 012 44! 78 1°96 808 9L ggz() JO 3 ‘sseul [BOIILID 0 %31 Ly g 91 73 ¢ 6 75 L1°9 +KYISUSD WOJE gez() g8z 0 000 g9% 0 160°0 $10L°0 iZAN0 % o10ut “)BS [ON} UL geg() [ 0 1 0 I 0 0/, a[ow ‘78S [ony ut ¥ Y, 01 01 8 8 9 9 1j ‘1939WBIP 910 0¥ 6¢ 8¢ L8 ¢ Ge I9qUINU d8B]) '¢1] 6EE 1OWN[OA [aN] [8UINXF *(383Y) MTA 009 :1omod [810], ¥ IUL, % (oW 2 + 349d % o[owt ¢ + JBN % 9[owW 8¢ :3[8S 9que[q ‘(*d0 + *dUL) % oloW [ + 240g % 90w 9% + JBN % 9[0W §¢ 3188 [oNg cez(] HLIM QETEN] SHOIOVEY TAIHONT] WAITIAYE-WAICOS NALTON ‘SNOINADOWO ‘NOIDEY-OMJ, 0 SOLISIHAIOVEVH]) ¥VATOAN TVLLIN]T g1 ATEV], 642 NUCLEAR ASPECTS OF MOLTEN-SALT REACTORS [cHAP. 14 capacity. The coefficient dR/dT may be related to the coefficient for linear expansion, &, of INOR-8, viz: dR d_T = Ro. Likewise the term dp/dT may be related to the coefficient of cubical ex- pansion, 3, of the fuel salt: do _ dT pB. From the nuclear calculations, the components of the temperature co- efficient were estimated, as follows: 1/0k _ ~5/0 E(ST)p, _=—(0.13£0.02) X 10~5/°F, R<6k k a\R),,, =+ 0.412 £ 0.0005, p 916) _ k< 3p e T 0.405 4 0.0005. The linear coefficient of expansion, o, of INOR-8 was estimated to be (8.01+0.5) X 1075/°F [5], and the coefficient of cubical expansion, 3, of the fuel was estimated to be (9.889 4 0.005) X 10~ 5/°F from a correlation of the density given by Powers [6]. Substitution of these values in Eq. (14-1) gives a e (3.80 4 0.04) X 10-5/°F Y| = s} for the temperature coefficient of reactivity of the fuel. In this calculation, the effects of changes with temperature in Doppler broadening and satura- tion of the resonances in Th and U235 were not taken into account. Since the effective widths of the resonances would be increased at higher tem- peratures, the thorium would contribute a reactivity decrease and the U?3% an increase. These effects are thought to be small, and they tend to cancel each other. 14-1] HOMOGENEOUS REACTORS FUELED WITH U23° 643 Additional coefficients of interest are those for U23% and thorium. For the 8-ft-diameter cores, N(U235)< ok ) 14 [0.17N,(U?35) x 10~19] k (9N dN,(U235) k. \ON(Th)/vw=s k \ON(UZ35)/)yw dN(Th) where w — 0.0595N (Th) x 10™1® N (Th = 0.0805 ¢ X 10719, In these equations, N (U?235) represents the atomic density of U%3° in atoms per cubic centimeter, N,(U235) is the critical density of U235, and N(Th) is the density of thorium atoms. Heat release in core vessel and blanket. The core vessel of a molten-salt reactor is heated by gamma radiation emanating from the core and blanket and from within the core vessel itself. Estimates of the gamma heating can be obtained by detailed analyses of the type illustrated by Alexander and Mann [7]. The gamma-ray heating in the core vessel of a reactor with an 8-ft-diameter core and 0.5 mole 9% ThF4 in the fuel salt has been estimated to be the following: Source Heat release rate, w/cm3 Radioactive decay in core 1.4 Fission, n—y capture, and inelastic 5.2 scattering in core n—y capture in core vessel 4.5 n—y capture in blanket 0.3 Total 1.4 Estimates of gamma-ray source strengths can be used to provide a crude estimate of the gamma-ray current entering the blanket. For the 8-ft- diameter core, the core contributes 45.3 w of gamma energy per square centimeter to the blanket, and the core vessel contributes 6.8 w/cm?, which, multiplied by the surface area of the core vessel, gives a total energy escape into the blanket of 9.7 Mw. Some of this energy will be reflected into the core, of course, and some will escape from the reactor vessel, and 644 NUCLEAR ASPECTS OF MOLTEN-SALT REACTORS [cHAP. 14 therefore the value of 9.7 Mw is an upper limit. To this may be added the heat released by capture of neutrons in the blanket. From the Oecusol-A calculation for the 8-ft-diameter core and a fuel salt containing 0.5 mole % ThF4 it was found that 0.176 of the neutrons would be captured in the blanket. If an energy release of 7 Mev/capture is assumed, the heat release at a power level of 600 Mw (heat) is estimated to be 8.6 Mw. The total is thus 18.3 Mw or, say, 20 4+ 5 Mw, to allow for errors. No allowance was made for fissions in the blanket. These would add 6 Mw for each 1% of the fissions occurring in the blanket. Thus it appears that the heat release rate in the blanket might range up to 50 Mw. 14-1.2 Intermediate states. Without reprocessing of fuel salt. The nu- clear performance of a homogeneous molten-salt reactor changes during operation at power because of the accumulation of fission products and nonfissionable isotopes of uranium. It is necessary to add U235 to the fuel salt to overcome these poisons and, as a result, the neutron spectrum is hardened and the regeneration ratio decreases because of the accompanying decrease in 7 for U235 and the increased competition for neutrons by the poisons relative to thorium. The accumulation of the superior fuel U233 compensates for these effects only in part. The decline in the regeneration ratio and the increase in the critical inventory during the first year of operation of three reactors having 8-ft-diameter cores charged, respectively, with 0.25, 0.75, and 1 mole 9, ThF4 are illustrated in Fig. 14-9. The criti- cal inventory increases by about 300 kg, and the regeneration ratio falls about 16%,. The gross burnup of fuel in the reactor charged with 1 mole 7, ThF4 and operated at 600 Mw with a load factor of 0.809, amounts to about 0.73 kg/day. The U235 burnup falls from this value as U233 assumes part of the load. During the first month of operation, the U235 burnup averages 0.69 kg/day. Overcoming the poisons requires 1.53 kg more and brings the feed rate to 2.22 kg/day. The initial rate is high because of the holdup of bred fuel in the form of Pa233. As the concentration of this iso- tope approaches equilibrium, the U235 feed rate falls rapidly. At the end of the first year the burnup rate has fallen to 0.62 kg/day and the feed rate to 1.28 kg/day. At this time U233 contributes about 129, of the fissions. The reactor contains 893 kg of U235, 70 kg of U233, 7 kg of Pu23®, 62 kg of U?3%, and 181 kg of fission products. The U236 and the fission products capture 1.8 and 3.89, of all neutrons and impair the regeneration ratio by 0.10 units. Details of the inventories and concentrations are given in Table 144. With reprocessing of fuel salt. If the fission products were allowed to ac- cumulate indefinitely, the fuel inventory would become prohibitively large and the neutron economy would become very poor. However, if the fission products are removed, as described in Chapter 12, at a rate such that the 14-1] HOMOGENEOUS REACTORS FUELED WITH U235 645 0.7 [ | 2 © - c — 2 B o 1 mole % ThFy4 in Fuel Salt > 0.75 ® 2 — 0.50 | l—-—Processing Begins Inventory, Kg No Processing ——s]s— Processing Begins 100 |- || mole % ThF4 075 0.50 l 1 2 3 0 | Time of Operation, years Fia. 14-9. Operating performance of two-region, homogeneous, molten flyoride- salt reactors