UNCLASSIFIED OAK RIDGE NATIONAL LABORATORY QOperated By UNION CARBIDE NUCLEAR COMPANY 0 R N L (44 POST OFFICE BOX X (ENTRAL F"-ES NUMBER OAK RIDGE, TENNESSEE X822 ........ 58-2-46 6S DATE: February 5, 1958 COPY NO. SUBJECT: A MOLTEN SALT NATURAL CONVECTION EXTERNAL TRANSMITTAL REACTOR SYSTEM AUTHORIZED TO: Iisted Distribution FROM: F. B. Romie, AMERTCAN-STANDARD, and B. W. Kinyon £ Q . NOTICE This document contains information of a preliminary nature and was prepared primarily for internal use G at the Oak Ridge MNationel L.aboratory. It is subject e : tc revision or correction and thersfore does not ‘ represent a final report. UNCLASSIFIED w5 oot comtractor of the Commissicn to the extent that such employee or contractor grepares, handles This report was prepared os an acceunt of Gevernment sponsored work, Neither the United States, nor the Cbmmission,. oy any perébfl acting on behalt of the Commissiont A, Makes any worransy or representation, express or implied, with respect to the acecuracy, completeness, o usefulness of the informotion contained in this report, or thet the use of eny information, opperatus, methed, or process disclosed in this report may not infringe privately owned rights; or B. Assumes any [iabilities with respeci to the use of, or for domages resulting from the use of 6ny informetion, eppurates, method, or process disclosed in this report. As used in the above, “‘person acting or beholf of the Commission® inciwdes any employee or or distributes, or provides cccess to, any infeemation pursuent to his employmant or coniract with the Cemmission. i " UNCLASSIFIED ii. o Abstract Fuel-salt volumes external to the core of a molten-salt reactor are calculated for a system in which the fuel salt circulstes through the core and primary exchanger by free convection. In the calculation of these volumes, the exchanger heights above the core top range from 5 to 20 ft. Coolants considered for the primary exchanger are g second molten salt and helium. External fuel holdup is found tc be the same with either cooclant. Two sets of terminal temperatures are selected for the helium. The first combination permits steam generation at 850 psia, 90001?e The second set is selected for a closed gas turbine cycle with an llOOOF turbine inlet temperature. Specific power (thermal kw/kg 235) is found to be about 900 kw/kg, based on initial, clean conditions and a 60 Mw (thermal) output. A specific power of 1275 kw/kg is estimated for a forced convection system of the same rating. UNCLASSIFIED UNCLASSIFIED 1. A MOLTEN SALT NATURAL CONVECTION REACTOR SYSTEM F. E. Romie¥ B. W. Kinyon Introduction One of the problems of a circulating-ligquid-fuel reactor is the provision of reliable, long-lived fuel circulating pumps. This problem is eliminated for a gsystem in which the fuel is circulated through the primary exchanger and reactor core by natural convection. The advantages of omitting the circulating pump and its attendant problems of meintenance and replacement are purchased at the price of ificreased fuel-salt volume in the primary exchanger and in the convection risers. There are applications for a resctor system in which the premium placed on reliability and ease of maintenance could meke the convection system attractive. The purpose of this report is to investigate a free-convection reactor system in crder to afford a basis for assessing its merits. Seleefiion of Reactor Conditions A schematic representation of the reactor and primary heat exchanger is shown in Figure (1). The fuel salt returned to the core after cooling in the heat exchanger enters the bottom of the reactor sphere and leaves through the top. The temperature of the fuel salt entering the heat exchanger is specified to be 1225°F, & temperature which available corrosion information indicates is consistent with long-term life for the system. With this temperature and a thermal output of 60 Mw, it is possible to consider selection of steam turbine conditions of conventionsl plants. Thus tem- perature conditions in the fluids passing