EAS N AND F DESIG OAK RIDGE SCHOQL OF REACTOR TECHNOLOGY 53'77 t%% h T ims cxorument consist ol 1 9\3 Lo o 2O of B3 A9pas; &Maaafiimifi Reactor Design snd Feasibility Study "HIGH PERFORMANCE MARINE REACTOR" Prepapred by: K. H. Dufrane, Group Leader T. G. Barnes C. Bicheldinger W. D. Lee N. P. Otto C. P. Patterson T. G. Proctor R. W. Thorpe R. A, Watson August 1957 26, 28, 29, 30. 31. 32. 33. 34, 36. 37 38. 39. 40 '”ll'l . A. M. Weinberg J. A. Swartout R. A. Charpie W. H. Jordan Lewis Nelson D. C, Hamilton E. P. Blizard L. B. Holland W. B, Kim'ley W. R. Gall A. P, Frags J. A. Lane P. R. Kasten A. F. Rupp E. 8. Bettis C. E. Winters R. B. Briggs A. L., Boch W. T. Purgerson R. V. Meghreblian L. R. Dresner K. H. Dufrane T. G. Barnes C. Eicheldinger W. D. Lee N. P. Otto C. P. Patterson T. G. Proctor R. W. 'I'hOl’pe - A, Watson ho-h3, REmp Library 4hohs. 4666, O7-72. ORSORT Files 73-87. TIsE 88. P. P. Eddy, Maritime Reac Martin Skinner Central Research Library Laboratory Records PREFACE In Septenber, 1956, a group of men experienced in various scientific and engineering flelds embarked on the twelve months of study which culmi- nated in this report. For nine of those months, formal classroom and student laboratory work occupled their time. AT the end of that period, these nine students were presented with a problem in reactor design. They studied it for ten weeks, the final period of the school term. This is a summsry report of their effort. It must be realized that, in so short s time, a sgtudy of this scope can not be gusranteed complete or free of error. This "thesis" is not offered as a polished engineering report, but rather as & record of the work done by the group under the leadership of the group leader. It is issued for use by those persons competent to assess the uncertainties inherent in the results obtained in terms of the preciseness of the technical data and analytical methods employed in the study. In the opinion of the students and faculty of ORSORT, the problem.has served the pedagogical purpose for which it was intended. The faculty Joins the authors in an expression of appreciation for the generous assistance which various members of the QOak Ridge National Laboratory gave. In particular, the guidance of the group consultant, A. P. Frags, is gratefully acknowledged. Lewis Nelson for The Faculty of ORSORT W, ABSTRACT For marine applications & circulating fuel, fused fluoride salt reactor gsysten appears to qffer'a sfibStantially reduced specific weight (1bs per shaft horsepower) ovér'current and planned reactor systems, Such a weight reduction would make nuclear power'feasible.forfsurface ships smaller than 7500 tons displacement, the cufrent_mifiifiufi for.pfesent and proposed reactor systems, as well as overall performance improvements for larger vegsels, Keeping within the bounds of currently available technology and proven practices, reactor-steam system capsble of developing 35,000 SHP with an overall specific weight of approximately 6.4 1bs/SHP is indicated, The partieulaf installation of this sysiem aboard a 931 class destroyer of 3-4000 tons displacement was found feasible, When this system is used in conjunction with the.conventional steam systenm to provide‘fuel—oil for shielding, an overall reactof plant weight of 5/ lbs/SHP is realized, In addition, the future potential of this design concept was investipated utilizing unproven but indicated feasibie teéhnology advancements, Speéific weights on the or&er.of 54 1bg/SHP were ‘found possible in this power range; ACKNOWLEDGEMENT The sauthors wish to take this opportunity to express their appreciation to the many people throughout Osk Ridge National Laboratory who so_freely allowed us to infringe upbn their spare moments to gain the benefits of their experiences and knoyledge. The group's special thanks goes to A, P, Fraas and W. R, Chambers, our group advisors, for their guldance and for selecting the study. Our particular indebtedness to the many pecople of the ANP Project‘at ORNL is attested by our numerous. references to their work. Also we wish to acknowledge the personal aid received from specialists v of the Bethlehem Steel Shipbuilding Division, Ingalls Shipbuilding Company, Babcock and Wileox, and the Knolls Atomic Power Laboratory. -6- TABLE OF CONTENTS- 1.0 Summary, Descrifition and Conclusion 2.0 3.