fueled with U235, Core diameter, 8 ft; total power, 600 Mw (heat); load factor, 0.80. equilibrium inventory is, for example, equal to the first year’s production, then the increase in U235 inventory and the decrease in regeneration ratio are effectively arrested, as shown in Fig. 14-10. The fuel-addition rate drops immediately from 1.28 to 0.73 kg/day when processing is started. At the end of two years, the addition rate is down to 0.50 kg/day, and it continues to decline slowly to 0.39 kg/day after 20 years of operation. The nonfissionable isotopes of uranium continue to accumulate, of course, but these are nearly compensated by the ingrowth of U233, As shown in Fig. 14-10, the inventory of U235 actually decreases for several years in a typical case, and then increases only moderately during a lifetime of 20 years. The rapid increase in critical inventory of U235 during the first year can be avoided by partial withdrawl of thorium. In Fig. 14-10 the dashed lines indicate the course of events when thorium is removed at the rate of 1/900 per day. Burnup reduces the thorium concentration by another 646 NUCLEAR ASPECTS OF MOLTEN-SALT REACTORS [cHAP. 14 0.7 o o £ 0.6 B 5 1 mole % ThF4 In Fuel Salt o | S=__ — §7 0.5 S —— Decreasing ThF 4 © T T e——— e Y | | l I 1200 T T T T T T T 1 ™ / | 2: 1000— / B 6 1 mole % ThF4 In Fuel Salt ’ 800 m — —— — «Q o . - Decreasing ThF , s 600 777 > o y % 777 / 7 > T 7777 RS \ SIS, /¢ L L RN Y L Dump Tank — F1gc. 15-4. Bellows-sealed, mechanically operated poppet valve for molten-salt service. - 15-4. SysteEm HEATING Molten-salt systems must be heated to prevent thermal shock during filling and to prevent freezing of the salt when the reactor is not operating to produce power. Straight pipe sections are normally heated by an elec- tric tube-furnace type of heater formed of exposed Nichrome V wire in a ceramic shell (clamshell heaters). A similar type of heater with the Ni- chrome V wire installed in flat ceramic blocks can be used to heat flat surfaces or large components, such as dump tanks, ete. In general, these heaters are satisfactory for continuous operation at 1800°F. Pipe bends, irregular shapes, and small components, such as valves and pressure- measuring devices, are usually heated with tubular heaters (e.g., General Electric Company *‘Calrods”) which can be shaped to fit the component or pipe bend. In general, this type of heater should be limited to service at 15-5] JOINTS 669 1500°F. Care must be exercised in the installation of tubular heaters to avold failure due to a hot spot caused by insulation in direct contact with the heater. This type of failure can be avoided by installing a thin sheet of metal (shim stock) between the heater and the insulation. Direct resistance heating in which an electric current is passed directly through a section of the molten-salt piping has also been used successfully. Operating temperatures of this type of heater are limited only by the corrosion and strength limitations of the metal as the temperature is increased. Experience has indicated that heating of pipe bends by this method is usually not uniform and can be accompanied by hot spots caused by nonuniformity of liquid flow in the bend. 15-5. JoINTS Failures of some system components may be expected during the de- sired operating life, say 20 years, of a molten-salt power-producing reactor; consequently, provisions must be made for servicing or removing and re- placing such components. Remotely controlled manipulations will be required because there will be a high level of radiation within the primary shield. Repair work on or preparations for disposal of components that fail will be carried out in separate hot-cell facilities. The components of the system are interconnected by piping, and flanged connections or welded joints may be used. In breaking connections between a component and the piping, the cleanliness of the system must be pre- served, and in remaking a connection, proper alignment of parts must be re-established. The reassembled system must conform to the original leaktightness specifications. Special tools and handling equipment will be needed to separate components from the piping and to transport parts within the highly radioactive regions of the system. While an all-welded system provides the highest structural integrity, remote cutting of welds, remote welding, and inspection of such welds are difficult operations. Special tools are being developed for these tasks, but they are not yet generally available. Flanged connections, which are attractive from the point of view of tooling, present problems of permanence of their leak- tightness. Three types of flanged joints are being tested that show promise. One is a freeze-flange joint that consists of a conventional flanged-ring joint with a cooled annulus between the ring and the process fluid. The salt that enters the annulus freezes and provides the primary seal. The ring provides a backup seal against salt and gas leakage. The annulus between the ring and frozen material can be monitored for fission product or other gas leak- age. The design of this joint is illustrated in Fig. 15-5. 670 MOLTEN-SALT REACTOR HEAT-TRANSFER EQUIPMENT [CHAP. 15 —.H.<-Gap (~1/16in.) -Soft-Iron or Copper Seal Ring S \“§‘ \ 23— Air Channel for Cooling A N Frozen -Salt Seal AN 1 BN BN \ . JW% \\ ‘\X&/Weld of Flange to Loop Tubing Salt Flow ; Ring Insert to Provide Labyrinth p for Salt Leakage ; ~—~Frozen-Salt Seal Narrow Section to Reduce Heat Transfer from the Molten Salt in the Loop Tubing \ X SO IO 35 1 % 0 VY2 1 = e T ] Inches Air Inlet to Cooling Channel Between Flange Faces Indicate Region of Frozen-Salt Seal; Indicate Region of Transition from Liquid to Solid Salt Fig. 15-5. Freeze-flange joint for 1/2-in.-OD tubing. +— Seal Material Insert { Shown Before Being | Fused to Form Seal AV /\V 77 7 72 W\ X\ i /¥ 7N\ 7 - N o NS - Weld of / Inches Flange to Loop Tubing Fi1a. 15-6. Cast-metal-sealed flanged joint. 15-6] INSTRUMENTS 671 Weld of Flange to Loop Tubing Copper Ring — nches Figc. 15-7. Indented-seal flange. A cast-metal-sealed flanged joint is also being tested for use in vertical runs of pipe. As shown in Fig. 15-6, this joint includes a seal which is cast in place in an annulus provided to contain it. When the connection is to be made or-broken the seal is melted. Mechanical strength is supplied by clamps or bolts. A flanged joint containing a gasket (Fig. 15-7) is the third type of joint being considered. In this joint the flange faces have sharp, circular, mating ridges. The opposing ridges compress a soft metal gasket to form the seal between the flanges. 