through the primary exchanger have been * On loan from AMERICAN-STANDARD, Atomic Energy Division, February 1957 to February 1958. URCLASSIFIED CEen it o L pod UNCLASSIFIED 2. selected to permit generation of 850 psia, QOOOF stesm. For a thermal output of 60 Mw these steam conditions would give a generator output of about 22 Mw (37% efficiency). The IQESOF maximum sglt temperature also allows consideration of a closed gas turbine cyecle using helium at a turbine inlet temperature of 1100°F. For this tem- perature and the postulated cycle parameters summsrized in Table I# the turbine output would be about 19 Mw (30.8% efficiency). The importance of desirable nuclear properties for the fuel salt takes precedence over the thermal properties in the selection of & fuel salt. For this reason a mixture of lithium and beryllium fluoride salts is selected. For such mixtures the'viscosity increases and the melting point decreases with increasing beryllium content. Mixture 130 (63% TiF, 36% BeFé;:vl% UF) on a mole % basis), with a melting point of 850°F, has been selected as giving a reasonable combination of these two properties. The pertinent physicel properties of this salt are given in Table II. Table I. ASSUMED GAS TURBINE CYCLE PARAMETERS Turbine inlet temperature 11000F Compressor inlet temperature (both stages) 1000F Compressor adisbatic efficiency 874 Turbine adiabatic efficiency 89% Regenerator effectiveness 884 Compressor pressure ratio (1.52/stage) 2.3 Pressure losses, I AP % P Calculated Performance - Helium Thermal efficiency 30.8% Helium temperature at salt exchanger inlet 6760F Regenerator NTU T-33 Cycle output decreases 2.1% for each percentage point increase in AP P __________ 4 * Attainment of the helium compressor and turbine efficiencies postulsted in hid Table I has yet to be demonstrated. UNCILASSTFIED T g BT e £ ;,‘”3 A e U b 0 o 9 URCLASSIFIED 3. The secondary fluid in the primary exchanger is limited by compatibility con- (13 siderations to either a gas or a molten salt. Both possibilities are treated. In the case of the molten salt cooling the heat from the reactor would be trans- ferred from the fuel salt, to the coolant salt, tc sodium, to water. A similar system is described ir Reference 1. Good compatibility with the fuel salt and low fusion temperature {6SOOF) are the principal reasons for selection of Mixture 84 (35% 1iF, 27% NaF, 38% BeF,) as the coolant salt. Table TIT. PHYSICAL PROPERTIES OF SALT MIXTURES 130 AND 84 Mixture 130 Mixture 8h Unit heat capacity, Btu/Ib°F 0.62 C.59 Thermal conductivity, Btu/hr-ft-CF 3.5 3,2 Density, Ib/ft° (t - OF) 136.4-0.0121t 1%9-0.0142t Viscosity, Ib/hr-ft, 1000°F: 36.6 29,1 1200°F: 19.2 1k,2 Fusion temperature, °p 850 6L0 The use of a gas as a cocolant offers the advantage that, in the case of a steam cyele, it would be the only fluid intermediate between the fuel salt and the steam. In the case of the gas turbine éycle there Would be no fluid inter- mediate between the working fluid and fuel salt. The use of gas rather than salt also eliminates heaters for the coolant salt and sodium circuits and decreases the number of drain tanks required for the system. Gases sultable as a coolant are hydrogen; helium and nitrogen. Helium is specifically considered in the following because of its good heat transfer properties and nuclear and chemical inertness. Hydrogen would be preferable from the point of view of heat transfer and pumping UNCLASSIFIED Sa 05 g UNCIASSIFIED \ considerations, while nitrogen, due to its higher molecular weight, would be attrac- tive for use with a gas turbine cycle. The low moclecular weights of hydrogen and helium require the use of a large number of compressor and turbine stages. Nitrogen would offer the advantage of using presently-developed compressor and turbine equipment. Optimum Riser Conditions The hydrostatic differefntisl hesd causing the fueli-sglt flow ig proportional to the product of the temperature change of the fuel salt and the height of the exchanger above the reactor. The pressure drop