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 Introduction Réactor Fuel Materials Heafi Exchangers and Steam Generators Potential | Conclusions Introduction 2.1 2.2 2.3 2.h 2.5 2.6 Need for a High Performance Marine Reactor Ship Installation | Design Philosophy Reactor Comparison and Selection Advantages and Dlsadvantages of Fused Salt Reactors Additional Applications OveralltPower Plant Description 3.1 3.2 3.3 3.4 3.5 3.6 3.7 . Introduction Alternate Approsach Reactor General Shielding Weight Comparison of Nuclear and Conventional System Hazard Evaluation Page 15 15 16 18 20 20 21 22 24 2l 25 26 28 30 33 35 35 38 39 43 42 43 b5 L Page v 4,0 Fuel and Secondary Fluid W6 4,1 Fuel b6 4.1,1 Introduction hé 4,1.,2 Coupogition Wt 4.1.,3 Corrosion 51 4,1.% Physical and Thermal Properties 53 4,1.5 Nuclear Properties 54 4,1.6 Availability and Cost B 4,1.7 Fuel Addition 55 %g: 4.1.,8 Fuel Reprocessing 55 J k,2 Secondary Fluid 57 4.2.1 Introduction 57 4,2,2 Physical and Thermal Properties 58 4,2.,3 Disadvantages of Fluid 59 5.0 Material Selection 61 5.1 Structural | : 61 5.2 Moderator 65 5.3 Reflector 66 . 5.4 . Poisoned Modetator Section 66 ‘ffi 5.5 Design Properties of Materials | 67 ¢ 6.0 Reactor and Primary Heat Exchanger Design 68 6.1 Introduction 68 6.2 Reactor - Types Cofisidered and Selection 68 6.2.1 Internal Arrangement 69 6.2;2 Vessel Design 73 6.2.3 Structural Arrangement | 75 7.0 8.0 S~ 6.2.4 Fuel Pumps 6.2,5 Pressurizes and Expansion Chamber 6.3 Primary Heat Exchangers 6.3.1 Design Criferia_ 6.3.2 Basic Design 6.3.3 Parameter Study 6.3.4 Stress Considerations Steam Generating System 7.1 Introduction 7.2 Molten Salt Cycle Selection 7.3 Steam Generator " T.3.1 Types Consgsidered T.3.2 Design of Selected Steam Generator 7.4 Superheater 7.5 Auxiliary Equipment 7.6 Part Load Operation Reactor Analysis 8.1 Nuclear Configuration 8.2 Parameter Study 8.2.1 Cross Sections 8.2,2 Summary of Results 8.2.3 Control Rod Study Page 76 76 78 78 78 83 8k 88 88 88 91 91 ok 97 98 102 110 110 2113 114 116 117 o~ a3 oy v B € 9.0 10.0 8.3 Nuclear Design 8.3.1 Criticality 8.3.2 Self Shielding 8.3.3 Burnup and Fisgion Product Poisons 8.3.h Prompt Temperature Coefficients 8.3.5 ZXenon Poison 8.3.6 Delay Neutron Loss 8.3.7 Excess Reactivity Requirements 8.3.8 Control Requirements Shielding 9.1 Introduction G.2 Neutron Flux Calculation 9.3 Secondary Salt Activation 9.4 Dose Tolerance levels 9.5 General Shield Arrangemenf 9,6 Primary Shield 9.7 Secondary Shield Heat Balance and General Aspects of the Steam System 10.1 Introduction 10,2 Steam Requirements 10,3 Condensate and Exhaust Heat 10,4 Heat Addition in the Steam Generating System 10.5 Comparison of Efficiencies Page 123 124 12k 129 129 131 131 132 133 135 135 135 141 142 143 145 146 154 154 154 157 158 158 w1 0w Page 11,0 Overall Power Plant Particulars - 160 | 11.1 Introduction | 160 ¥ 11,2 General Arrangement - 160 11.3 Power Plant Control | 165 11,4 Emergency Operation 171 11,5 Maintenance 178 11,6 Removal and Disposal of Volatile Fission Products _ 180 11,7 PFuel Loading | . | : * 182 11,8 Pumps, Valves and Blenders | 184 12,0 Modified Approach 189 « 13,0 Weight Summary | 191 14,0 Future Potential | | 196 ) - Figuré No. 2-1 3-1 3-2 3~3 h-1 b2 4-3 Lok 5-1 5-3 6-1 1] LIST OF FIGURES Partial Summary of Current and Proposed Nuclear Marine Installations Flow Diagram Reactor Diagram Specific Weight Comparison