15—-6. INSTRUMENTS Sensing devices are required in molten-salt systems for the measurement of flow rates, pressures, temperatures, and liquid levels. Devices for these services are evaluated according to the following criteria: (1) they must be of leaktight, preferably all-welded, construction, (2) they must be cap- able of operating at the maximum temperature of the fluid system, (3) their accuracies must be relatively unaffected by changes in the system tem- perature, (4) they should provide lifetimes at least as great as the lifetime of the reactor, (5) each must be constructed so that, if the sensing element fails, only the measurement supplied by it is lost. The fluid system to which the instrument is attached must not be jeopardized by failure of the sensing element. 15-6.1 Flow measurements. Flow rates are measured in molten-salt systems with orifice or venturi elements. The pressures developed across the sensing element are measured by comparing the outputs of two pressure- measuring devices. Magnetic flowmeters are not at present sufficiently sensitive for molten-salt service because of the poor electrical conductivity of the salts. 672 MOLTEN-SALT REACTOR HEAT-TRANSFER EQUIPMENT [cHAP. 15 15-6.2 Pressure measurements. Measurements of system pressures re- quire that transducers operate at a safe margin above the melting point of the salt, and thus the minimum transducer operating temperature is usually about 1200°F. The pressure transducers that are available are of two types: (1) a pneumatic force-balanced unit and (2) a displacement unit in which the pressure is sensed by displacement of a Bourdon tube or diaphragm. The pneumatic force-balanced unit has the disadvantages that loss of the instrument gas supply (usually air) can result in loss of the measurement, and that failure of the bellows or diaphragm would open the process system to the air supply or to the atmosphere. The displacement unit, on the other hand, makes use of an isolating fluid to transfer the sensed pressure hydro- statically to an isolated low-temperature output element. Thus, in the event of a failure of the primary diaphragm, the process fluid would merely mix with the isolating fluid and the closure of the system would be unaffected. 15-6.3 Temperature measurements. Temperatures in the range of 800 to 1300°F are commonly measured with Chromel-Alumel or platinum- platinum-rhodium thermocouples. The accuracy and life of a thermocouple in the temperature range of interest are functions of the wire size and, in general, the largest possible thermocouple should be used. Either beaded thermocouples or the newer, magnesium oxide-insulated thermocouples may be used. 15-6.4 Liquid-level measurements. Instruments are available for both on-off and continuous level measurements. On-off measurements are made with modified automotive-type spark plugs in which a long rod is used in place of the normal center conductor of the spark plug. To obtain a con- tinuous level measurement, the fluid head is measured with a differential pressure instrument. The pressure required to bubble a gas into the fluid 1s compared with the pressure above the liquid to obtain the fluid head. Resistance probe and float types of level indicators are available for use in liquid-metal systems. 15-6.5 Nuclear sensors. Nuclear sensors for molten-salt reactors are similar to those of other reactors and are not required to withstand high temperatures. Existing and well-tested fission, ionization, and boron tri- fluoride thermal-neutron detection chambers are available for installation at all points essential to reactor operation. Their disadvantages of limited life can be countered only by duplication or replacement, and provisions can be made for this. It should be pointed out that the relatively large, negative temperature coefficients of reactivity provided by most circulating- fuel reactors make these instruments unessential to the routine operation of the reactor. CHAPTER 16 AIRCRAFT REACTOR EXPERIMENT* The feasibility of the operation of a molten-salt-fueled reactor at a truly high temperature was demonstrated in 1954 in experiments with a reactor constructed at ORNL. The temperature of the fuel exiting from the core of this reactor was about 1500°F, and the temperature of the fuel at the inlet to the core was about 1200°F. The reactor was constructed before the mechanism and control of corrosion by molten salts had been fully explored, and therefore the experimental operation of the reactor was of short duration. Since the work was supported by the Aircraft Reactors Branch of the Atomic Energy Commission, the reactor was called the Air- craft Reactor Experiment (ARE).} | The ARE was a thermal reactor in which moderation was accomplished by BeO blocks through which the fluoride fuel was circulated in Inconel tubes arranged in a symmetrical, heterogeneous matrix. The Inconel ves- sel containing the core was essentially a right cylinder, approximately 52 in. OD and 44 in. in height, with 2-in.-thick walls. The fuel passages consisted of 1-in.-diameter Inconel tubes arranged in six parallel circuits, and each circuit, by the use of reverse bends at top and bottom of the core, made eleven passes through the core. The fuel passages did not traverse the peripheral BeO blocks which served as a reflector around the core of the reactor. A top view of the BeO blocks and the Inconel tubes is shown in Fig. 16-1. The moderator and reflector blocks were cooled by circulating liquid sodium from the bottom to the top of the pressure vessel. The sodium permeated all interstices of the BeO and flowed rap- idly through 1/2-in. vertical holes in the reflector sections of the BeO. An elevation drawing of the reactor which illustrates these features is presented in Fig. 16-2, and a photograph of the reactor vessel that was taken before assembly of the thermal shield is shown in Fig. 16-3. Since the purpose of the operation of this experimental reactor was to study the behavior of the circulating-fluoride-fuel system and to identify the problems associated therewith, the power output of the reactor was not utilized but, rather, was simply dumped as heat. The heat-removal system is shown schematically in Fig. 16-4. The fuel was circulated through a finned-tube radiator type of heat exchanger. This radiator was located within a sheet-metal housing of a toroidal shape. In another part of the toroidal housing there was a second finned-tube radiator through *By E. S. Bettis and W. K. Ergen. R. C. Briant et al., Nuclear Science and Engineering, Vol. 2, No. 6, 795-853 (1957). 