available to the exchanger is thg hydrostatic head minus frictional losses oecurring in the riser pipes. These fric- tional losses decresse rapidly with increasing riser diameter, but at the expense of an increased salt hcldup volume in the riser. If the fuel-salt mass rate and temperature change are both fixed, then, for the attainment cf a specified pressure drop available to the exchanger, there is one combination of riser diameter and height of exchanger for which the salt volume in the riser system will be a minimum. The exiStefice of this minimum is illustrated in Figure (2), in which the variation of salt volume in the riser and of riser diameter are shown as a function of exchanger height. Tablé ITT summarizes 8 set of optimm riser conditions used in this reporta Flgure (2) and Table IV are based on the riser friction data given in Table III. The frictional losses in the riser are found to be determined primarily by the expansion and contraction losses and are insensitive to wall friction. Thus replace- ment of a single riser by two sete of risers having an equal height and total cross sectional ares will make essentially the same pressure drop available to the exchanger at the same riser holdup volume. This fact can be of use in decreasing the mechanical rigidity of the piping and indicates that the use of two exchangers with separate risers in place of one will not require an increased fuel holdup volume in the risers. G [BO6 UNCILASSIFIED UNCLASSIFIED G Table TII. FRICTION IOSSES IN THE RISERS Expansion plus contraction loss on entering and leaving core L.5 ft on entering and leaving exchanger headers 1.5 £t Equivalent length of pipe added for bend~loss allowance 8 ft Flow friction factor .02 Tength of piping 2H + 17 ft Priction loss asscciated with risers =(o.02[8+17+23)+3 we D (%)2 D Z2gp Hydrostatic head for flow = Ofl@f (fi + %) Salt volume in riser = xpF (17 + 2H i Part Load Operation For a given core-riser-exchanger system the flow rate of the fuel salt is determined in terms of the difference in fuel salt temperature in the two legs of the riser and the flow frietion characteristic of the system. The part load operation of a given system is shown in Figure (3). For the molten salt reactor the average of the core inlet and exit temperatures is effectively a constant in- dependent of heat output. Thus, if at rated conditions the salt temperature entering the exchanger (leaving the core) were 12250F, the salt temperature leav- g ing the core at half-load would be, based on Figure (3}, ll870F, and the temperature of the salt returned to the core would be 1059OF compared to lOOOOF at full-losad. Gel o 0o UNCLASSTFIED £ s gy 500 % LASSIFIED Temperature change of fuel salt, OF Exchanger height, ft Salt volume in risers, ft5 Diameter of riser, Tt Pressure drop across exchanger, lb/ £t2 Salt velocity in riser, ft/sec 200 65 1.78 1.5 OPTTMUM RISER CONDITIONS 200 10 19.5 1.656 13.4 1.74 Q = 60 Mw a = 0.,0121 200 105.5 1.537 27.8 2.02 Table IV. 1b 2 r£=-Op 225 55 1.602 Tl 106)"' 225 10 68 1.525% 15.6 1.79 225 20 89 1.418 31.8 2.10 250 L8 1.497 8.8 1.69 250 10 58.5 1.41 17.8 1.90 UNCLASSITFIED 250 20 7T 1.%18 36.0 2.18 UNCLASSIFIED Salt-Cooled Heat Exchanger S The términal temperatures selected for the coolant salt are 8750F at inlet to the primary exchanger and 10250F at exit. These two temperatures remain un- changed for all salt-cocled exchanger designs considered. The BTSOF entrance temperature is 25°F above the fusion temperature of the fuel galt and QBSOF above the fusion temperature of the coolant salt. Heat transfer data used in the cal- culations are given in Table V. Table V. PRIMARY HEAT TRANSFER AND FRICTION DATA Fuel Salt Nusselt modulus for fuel salt* k.0 Friction factor 64/Re Entrance and exit losses for salt, velocity heads 1.5 Thermal resistance per unit tube length 1 NurziR8 = 0.0227 Coolant Salt Thermal resistance per unit tube length¥* 0.0032 Exchanger Tubes Thermal conductivity, Btu/hr-ft-°F 1.5 Wall thickness, in. 0.059 Thermal resistance per unit length: 0.42 