in 1b/shp Phase Diagram of the Three-Component NaF—Zth-UFh System Partial Pressure of ZrF, Based on the Assumption that Only NaF and ZrF) %xist in the Vapor Phase Fused Salt Fluoride Volatility Uranium Recovery Process The System LiF-NaF-BeF, Design Curve for As-Received Inconel Tested in Fused Salt No, 30 at 1300°F| Comparison of the Stress Rupture Properties of As- Received Inconel Tested at 1300, 1500 and 1650°F in Argon and Fused Salt No. 30 Effect of Section Thickness on Creep-Rupture Properties of As-Recelved Inconel Tested in Fused Salt No, 30 at 1500°F under 3500 psi Stress Estimate of Weight Per Power Ratio vs Primary Heat Exchanger Outer Diameter for Various Tube Spacings Proposed Steam Generator Proposed Superhea ter Proposed Basic Arrangement Salt Viscosity Friction Factor Heat Transfer Correlatlon Pump Equivalent Weight Page 29 36 41 by 49 50 56 60 62 63 64 85 103 104 105 106 107 108 109 Figure No, g1 g-2 8-3 Bl 8-5 8-6 817 8-8 9-14 9-1B 9=2 9=3 10-1 11=1 112 11-3 11-6 11=7 -]2= Reacfivity vs Mass U=235 Reactivity'vs U~235 Concentration Reactivity vs U=235 Mass Radial Flux Distribution Radial Power Distribution Thermal Flux Distribution in Unit Lattice Cell Percent Reactivity loss During Lifetime Due to Burnup and Fission Product Polsons Core Reactivity vs Cofitrol Rod Position Neutron Flux Plot -~ Core té Primary Shield Lead Neutron Flux Plot in Primary Shield Tank Secondary Shield Reactor Compartment Predicted Steam Balance for Reactor Powered System at 359,000 shp | General Arrangement Simulation Flow Sheet Reactor Power and Temperature vs Time for a Ramp, Change in Power Demand Reactor Power and Temperature vs Time for a Step Change in Power Demand Reactor Power and Temperature vs Time for Step Change in Reactivity Output Steam Temperature vsg Load Reactor Power and Steam Temperature vs Time for a Linear Change in Power Demand Page 118 119 125 126 127 128 130 134 139 140 144 149 155 163 172 173 17Th 175 176 177 Figure Nog A<5,1 A=5,2 A=5,3 A=6,1 A=11,1 A"“ll 92 A=11.3 A=11,/ A=11,5 =1 3 InconelhDesign Data Inconel Degign Data Inconel Design Data Moderator Rod Tefiperature Distribution Analog Represgentation of Fuel Loop Analog Representation of Salt Side of Primary Heat Exchanger and Superheater Analog Representation of Salt Side of Steam Generator Method of Generating Power Demand Voltages Analog Repregentation of Control System Page 20k 205 206 214 2hg 250 251 I252 253 ~14e APPENDIX Pago i 5el Materials Data ¢« o ¢ o o o o o 0 2 s o o o o o o 202 6.1 Jugtification of Moderator Material . . o . « , 207 6,2 Calculations for Final Design of Primary Heat Exchanger o« o o s o o 6 06 6 06 6 o 6 0 o 0 0 o o 215 7.1 Steam Generating System o o o o o o o o o o o o 226 8,1 Three=Group Cross Sections. o o o o o o o o o o 2ko 8.2 Perturbation Technique o o o o o o o o o o o o 21 8,3 Burnup and fission Product Poisons . « o « . & 2h3 8.4 Prompt Neutron Lifetime . « o o o o o o o o o o 25 ) 11,1 Degceription of Simulation Program o o o o o o 246 . 11,2 BExpansion Chamber Heating Calculations . , . . 248 - 13,1 Breakdown of Basic Reactor Powered System Components Weigh‘ba © &6 B © & 6 © © B ° o 6 o6 o 255 =15 1,0 SUMMARY, DESCRIPTION AND CONCLUSIONS 1,1 Introduction This report covefs a study of the feésibility of & high performance marine reactor (HPMR) utilizing a circulating fuel, fused salt reactor concept., The definition of high performance as considered in this report is low specific velght in terms of total power plant weight per shaft horsepower, By significantly decreasing specific weight below that which is currently found feasible with present and proposed systems, reactor installations on a lighter class of ships is now possible. This wduld also offer potential improvements for all heavier classes, A design study was made for a reactor system of this type to power a 931 class destroyer of 3-4000 tons displacgmenta The reactor and steanm generating equipment simply replaced one of the present boiler rooms on this c¢lass ship and duplicated the steam conditions (950°F, 1200 psig) supplied to the propulsion machinery. An overall specific weighi of 59 ibs/SHP was achieved for the 35,000 SHP delivered per boiler room, This is comparable with the presently installed oil-fired system including fuel.