673 674 AIRCRAFT REACTOR EXPERIMENT [cHAP. 16 Fie. 16-1. Top view of the reactor core of the ARE. Hexagonal beryllium oxide blocks serve as the moderator. Inconel tubes pass through the moderator blocks to carry the molten-salt fuel. which plant water flowed. A large centrifugal blower circulated the coolant gas (helium) in the toroidal loop so that heat was picked up from the fuel radiator and dumped into the water radiator. An identical arrangement of radiators and blower was used for cooling the sodium used as the moderator-reflector coolant. In the interest of safety (for removal of afterheat in the event of a pump failure), the so- dium circuit was installed in duplicate so that an entire sodium cooling system was availlable as a spare. These two sodium loops were operated alternately during the experiment in an effort to keep a check on the op- erability of each loop. Had one loop failed to operate, the experiment would have been terminated for lack of a spare cooling system. The control system of the reactor was based on conventional practice. The three safety shim rods were actuated by electrically driven lead screws which moved electromagnets in a vertical plane. When these magnets were driven to their lowest extremity, an armature was engaged to which the shim (poison) rods were attached. Loss of current in the electromagnets would allow the rods to fall under the action of gravity into thimbles in the central region of the core. The regulating rod was a simple stainless- steel pipe which was rigidly attached to a rack driven by a reversible elec- CHAP. 16] Reflector Coolant Fic. 16-2. Elevative section of the Aircraft Reactor Experiment. Safety Rod Assembly Tubes Fuel Tubes AIRCRAFT REACTOR EXPERIMENT Regulating Rod RN SHUEOUIRNER 7 N % A, v V.78 V% | ] A Assembly, L4 NN P Y SN t NN 27277777 NN 77770807 NN\ 77777477, NS NSNS 5o R R iY77 7/ P A S N N N DN SN V« N ?&\\\\\\\\\\\\\\\\\\\\\\\\\\S V0 gaRkzzza 1y s & B g | m /] o i o [i J j>\\\\V v AN N ol i d N7 Z280\ \\Z N D eSS OO e M IS N ekl o KA AN\ AN Z2 st 2 Y S OSSN Ll S SN AN\ 7222 N iz Y S S NN 7 2s NSNS - AWA 4 RSN e S I ez N« R A A7 By d ‘—-W\\‘\\\\\WWWM\\\\Y\\\ //////IIIIIII< LA N NNV Fuel Inlet Manifold Vi NN\ Miiic e S l '\ |_|_ R0 Y S / AN, EHEHEHHREHREHK 72000204 SIS AR N YV \ Y > P N WA 4 NV N N §Z§ % § §/ o 0O — O - Helium — Reflector 7] Reflector Coolant Fic. 16-4. Schematic diagram of the heat-removal system for the ARE. be pressurized from the tanks into the system and could be drained back into the tanks after the experiment was over. Dry helium was used for operating penumatic instruments and for pressurizing the liquids into the system from the tanks. Pumps for both the sodium and the molten-fluoride mixture consisted of sump-type centrifugal pumps with overhanging shafts. The pumps were mounted vertically, and a gas space was provided between the liquid level and the upper bearings of the pump. The pumps were located so that the free-liquid surface in the sump tank was the high point in both the fuel and the sodium circuits. The sump tank of the pump also served as an expansion tank for the liquid. The isometric drawing of the fuel system presented in Fig. 16-5 indicates the relative levels of the components. Both of the liquid systems, fuel and sodium, were fabricated entirely of Inconel, and all closures were made by inert-gas-shielded electric-arc (Heliarc) welding. The welding procedure was adopted after extensive experimental research and developmental work, and meticulous care was exercised in all welding operations. The entire reactor system, that is, the reactor vessel, heat exchangers, pumps, dump tanks, piping, and auxiliary equipment (with the exception of control rod drives), was located in con- crete pits below ground level. After the reactor was brought to criticality by manual fuel injection, concrete blocks were placed on top of the pits to complete the shielding of the system as required during power operation. Fuel was added as a molten mixture of NaF and UF4 (enriched in U23%) after the sodium system had been heated and filled with sodium and the fuel system had been heated and filled with fuel carrier—a molten mixture of NaF and ZrF.. The fuel additions were made into the sump of the fuel pump through the use of a temporary enrichment system that was capable of injecting (by manual operation) a few hundred grams of fuel mixture 678 AIRCRAFT REACTOR EXPERIMENT [cHAP. 16 Standby Fuel Pump A2 7 7 ¥, Frangible M Disk Valve R Heat Exchanger &y No. 2 b Heat Exchanger No. 1 o Hot Fuel Dump Tank Carrier Fill Line 4 Reserve Tank No. 1 Fill Tank No. 2 F16. 16-5. Layout of the fuel system components for the ARE. at a time. This method of fuel addition was laborious and time-consuming, but it effectively and safely enriched the reactor to a critical concentration. The reactor was taken to criticality essentially without incident. The total amount of U235 added to the system to make the reactor critical was approximately 61 kg, but small amounts of fuel were withdrawn from the system for sampling and in trimming the pump level. The uranium con- centration at criticality was 384 g/liter of fluoride mixture. The calculated volume of the core was 38.8 liters at 1300°F, and thus the clean critical mass of the reactor was 14.9 kg of U235, It was demonstrated that the reactor had an over-all temperature co- efficient of reactivity of —6 X 10~5 (Ak/k)/°F. As was anticipated, the fast negative temperature coefficient of reactivity (associated with the fuel expansion coefficient) served to stabilize the reactor power level. From a power lever of 200 kw upward, the temperature coefficient controlled the system so precisely that the reactor responded to load demands in a thoroughly reliable manner. The response of the reactor was demonstrated in a number of experi- ments, one of which is described in Fig. 16-6. The abscissa, to be read from right to left, is the time in minutes, and the print-outs from recorders giving the reactor inlet and outlet temperatures are the ordinate. Initially, in this experiment, the reactor was operating at low power. Then the heat CHAP. 16] AIRCRAFT REACTOR EXPERIMENT 679 Reactor Inlet and Outlet Tube Temperatures 0045 hundreds o‘f" degrees F 0040 T 11111 teed Shim Rods Ins 0035 , Fuel and Na Blowers ‘ Low Temperature 3 0030 Interlock o Reduced Blower ~ 0025 | Turlned o? Fuel o ) i1]'| Helium B — Cooled & 1'{ Helium Blower > 0020 'Broughf (FU" Speed) 2 Inserted Regulating Rod 0015 Power Reactor Subcritical Extraction Turned Off Fuel Helium 0010 Blower 0005 2400 > 2355 l1l{Turned on System o Helium Blower w 2350 (Full Speed) g 2345 ‘ Inserted Shim Rods = A Fuel System 2340 : Helium Blower Speed to Zero 2335 | B o 2330 °:~ 2325 > 2 2320 System 2315 Blower on ° _ Increase Fuel 2310 s | System Helium Blower » Speed to Maximum 2305 1 bl IR | 1 "LReactor Critical: 2300 .. 