in. ID tubes 0.00272 0.634 in. ID tubes 0.00189 ¥ (QGraetz modulus is less than 5¢0 for all designs. *¥% This value is estimated attainable without excessive pressure loss by appropriate shell-side baffling. (Doubling this resistance would in- crease the over-all resistance by only 11%.) ' Figures (4) and (5) summarize the results of exchanger calculations. In- spection of these figures shows that increasing the height of t@e exchanger above the reactor has no effect on the total length of exphangef tubing required and o UNCLASSIFIED SR e ] 5 Ty Q b e s GOY UNCLASSIFIED ] thus no effect on the fuel-salt volume within the exchanger tubes; however, in- creésing the exchanger height leads to an increased length of the exchanger and a decrease in the number of tubes in the exchanger. Thus the tube bundle diameter is decreased and, consequently; the volume of fuel salt contained in the exchanger header. However, the net result on salt volume external to the reactor is a slight increase with increasing exchanger height. This increase is due to the increased sglt volume in the risers. If the number of exchanger tubes were the constant parametef rather than the tube diameter, then an increase in exchanger height would lead to a decrease in total holdup volume and a smaller diameter tube. Decreasing the internal diameter of the exchanger tubes from 0.63L4 to 0.42 in. produces, for a 10 ft height of the exchanger, about a 23% decrease in external holdup volume, but at the expense of a 2.3-fold increase in the number of tubes. Variation of the fuel-salt temperature drop over the range from 200 to 250°F PTro- duces a relatively minor effect on the exchanger designs. The low thermal resistance of the coolant salt relative to that of the laminarly-flowing fuel salt results in a small temperature difference hetween the exchanger shell and tubes. Thus thermal stresses due to differential expan- sion of t;bes and shell should prove small enough to permit & straight tube design despite the rather large temperature difference between the two salts. Gas-Cocled Exchangers for Steam Cycle For helium cooling of the primary exchanger, the terminal temperatures of the helium were fixed at BSOOF for inlet to the exchanger and lOZSOF for outlet. The drop in fuel-salt temperature wae held at 225°F. The 850°F inlet temperature insures that the fuel salt will not freeze ifi the tubes, and the 1025°F exit tem- perature is consistent with the generation of 900°F steam. The exchanger designs, s UNCLASSIFIED g Y Sred LG d) UNCLASSIFIED 9. which are summarized in Table VI, are based on the use of 0.634 in. ID tubing with copper fins. Dimensions of the finned tubes (Table VII) were scaled-up from descriptions in Reference 2, from which the flow fractiéfi and heat transfer data wére also obtained. With a single-pass cross flow exchanger, the salt flowing in the row of tubes first contacted by the entering helium would leave the exchanger at a'considerably lower temperature than that in the last row of tubes. With laminar flow, the flow resistance is directly proportional to the viscosity. Thus the salt velocity in the first row would be less than in the last row, and the mean temperature of the first row conseqfiently lower than would be calculated cn the basis of uniform flow in all tubes. The effects leading to non-uniform salt flow are mutually aggravating, with the result that a single-pass cross flow exchanger is unacceptable. A multi- pass exchanger is thus required if a cross flow exchanger design is to be used. A counterflow exchanger has the optimum configuration, both from the point of view of obtaining uniform flow distribution and minimizing the required over-all conduc- tance of the exchanger. | The over-all dimensions given for the helium-cooled exchanger gorrespond to a three-helium-pass cross flow exchanger. The column labeled "depth" refers to the distance traveled by the helium in one pass (i.e., distance from front to rear of tube array}. The "width" refers to the transverse dimension of the exchangers. The fractional pressure loss, AP/P, listed in the table is based on & helium pressure level at the exchanger of 100 psia. The fractional pressure loss is in- versely proporticnal to the square of the pressure level. Thus, for example, if the pressure of the helium were increased to 200 psia, the AP/P values shown ifi Table VI would be quartered (without change in any other entries in the table). UNCLASSIFIED C_:" B oy mE 1y ELE For g UNCIASSIFIED 10. ......... B Table VI. GAS-COOIED EXCHANGER - STEAM CYCIE 0.634k ID Finned Tubing Total Iength AP Tube Salt Frontal Area Helium of tubing P Depth Volume per pass Re No. £L (100 psia He) ft £t £2 500 k8,300 .0000758 264 106 1060 1000 43,000 000438 469 95.1 530 2000 39,200 .0026 .855 86 265 LOCo 36,200 0177 1.582 79.3 132 _ Iength per Header Total Salt Helium Number of tube Volume Yolume Width Re No. Tubes £t £43 ££3 £t 5 ft high: riser vol - 55 cu ft, AP = 7.k psf 500 5520 8.75 2l 185 410 1000 5230 8.22 23 173 194 2000 5000 7.8k o2 163 102 LO00 4800 7.5h 21 155 5k 10 £t high: riser vol - 68 cu ft, APsalt = 15.6 psf 500 3800 12.7 16.7 191 250 1000 3600 12.0 15.8 179 133 2000 3440 11.4 15.1 169 69.4 4000 3310 10,9 14.5 162 36.2 20 £t high: riser vol - 89 cu ft, AP 1y = 31.8 psf 500 2660 18.2 11.7 207 175 1000 2520 17.1 11.1 195 93,2 2000 2k10 16.3 10.6 186 48.8 4000 2320 15.6 10.2 178 25.4 UNCIASSIFIED et e e €3 €2 Hae o b éj i e UNCLASSIFIED 11. Table VII. DESCRIPTION OF FINNED TUBE* Internal tube diasmeter, in. 0.63h Wall thickness, in. 0.059 Tube material INOR-8 Fin material Copper Fin area/total area 0.876 Fin thickness, in. 0.03%9 Tube spacing in plane normal to gas flow, in. | 1.745 Tube spacing parallel to flow, in. | 1.4%0 Heat transfer area/unit volume, 1/ft 76.2 * Geometrically similar to surface CF-8.72 (c) in Reference 2. The fraction, s, of the plant output used in pumping the helium through the exchanger is given by the equation, in which 7y 1s the specific heat rate, T +the mean gas temperature (OR), AT the temperature change of the gas, Mo the blower efficiency, and np the plant efficiency. For a blower efficiency of 80% and plant efficiency of 37%, one obtains, AP S - 1097“"?"‘ Thus if 2% (8 = .02) of the generator output is used to blow helium through the exchanger, the value of AP/P would have to be 0.00187. In order to facilitate discussion of the calculated results summarized in Table VI, it is convenient to define the reference heat exchanger described in Table VIII. UNCLASSTIFIED IS e LR . e b L . x ] R Y @ £oe ’ i ok UNCLASSIFIED 12. Table VIII. e P REFERENCE GAS~COOLED HEAT EXCHANGER AP/P 0.00187 Pregsure level, psia 100 Exchanger height, ft 10 Tube length, ft 11.5 Number of tubes 3240 Iength of tubes, ft 8 Total salt volume, ft3 172 Depth, in. 9 % Helium Reynolds Modulus 1750 If the pressure level of the helium were increased from 100 to 200 péia and the exchanger altered in order to maintain AP/P constant, then the exchanger length would be decreased from 78 ft to 48 ft and the depth increassed from 0.77 to 1.2 ft. Accompanying changes in the number of tubes, tube length and salt holdup volume would be relatively negligible. If the power expended in pumping the 100 psia helium through the exchanger were incressed from 2% to 4% of generator output, (i.e., %? = 0.00%74), the depth and length of the exchanger would change to 0.93 ft and 61 ft, respectively. Corre- sponding changes in salt holdup volume, number of tubes and tube length would be relatively negligible. For a 10 ft height of the exchanger above the reactor, the calculated salt fiolume external to the core is 172 cu ft for the reference exchanger. This can be compared to & fuel-salt volume of 160 cu ft for the salt-cooled exchanger of the same height, using exchanger tubes of the same internal diameter. If the reactor core diameter is taken as 8 ft, corresponding to a fuel-salt volume of UNCIASSIFIED Gal o UL4 URCLASSIFIED 13, 268 cu ft, then these figures indicate that gas cooling can be used with a fuel~ salt volume increase of 12 cu ft in a total volume (exclusive