1 Operating | e 2255 Fic. 16-6. Chart of inlet and outlet temperatures for the ARE as influenced by various experimental procedures. extraction from the fuel was slowly increased and there was, first, a re- sultant decrease in the temperature of the fuel which reached the reactor inlet from the heat exchanger. This increased the reactivity and the re- actor power, as indicated by the temperature rise at the reactor outlet. The spread of inlet and outlet temperatures corresponds to a power level of 2.5 Mw. When the heat extraction was reduced, the inlet temperature 680 AIRCRAFT REACTOR EXPERIMENT [cHAP. 16 rose and the outlet temperature fell until the two temperatures became nearly coincident. As may be seen, the control rods did not determine the power output; they only influenced the average temperature. Insertion of the shim rods decreased the temperature. Another rapid increase in the power demand on the fuel system again spread apart the inlet and outlet temperature recordings, and full insertion and full withdrawal of the regulating rod depressed and then raised both temperatures simultaneously. Next, the power extraction was stopped and the regulating rod was in- serted to make the reactor subecritical. The third spread of the temperatures in Iig. 166 was a result of a demonstration which showed that the reactor could be brought to criticality, without use of the rods, by the power demand alone. Power extraction from the sodium system cooled the reactor to make it critical, and power extraction from the fuel again caused the spread of inlet and outlet tem- peratures. The remarkable stability of the system made it unexpectedly possible to demonstrate that no more than 59, of the Xel3% was retained in the molten fuel. It had been computed that the xenon poisoning after 27 hr of operation at full power would amount to 2 X 1073 in Ak/k if all the xenon formed stayed in the fuel until it decayed. This level of poisoning was less than would be expected from the usual equations, partly because the fuel spent only one-fourth of the time in the core and was thus effec- tively only subjected to one-fourth of the flux, and partly because many of the neutrons had energies above the large Xe!3% absorption resonance. As little as 59 of this computed poisoning would have been detectable, but none was found. There was a small leakage from the gas volume above the liquid surface of the fuel pumps which made operation at a high power level somewhat awkward, but danger to operating personnel was circumvented by operat- ing with the reactor pit at a subatmospheric pressure and remotely ex- hausting the pit gases to the atmosphere at a location where they were adequately dispersed. The entire program of experiments that had been planned for the reactor was completed satisfactorily. The reactor was shut down after a total power production of 96 Mwh, and it was later dismantled. The fuel and sodium systems had been in operation for a total of 462 and 635 hr, re- spectively, including 221 hr of nuclear operation, with the final 74 hr of operation in the megawatt range. CHAPTER 17 CONCEPTUAL DESIGN OF A POWER REACTOR* The design of a homogeneous molten-salt reactor of the type discussed in the preceding chapters is described below. The choice of the power level for this design is arbitrary, since the 8-ft-diameter reactor core, chosen from nuclear considerations, is capable of operating at power levels up to 1900 Mw (thermal) without excessive power densities in the core. An electrical generator of 275-Mw capacity was chosen, since this is in the size range that a number of power companies have used in recent years. It is estimated that about 6% of the power would be used in the station, and thus the net power to the system would be about 260 Mw. Two sodium circuits in series were chosen as the heat-transfer system between the fuel salt and the steam. Delayed neutrons from the circulating fuel will activate the primary heat exchangers and the sodium passing through them. A secondary heat-exchanger system in which the heat will transfer from the radioactive sodium to nonradioactive sodium will serve to prevent radioactivity at the steam generators, superheaters, and re- heaters. The fuel flow from the core is distributed among four primary heat exchangers which serve as the first elements of the four parallel paths for heat transfer to the steam. A single primary heat exchanger and path 1s provided for the blanket circuit. Plan and elevation views of the reactor plant are shown in Figs. 17-1 and 17-2, and an isometric drawing showing the piping of the heat-transfer systems is shown in Fig. 17-3. The reactor and the primary heat exchangers are contained in a large rectangular reactor cell, sealed to contain any leakage of fission-product gases. All operations in the cell must be carried out remotely after the reactor has operated at power. The principal char- acteristics of the plant are listed in Table 17-1. | 17-1. FUEL AND BLANKET SYSTEMS 17-1.1 Reactor vessel. The reactor vessel and the fuel and blanket pumps are a closely coupled assembly (Fig. 17-4) which is suspended. from a flange on the fuel pump barrel. The vessel itself has two regions— one for the fuel and one for the blanket salt. The fuel region consists of the reactor core surmounted by an expansion chamber, which contains the single fuel pump. The blanket region completely surrounds the fuel region, and the blanket salt cools the walls of the expansion chamber gas space and shields the pump motor. The floor of the expansion chamber is *By L. G. Alexander, B. W. Kinyon, M. E. Lackey, H. G. MacPherson, L. A. Mann, J. T. Roberts, F. C. VonderLage, G. D. Whitman, and J. Zasler. 681 682 CONCEPTUAL DESIGN OF A POWER REACTOR [cHAP. 17 Primary Sodium Pump (1 of 5) Reactor . Intermediate HeITt Primary Sodium To Secondary Exchanger Ce Sodium Heat Exchanger (1 of 5) Blanket Pump Primary Secondary Air Lock Shield Secondary Sodium Circvuits Shield Fuel Pump Boiler Hot Maintenance Area\ (1 of 5) A T XA p 3 Maintenance :, T .4_:5\:?0 . T Turbo- ¢ Heater | y Generator ' Removal ' Fuel Drain-g . | Tank b el Fd Rl | [ ree—= | spe—Oflcl L L + ’ ; | (lr‘. R h Chemical " eheater ¢ , Processing : (1 of 4) Control /s S 2 Blanket Superheater Fuel Chemicol EnriCher B|Onkef H.eCIf. (] Of 5) Processing Transfer Circuit / F.Ud Expansion Secondary Sodium Blanket Drain Enricher Pump (1 of 5) Blanket BlanketTo-Sodium Chemical Heat Exchanger . Processing Fuel-To-Sodium Heat Exchanger (1 of 4) Fia. 17-1. Plan view of molten salt power reactor plant. a flat disk, 3/8 in. thick, which serves as a diaphragm to absorb differential thermal expansion between the core and the outer shells. 17-1.2 Fuel pump. The fuel pump is of the type illustrated in Chap- ter 15 (Fig. 15-3) and is designed to have a capacity of 24,000 gpm. It is driven by a 1000-hp motor with a shaft speed of 700 rpm. This pump incorporates three major advanced features that are being developed, but which are not present in any molten-salt pump operated to date. These are a hydrostatic lower bearing to be operated in the molten salt, a laby- rinth type of gas seal to prevent escape of fission-product gases up the shaft, and a hemispherical gas-cushioned upper bearing to act as a com- bined thrust and radial bearing. These advanced features are intended to provide a pump with greater resistance to radiation damage and less complex auxiliary equipment than necessary for pumps presently used for molten salts. 