of volume in expan- sion tank) of 440 cu ft. The major uncertainty in comparing the externsl fuel-salt volumes for the gas-céoled and salt-cooled exchangers lies in the uncertainty in the salt volume which should be attributed to the exchanger headers. TFor the present calculations the header volume hae been assumed to be 0.00438 £t per 0.634 in. ID tube for both the salt-cooled and gas-cocled exchangers. A more detailed design study is required to fix the header volume more accurately. Use of a smaller tube diameter for the gas-cooled exchanger would result in a8 decreased fuel-sglt holdup volume at the expense of an increased number of shorter tubes. The effect of using a 0.42 in. ID tube instead of the 0.63L4 in. tube used in the calcuiations shcould parallel the behavior previously shown for the salt-cooled exchanger. A possible configuration for the three-pass exchanger is shown in Figure (6). The configuratién permits attachment of the riser pipes to the headers with suffi- cient flexibility to minimize thermael stresses and provides a cylindrical containment for the pressurized helium. Use of two such cylinders to contain the reference exchanger would result in twe exchangers, each 19 1/2 ft long. If the helium pres- sure were ipcreased to 200 psia, the length of each would be reduced to 12 ft. Gas-Cooled Exchangers for Gas Turbine Cycle The '‘major problem encountered in designs of exchangers for use with a gas turbine cycle is the low temperature of the helium entering the helium~-fuel-salt exchanger. For the cycle parameters given in Table I, this temperature is 676°F, & temperature lThOF below the fusion temperature of the salt. Exploratory calcu- lations indicate that this low temperature would freeze the salt if an exchanger uging the finned surface described in Table VII were used. e . UNCTASSIFIED chin (i1 % UNCIASSTIFIED 1k, For minimum fuel-sglt holdup in the exchanger tubes, the optimum solution ............... to the problem of meintaining a wall surface temperature above the freezing point is found in the use of a counterflow exchanger, in which the gas-side thermal resistance decreases in the direction of gas flow .in such a msnner that the wall temperature does not drop below a prescribed value. In the csse of a longitudi- nelly finned tube, the gas-side thermal resistance variation required could be effected by increasing the fin height along the tube until the full fin height is gttained. For the remainder of the tube length, the fin height would be un- changed and the tube wall temperature would increase above the prescribed minimum. The exchanger calculations presented for the gas turbine cycle presume the use of such an exchanger with a minimum wall-fuel-salt interfacial temperature of 9OOOF. Results of the exchanger calculations are presented in Figuref?}-e The abscissa, ¢ , is the ratio of the sum of the gas and wall thermal resistances per unit tube length to the fuel-salt thermal resistance per unit length, = Rw + Rg . The gas-side thermal resistance is defined as that which is R obtainedswith the unaitered finned tubing. Using the thermal resistance data ¢ presented in Table V, it is easily shown that the gas-side thermal resistances OF-ft-hr Btu 8 l/z-fold increase in gggwside registance corresponds to an increase of approx=- corresponding to ¢ values of 0.25 and 1.5 are 0.003785 and 0.0321 . This imately 40% in external fuel-salt holdup volume, number of tubes and length per tube. No data are presently available defining the hesat transfer and flow friction characteristics of tube bundles using longitudinally finned tubes. Thus the exchanger designs are incomplete. Hafiever, the number of tubes, tube length and total external salt volume can be estimated with reasonsble accuracy if it is assumed, as seems likely, that the same value of ¢ can be realized with longitu- dinally finned tubes as with circumferentially finned tubes. For the steam cycle R UNCLASSIFIED UNCLASSIFIED 15. helium-cocled reference exchanger the value of ¢ 1s C,47. Using this value