17-1.3 System for removal of fission-product gases. About 3.5%, of the fuel passing through the fuel pump is diverted from the main stream, 17-1] FUEL AND BLANKET SYSTEMS 683 Primary Sodium To Secondary Sodium Heat Exchanger (1 of 5) Fuel To Sodium Heat Exchanger (1 of 4) Jet Pump Superheater Fuel crmary (1 of 5) (1 of 5) Manipulator Scm ' Pump Removable Boiler (1 of 5) Air Lock (1 of 5) Hot Maintenance Area Concrete Slabs Turbine-Generator O Maintenance Area Heater Removal Steam Header Fuel Tank Blanket To Sodium Blanket Pump Heat Exchanger S:Z?fr:dP%Zp (1 of 3) (1 of 5) Boiler Feed Water Pump Primary Sodium Secondary Sodium (1 of 5) Drain Tank Drain Tank (1 of 5) (1 of 5) Fia. 17-2. Elevation view of molten salt power reactor plant. F1c. 17-3. Isometric view of molten salt power reactor plant. Section A-A Siphon Drain Fuel Pump Motor Fuel Line To Heat Exchanger < 0123 435 Scale—Feet Fuel Return i F1g. 17-4. Reactor vessel and pump assembly. / / Molten Salt NN ~ Blanket Pump Motor Blanket Expansion Tank Wil Blanket Return . Fuel Expansion Tank Breeding Blanket 17-1] FUEL AND BLANKET SYSTEMS 685 TABLE 17-1 REACTOR PLANT CHARACTERISTICS Fuel Fuel carrier Neutron energy Moderator Primary coolant Power Electric (net) Heat Regeneration ratio Clean Average (20 yr) Blanket salt Refueling cycle at full power Shielding Control Plant efficiency Exit fuel temperature Steam Temperature Pressure Second loop fluid Third loop fluid Structural materials - Fuel circuit Secondary loop Tertiary loop Steam boiler Steam superheater Active-core dimensions Fuel equivalent diameter Blanket thickness Temperature coefficient, (Ak/k)/°F Specific power Power density Fuel inventory Initial (clean) Average (20 yr) Clean critical mass Burnup > 909, U235F, 62 mole 9, LiF, 37 mole 9, BeFs, 1 mole 9, ThF, Intermediate LiF Ber Circulating fuel solution, 23,800 gpm 260 Mw 640 Mw 0.63 0.50 71 mole 9, LiF, 16 mole 9, BeF, 13 mole 9, ThFy Semicontinuous Concrete room walls, 9 ft thick Temperature and fuel concentration 44.39, 1210°F at approximately 83 psia 1000°F, with 1000°F reheat 1800 psia, Sodium Sodium INOR-8 Type-316 stainless steel 5%, Cr, 19, Si steel 2.59%, Cr, 19, Mo steel 5% Cr, 19, Si steel 8 ft 2 ft —(3.84+0.04) X 10°5 1000 kw/kg 80 kw/liter 604 kg of U235 1000 kg of U235 267 kg of U235 Unlimited 686 CONCEPTUAL DESIGN OF A POWER REACTOR [cHAP. 17 Y ~—~Blanket Pump A Blanket Expansion SCFM ) Tank — 0.1 CFS Fuel Expansion 1210% Blanket Bypass Tank 1.8 CFS 0.1CFS 1250°F Fuel Bypass i anke Y < - A 235 F13 785 Ft3 61 Ft3 16 F3 v 235 Ft3 785 Ft3 61 F13 16 Ft3 738%F 212°F 62°F —40°F Av Av ~Av Av Y y i Coolant Coolant Coolant Coolant Fic. 17-5. Schematic flow diagram for continuous removal of fission-product gases. mixed with helium from the pump-shaft labyrinth seal, and sprayed into the reactor expansion tank. The mixing and spraying provides a large fuel-to-purge-gas interface, which promotes the. establishment of low equilibrium fission gas concentrations in the fuel. The expansion tank provides a liquid surface area of approximately 26 ft? for removal of the entrained purge and fission gas mixture. The gas removal is effected by the balance between the difference in the density of the fuel and the gas bubbles and the drag of the opposing fuel velocity. The downward surface velocity in the expansion tank is less than 1 in/sec, which should allow all bubbles larger than 0.008 in. in radius to come to the surface and escape. In the Aircraft Reactor Experiment at least 97% of the fission-product gases were continuously purged by similar techniques. With a fuel purge gas rate of 5cfm, approximately 350 kw of beta heating from the decay of the fission-product gases and their daughters is deposited in the fuel and on metal surfaces of the fuel expansion tank. This heat is partly removed by the bypass fuel circuits and the balance is transferred through the expansion tank walls to the blanket salt. The mixture of fission-product gases, decay products, and purge helium leaves the expansion tank through the off-gas line, which is located in the top of the tank, and joins with a similar stream from the blanket expansion tank (see Fig. 17-5). The combined flow is delayed approximately 50 min in a cooled volume to allow a large fraction of the shorter-lived fission products to decay before entering the cooled activated-carbon beds. The 17-2] HEAT-TRANSFER CIRCUITS AND TURBINE GENERATOR 687 capacity of the carbon beds will hold krypton from passing through for approximately 6 days, and xenon for much longer times. The purge gases, essentially free from activity, leave the carbon beds to join the gases from the gas-lubricated bearings of the pumps. The gases are then compressed and returned to the reactor to repeat the cycle. Ap- proximately every four days the gas stream is diverted from one set of carbon beds to the other. The inactive bed is then regenerated by warming it to expel the Kr3% and other long-lived fission products. It will probably be economical to recover some of these gases; others may be expelled to the stack. 17-2. HeAT-TRANSFER CirRcuIiTs AND TURBINE (GENERATOR The primary heat exchangers are designed to have the fuel on the shell side and sodium inside the tubes. This arrangement makes full use of the superior properties of sodium as a heat-transfer fluid and appears to yield the lowest fuel volume. The heat exchangers, which are of semicircular construction, as shown in Fig. 17-3, provide convenient piping to the top and bottom of the reactor. The thermal characteristics of the primary heat exchanger, to- gether with the characteristics of other heat exchangers of the reactor system, are listed in Table 17-2. The sodium in the intermediate heat-transfer system (see Fig. 17-6) 1s heated by the fuel in the primary heat exchanger and is pumped out of the reactor cell and through the reactor cell shield to adjacent cells, which con- tain the secondary sodium-to-sodium heat exchangers and the pump. No control of intermediate sodium flow is required, so there are no valves and a constant speed centrifugal pump is used. To permit the sodium to be at a lower pressure than the fuel in the primary heat exchanger, the pump for the intermediate sodium is in the higher temperature side of the circuit. The secondary heat exchangers are of the U-tube in U-shell, counterflow design, with the intermediate sodium in the tubes and the final sodium on the shell side. The final sodium circuit, except for the secondary exchanger, is outside the shielded area and thus available for adjustment and maintenance at all times. The principal problems in this circuit are concerned with the ad- justment of sodium temperature. Excessive thermal strains are prevented in the steam generator by limiting the temperature of the sodium entering it, and in the intermediate heat exchanger by the regulation of sodium flows so that too cold sodium is never returned to it. The hot sodium from the secondary exchanger is split into three streams with regulating valves