of ¢ in Figure (7} gives, for an exchanger 10 ft above the reactor, a total salt holdup volume of 160 £t°, s total length of exchanger tubing of 36,200 ft, with 3300 tubes 11 ft long. A possible configuration for the exchanger is shown in Figure (8). Comparison of Cooling Means Teble IX gives a compariscn of the three cooling systems considered for the conditions listed at the head of the table. The difference in external fuel-salt holdup volume among the three systems is not large and is therefore not a determin- ing factor in the selection of the coolant medium for the reactor. In genersl, the helium-cooled exchangers have much larger over-all volumes than the salt-cooled exchangers. This i;creased bulk is caused by the larger spacing required by the finned tubes and also by the large volume reguired by the helium headers. An increase in helium pressure will decrease the helium header volune and also should decresse the fuel-salt header volume. The fipper limit on helium pressure is prcbably set by consideration of the effects of a tube rupture on the fuel-salt system. It is of interest to note that other gas-cooled reactor stfidiesf apparently not limited by similsr considerations; have recommended gas pressures as high as 1000 psia. The 100 psia pressure specificslly considered in these calculsations is probably conservstively low. Table IX. COMPARISON OF COOLING MEANS Based on 60 Mw, 10 ft high exchanger - OF 0.634 ID tubes, AT, o7 = 225°F Method of Cooling No. of Tubes Total Tube Iength, £t Vol Outside Core, ftE Salt-cooled for steam 3150 36,000 161 generation Helium~-cooled for 3240 37,200 172 - steam generation e Helium-cooled for gas 3300 36,000 16% turbine cycle en o o UNCIASSIFIED Py ML UNCLASSIFIED 16. Comparison with Forced Convecticon System ........ Design of a forced convection resctor system (1) has led to an estimation of 0.56 cu ft of fuel salt externsl to the reactor core per thermal megswatt. For the free convection system, using C.634% ID tubes in the primary exchanger, the externsl volume is about 160 cu ft for 650 Mw, giving a specific volume of 2.67 cu ft/Mw. Using these numbers and initial-clean critical mass data (5), the fuel inventory for 60 Mw minimizes, for both the free and foreed convection systems, at a core diameter of about 8 ft. At.this diameter the specific powvers {kw thermal/kg 235) are 895 and 1275 kw/kg for the free and forced convection systems, respectively. The free convection system is thus estimated to require Lo% more inventory than the forced convection system. Addition of a small amount of thorium to the core salt would produce some breeding and a longer time intervsl between fuel additions. One-quarter percent thorium requires sbout an 87% increase in fuel inventory at a core diameter of & ft. UNCLASSIFIED . Sk T s oy 0 b L2 k.) &:} & (__‘} UNCLASSIFIED 17. References 1. Kinyon, B. W., and Romie, F. E., "Two Power Qeneration Systems for s Molten Fluoride Reactor"”, To be presented at the Nuclear Engineer- ing and Science Conference of the 1958 Nuclear Congress (March), Chicago, Illinois 2. Kays, W. M., and london, A. L., "Compact Heat Exchangers", The National Press, Palo Altc, California (1955) 3. Personal communiecaticn from J. T. Roberts, ORNL UNCLASSIFIED UNCLASSIFIED ORNL~LR-DWG 27943 EXPANSION DOME HEAT EXCHANGER 1COOLANT 1 DOWNCOMER Fig.1. Schematic of Natural Convection Reactor. 18 el 020 UNC-LR-DWG 27944 19 et UNC-LR-DWG. 27945 LR £LERER 20 68 ok r{j p:i Fa FIGURE 4 UNC-LR-DWG. 27946 ......... e 2 I0 12 20 HEIGHT OF EXCHANGER —FT iy poo 21 W Ued 50 40 30 20 FIGURE 5 UNC-LR-DWG. 27947 TEMP CHANGE OF FUEL SALT 250 °F 225 200 TOTAL LENGTH OF TUBING x 103 FT (BOTH 0.42 & 0.634 1.D. TUBING) 250 634 1.D, ,/ // / s ?zoo ;/ = \ EXCHANGER TUBE LENGTH,FT _ 250 - 4225 - ~200 0 20 EXCHANGER HEIGHT,FT ¢e R Figure 6 SALT HEADER Three-Pass Exchanger UNCLASSIFIED GRNL-LR- DWG 27948 ~PRESSURE SHELL FIGURE 7 UNC-LR-DWG. 27949 - " | mm UNCLASSIFIED ORNL-LR-DWG 27950 ac A K‘“ji = ifii ) éj Fig. 8. Schematic of Counterflow Salt to Helium Exchanger.