for control of the relative flows. One stream bypasses the steam system and goes directly to a blender; the flow in it is, of course, greatest at low power CONCEPTUAL DESIGN OF A POWER REACTOR [cHAP. 17 688 PINULIU0D 8 ¥l 01 G'GT 0¥ 1sd ‘doap aansserq G €l 6 €I L 61 8 01 sdy ‘Ay00[PA pPINg 9 &8 [ 9y 1 9% ¥ €l sdjo ‘eyer mo[q 6 A (233) (1) /N3¢ 00T ‘XNY 983y 9FBIAAY 003¢ 0083 1] ‘BaIB I9JsuBI) 160 vyl iadl MIN ‘A910edBO J9JSueI) 3B 9¢ 8¢ "ur ‘I9jewReIp dpung 868 0 Vil ut ‘(y) yond OVl a1¢ I9quIn N ¢ 1¢ L €3¢ 1} ‘qI3ue] 6¥0 0 8G0°0 "ur ‘sseuOIy} [BM 0S40 000 T "ul ‘I9jewBIp dApIsINQ [0998 ssoquIe)s gre—odAT, S—JONI [BLIR)B]AL DIOP QN ], ¢cs8 Gc6 Gco GLOT A, ‘PuUs pron 0801 0cT1 0cIl 01cI A, ‘Pue 101 saunyosaduia J, MO}I93unod MOFI)UNO?d ‘19Ys—[1 ut aqni—N ‘[IeYs—(] ur aqni— I28uBYIX9 Jo adA T, [°US soqng, SoqnLL [1PUS uoneao[ pmiyq wnipos A1epuodsg wWNIPOS AIBWILIJ wnipos Arswij es PNy pinq ¥ ¥ paanbax Jequun ) $LPBUDYITI WNIPOS-0]-UUNIPOS PUD [IN ] flunpuorsg flupwira J SUADNVHOXY LVAJ] ¥0d VIV(] Z—L21 @14v], HEAT-TRANSFER CIRCUITS AND TURBINE GENERATOR 689 17-2] ¥ 0I G 8 ¢ 01 69 (dwnd 30l) 2°¢ 18d ‘doap aunssar] LET 6 L 9 €6 9°¢ sdj ‘£j100[0A PIN[Y 66 90% 01¥ I4/q1 0001 10 8 91 G ¢ql G LG sdjo ‘©ye1 Mol q Ge 9L 00T (237) (1) /N3 0001 XN 8Oy SCIOAY 00¢¢ 09.1 0083 1] ‘BoIe I9JsuBI} 389 9°¢¢ G 68 %' 78 MIN ‘A310edeo 19jsuBI} J86OH L 62 €2 GG "ul ‘10)9WEIp s[pung 00T 00T 6L ut {(v) uord 008 08% G9¢€ IequIn N ¢ 91 ¢ 8T 13 ‘U38ua] G900 ¢60 0 08T 0 "Ul ‘SSOUNIIY] [[BM 0.0 0S.°0 z "Ul ‘19)oWBIP IPISINQ KONV 18 %1 ‘10 %S | AoV 18 %T 10 %S | 4onv oW %I 1D %9T [BLIOYEIAL P 2qN,J, 0%9 000T 129 0e6 129 0F2 do ‘PUe PIOD 000T 080T | 0001 080T 129 Gz8 d, ‘Pue 1°H $aungoadwia J, MOJJI9UN09 MOJJI9}UNOD MO}JI9)UNO09d “pySreng q19Ys—( U1 aqn3—-n ‘youofeg 108ueyoxe jo odA, seqn, PYS | seqny, PYS | seqny, 1°US uonyeoor P wnIpos wnipos WNIpos wedns AI1Bpuoosg wea)y AIBpuoosyg 1978 M K18pu099g pmiq ¥ ¥ b paambal zaquInN LoBUDYITI WDI)S-0)-WUNIPOY L2J0YIY L2702y Ladng LOIDLIUIL) UDIIS (ponunuod) g—,1 414v], 690 CONCEPTUAL DESIGN OF A POWER REACTOR [cHAP. 17 Low- and Boiler Turbines Turbine urbine Stop Relief Water Station Pump Sodium Superheater In Series Steamchest Pump Feedwater in Series Fig. 17-6. Schematic diagram of heat-transfer system. levels. The other two streams go to the superheater and the reheater, and are then combined with the bypass flow in the blender. On leaving the blender, the sodium stream is split again by a three-way valve into two streams; one enters a second bypass and goes directly to the main pump and the other enters a jet pump that keeps a large sodium flow recirculating through the boiler, which is of the Lewis type. The three-way valve is adjusted, at design point, so that about two-thirds of the flow goes to the jet pump and one-third bypasses the boiler. At low power levels the valve would be adjusted to give very low flows to the boiler. The centrifugal pump in this circuit has two speeds, full speed and one- fourth of full speed. The low-speed operation provides for better regulation of the sodium flow at very low power levels. The turbine selected uses 1800-psia steam at 1000°F with reheat to 1000°F and is rated at 275 Mw. It is a 3600-rpm single-shaft machine with three exhaust ends. The turbine heat rate is estimated to be 7700 Btu/ kwh, or 44.39%, cycle efficiency, while 7860 and 8360 Btu/kwh are the generator and station heat rates, respectively. With 69, of generator output used for station auxiliaries, 260 Mw is supplied to the bus bar. 17-3. REMOTE MAINTENANCE PROVISIONS Remotely controlled mechanized tools and viewing devices are provided in the reactor cell for making minor repairs and for removing and re- 17-4] MOLTEN-SALT TRANSFER EQUIPMENT 691 placing any component in the cell. The tools will be able to handle any pump, heat exchanger, pipe, heater for pipe and equipment, instrument, and even the reactor vessel, and, correspondingly, the components will be de- signed and located for accessibility and separation. The removal and replacement of components requires a reliable method of making and breaking joints in the pipe. Cutting and welding of pipe sections can be used, but in the low-pressure molten-salt system 1t 1s be- lieved that a flanged-pipe joint (see Section 15-5) may be satisfactory. All equipment and pipe joints in the reactor cell are laid out so that they are accessible from above. Directly above the equipment is a traveling bridge on which can be mounted one or more remotely operated manipu- lators. At the top of the cell is another traveling bridge for a remotely operated crane. At one end of the cell is an air lock that connects with the maintenance area. The erane can move from the bridge in the cell to a monorail in the air lock. Closed-circuit television equipment is provided for viewing the mainte- nance operation in the cell. A number of cameras are mounted to show the operation from different angles, and a periscope gives a direct view of the entire cell. | 17—4. MoOLTEN-SALT TRANSFER L QUIPMENT The fuel-transfer systems are shown schematically in Fig. 17-7. Salt freeze valves (see Section 15-3) are used to isolate the individual com- ponents in the fuel-transfer lines and to isolate the chemical plant from the components in the reactor cell. With the exception of the reactor draining operation, which is described below, the liquid is transferred frqm one vessel to another by a differential gas pressure. By this means, fuel may be added to, or withdrawn from, the reactor during power operation. The fuel added to the reactor will have a high concentration of UF4 with respect to the process fuel, so that additions to overcome burnup will re- quire transfer of only a small volume; similarly, thorium-bearing molten salt may be added at any time to the fuel system. The thorium, in addi- tion to being a design constituent of the fuel salt, may be added in amounts required to serve as a nuclear poison. For the main fuel drain circuit, bellows-sealed, mechamcally operated, poppet valves (see Section 15-3) will be placed in series with the freeze valves to establish a stagnant liquid suitable for freezing. Normally these mechanical valves will be left open. By melting the plug in the freeze line and opening gas-equalization valves, the liquid in the reactor will flow by gravity to the drain tank, and the gas in the drain tank will be transferred to the reactor system. Thus gas will not have to be added to, or vented from, the primary system. [cHAP. 17 CONCEPTUAL DESIGN OF A POWER REACTOR 692 Buibiny) §|pg jau| waysAG uipaqg 2 ||id VIDW $§NPOI{ Uoissi{—d uoidaaq MO |PUWION e JUSA—A A|ddng—g SAIPA 823343 A |- SA|PA Eo:&owlfim' aun spoH)— - ——— aur [ong :pusban "WI)SAS JOJSUBI} J[BS [ONJ JO WEBISLIP JBWAYIS °2~LT "BI] Ad ||||||||||||||||||||||||||||||||||||||||||||||||| ; (V4n—C4eg—d41) (V4y1-Cieg—4n ! i Jojpunion|4 AL £ o rfllfi ) 1abupydx3 JOOH - junl |[PMBIPYH M juo|d { Buissadoay CH PR ) 1 Y Joyduug Jayouug gezh 4l 10Op3Y | A= A I — | ey ’ m ;. 1 {Ad— Mm;lwym;lml 17-6] CHEMICAL REPROCESSING METHOD 693 17-5. FueL DRraIN TANK For the drain vessel design calculations, it was assumed that at 1200°F the fuel system volume would be 600 ft3. The design capacity of the drain vessel was therefore set at 750 ft3 in order to allow for temperature excursions and for a residual inventory. An array of 12-in.-diameter pipes was selected as the primary containment vessel of the drain system in order to obtain a large surface area-to-volume ratio for heat-transfer effi- ciency and to provide a large amount of nuclear poison material. Forty- eight 20-ft lengths of pipe are arranged in six vertical banks connected on alternate ends with mitered joints (see Fig. 17-8). The six banks of pipe are connected at the bottom with a common drain line that connects with the fuel system. The drain system is preheated and maintained at the de- sired temperature with electric heaters installed in small-diameter pipes located axially inside main pipes. These bayonet-type heaters can be re- moved or installed from one face of the pipe array to facilitate maintenance. The entire system is installed in an insulated room or furnace to minimize heat losses. The removal of the fuel afterheat is accomplished by filling boiler tubes installed between 12-in.-diameter fuel-containing pipes with water from headers that are normally filled. The boiler tubes will normally be dry and at the ambient temperature of about 950°F. Cooling will be accomplished by slowly flooding or “‘quenching” the tubes, which then furnish a heat sink for radiant heat transfer between the fuel-containing pipes and the low-pressure, low-temperature boiler tubes. For the peak afterheat load, a flow of about 150 gpm of water is required to supply the boiler tubes. The design criteria for this fill-and-drain system are that it always be in a standby condition immediately available for drainage of the fuel, that 1t be capable of adequately handling the fuel afterheat, and that it provide double containment for the fuel. The heat-removal scheme is essentially self-regulating in that the amount of heat removed is determined by the radiant exchange between the vessels and water wall. Both the water and the fuel systems are at low pressure, and a double failure would be required for the two fluids to be mixed. The drain system cell may be easily enclosed and sealed from the atmosphere because there are no large gas-cooling ducts or other major external systems connected to it. | 17-6. CHEMICAL REPROCESSING METHOD In this plant it is assumed that the core and blanket salts will be re- processed by the fluoride volatility process [1] to remove UFs. The UFg will be reduced by a fluorine-hydrogen flame process [2] and returned to ‘the reactor core. The blanket salt with its U233 removed will be returned to the blanket, since the buildup of fission products in the blanket salt will 694 CONCEPTUAL DESIGN OF A POWER REACTOR . [cHAP. 17 Fia. 17-8. Drain and storage tank for fuel salt of molten salt power reactor. not be an appreciable poison for many years. For purposes of the cost study, it is assumed that the spent fuel salt with its contained fission prod- ucts will be discarded after each processing, and new fuel salt will be pro- vided. Since several means of recovering the fuel salt appear to be feasible, this is a conservative assumption. The chemical processing is assumed to be an integral part of this reactor plant. 17-7. CosT ESTIMATES All the costs were calculated for a 260-Mw net-electrical-output plant operated at a load factor of 0.80. An annual charge of 149 was made for all capital items, other than the uranium inventory, for which a 4% annual charge was made. It is assumed that the reactor plant components will have been engineered and developed with separate funds not included in this estimate. 17-7] COST ESTIMATES 695 Table 17-3 gives the fuel cycle costs separately from the other reactor costs, so that a comparison can be made with comparable costs for solid- fuel-element reactors. TABLE 17-3 FueL .CycLeE Costs Operating costs (fuel cycle only) U235 consumed Fuel salt makeup (or recovery cost) Chemical plant operation (listed as part of Operation and Maintenance in Table 17-5) Caprtal charges (fuel cycle only) U233 and U235 inventory, 1000 kg at 49, Chemical plant and equipment $3,000,000 at 149, (included in capital costs in Table 174 and fixed cost in Table 17-5) Total fuel cycle cost $/year 2,260,000 760,000 500,000 3,520,000 680,000 420,000 1,100,000 4,620,000 mills/kwh 1.24 0.42 0.28 Do ot N | An estimate of the capital costs for the reactor plant is given in Table 17—4. This breakdown leads to the estimated power costs in Table 17-5. TABLE 174 CaritaL CosTts Land and land rights Structures and improvements Reactor system (including chemical plant) Steam system Turbine-generator plant Accessory electrical equipment Miscellaneous power plant equipment Direct costs subtotal Contingency subtotal General expense Design costs Total cost $500,000 7,500,000 20,039,000 3,750,000 11,750,000 4 600,000 1,250,000 49,389,000 10,216,000 7,500,000 2,450,000 $69,555,000 696 CONCEPTUAL DESIGN OF A POWER REACTOR [cHAP. 17 Note that in Table 17-5 the chemical plant capital and operating costs listed in Table 17-3 are part of the fixed cost and operation and mainte- nance, respectively, so that the amount listed for fuel charges is only 2.03 TABLE 17-5 PowEeEr CosTs Item Annual charge malls/kwh Fixed cost $9,738,000 5.34 Operation and maintenance 2,700,000 1.48 Fuel charges 3,700,000 2.03 Total annual charge $16,138,000 Total power cost 8.85 mills/kwh. This breakdown is in keeping with the philosophy of regarding the chemical reprocessing as an integral part of the reactor plant. The difference in cost between having the reactor on standby and having it on the line is less than 2 mills/kwh. REFERENCES 1. G. I. CatuErs, Uranium Recovery for Spent Fuel by Dissolution in Fused Salt and Fluorination, Nuclear Sct. and Eng. 2, 767 (1957). 2. S. H. SmiLey and D. C. BraTter, Conversion of Uranium Hexafluoride to Uranium Tetrafluoride, Progress in Nuclear Energy, Series III, Process Chem- istry, Vol. II. New York: Pergamon Press, in preparation. 697 BiBLioGRAPHY FOR PART II Conceptual design studies of molten-salt power reactors: BuLMER, J. J. et al.,, Fused Salt Fast Breeder, USALEC Report CF-56-8- 204(Del.), Oak Ridge National Laboratory, 1956. Davipson, J. K. and W. L. RosB, A Molten-salt Thorium Converter for Power Production, USAEC Report KAPL-M-JKD-10, Knolls Atomic Power Labora- tory, 1956. Davies, R. W. et al., 600 Mw Fused Salt Homogeneous Reactor Power Plant, USAEC Report CF-56-8-208(Del.), Oak Ridge National Laboratory, 1956. WEHMEYER, D. B. et al., Study of a Fused Salt Breeder Reactor for Power Production, USAEC Report CF-53-10-25, Oak Ridge National Laboratory, 1993.