‘ This speci= fic weight was achieved with a reactor afid steam generating equipment overdesign of approximately 30%, Indications are that if time had allowed a reiteration of the system size to the 10% overdesign factor used in most reactor systems, a specific weight reduction to at least 54 1lbs/SHP would have been achieved, These specific weights, which are approximately one half that of any planned syatem, were brought about by obtaining a small reactor package to minimize shielding and combining this with the production of high temperature steam to give godd steam plant efficiency, =] b The initial basic study incorporated a single intermediate loop utilizing gnother fused salt {also compatible with water) to transfer the heat from the fuel to the steam generator and superheater. This prevented activation of the steam and through the use of blenders the temperature of the salt entering the steam generator was feduced substantially to decrease the problem of " thermal stress, However,; this required that saconflary shielding be placed about the large volume of fhe steam generating equipment,- It was found that through the use of two intermediate loops the amount of secondary shielding could be.redueed and the overall specific weight releaaed from 58 to the 54 1bs/SHP, The comparable reduction fdr the case with 30% overdesign is from 65 to 59 1bs/SHP. Unfortunately sufficient time was not available to allow as detailed a study as that given to the single intermediate loop systen, Considerable use and reference has been made of the ANP studies and experimental work carried out at ORNL on fused salt reactors, This has allowed demofistrated components and materials to be incorporated directly into this plant, 1.2 Reagtor In order to achieve the primary overall objective of reduced specific weight it is desirable to keep the reactor size as small as possible in order fio minimize n;t only reactor weight but that of the primary and secondary éhielding as well, The compact reactor selected was cylindrical in shape with the fuel circulating up through a central critical region and then down through an ammular downcomer at its periphery.containing the primary heat exchangers (fuel to secondary fluid). The core is moderated by cylindriecsgl rods of beryllium oxide clad with Inconel that are gquisPaced throughout the core region, A nickel reflector surrounding the core plus an additional blankeflblack to =] e thermal neutrons shield the primary heat exchanger and prevents excessive activation of the secondary fluid, Because of the inherent stability that has been demonstrated with reactors of this type, poison rods are not needed for control but a single réd is placed at the core centerline to provide for reactor shutdown, mean témpérature change, and fuel burnup, | The reactor and steam generating system were designed to produce 125 MW which is a conservative overdesign of greater than 25%., This safety factor is considerably larger than felt necessary but was brought about by the necesgity of starting the reactor design before the details of the steanm _ system became available, An average core temperature of 1225°F with an 100°F difference across the core was selected as a compromise of weight and thermal efficiency against corrosion and thermal stress problems, The neutron flux gpectrum is largely intermediate giving rise to a fission distribution of 28% thermal, 63% intermediate and 9% fast, ~ The nickel reflector tends to hold up the thermal flux spectrum at the outer edge of the core and helps %o prpvide the favorable pesk to average power diétribution of approximately 1.4. The power density averaged over the core is 360 watts/cmB. | The reactor vessel itself is approximately 6,7 ft in diameter and 6,7 ft high.’ An expansion tank for the fuel is incorporated into the head design along with provisions for removing Xenon and other fission product gases., Three fuel pumps are also located in the reacfor head in a manner such that they may be replaced aboard ship. The reactor head is removable by unbolting and cutting a smell omega type seal weld, This allows replacement of the primary heat =18~ exchangers afid inspection of the core agsembly, However, it is recomménded that the reactor be removed from the ship prior to this operation in order to reduce the remote hgndling costs and problems., Also the feasibility of balancing the cost of discarding complete reactor assemblies against that of the design and operation of a remofe handling facility should be thoroughly investigated with the idea of reducing both overall costs and simplification of the basic reactor design, The primary reactor shield is made up of structural support steel along with approximately 5 inches of lead and 39 inches of water, The shield requirements are based mainly on the fission product and sodium decay gamma's and the delay and fission neutrons in the outer annular region containing the primary heat exchangers. These activities were found to be seversl order of magnitudes greaterAthan the prompt gamma and neutron radiation from the core, The secondar; shield for the basic_study enveloped both the reactor énd the steam generating equipment and incorporated a thickness ofJapproximately 4 = 6=1/2 inches of lead. This requirement is a direet function of the activation ‘of the sodium ions in the secondary fluid as it passes through the primary heat exchangers. 1.3 Fuel and Secondary Fluid In the selection of a fuel for this system, in addition to simpiy selecting a carrier for a critical amount of uranium, primary emphasis was placed on chdosing one that had been proven acceptable, This included its chem;cal stability, corrosion, nuclear, and physical properties. This selection was rather easily made since a large number of salts have been investigated =19- by ORNL and only a few found promising enough to warrant additional testing, A solution of sodium, zirconium and uranium fluorides was selected on the basis of reasonable nuclear and physical properties and because it had been used successfully in a reactor experiment (ARE)}, Also, extensive investigations have been made on its corrosion and physical properties in anticipation of its use in the Aireraft Reactor Test (ART), The vapor pressure of this salt is typically very low so that at operating temperatures the reactor vessel has to be pressurized only slightly to prevent pump cavitation, Thé actual composition of the fuel selected; closely approximating that of the ART except for exact uranium concentration, is 9% NaF, 453 23 and 6% UFA (mole percent), Uranium will be added to the system in the form of (NaF), UF,, Pellots 4° or dissolved golution of this salt would be added during operation of the reactor to compensate for uranium burn-up and to override fission product and corrosion poisons. It is anticipated that sufficient addition of fuel may be made throughout the life of the reactor to eliminate the necessity of rgplacing the original salfi loading. A basic ground rule requiring chemical aompatabiiity of the fuel, sécondary fluid, water and sea water was established. In view of this coupled with corrosion, heat transfer, radiation and chemical stability requirements, the selection of possible choices was narrowed down to a fused salt, Because of the difficulties involved in preventing this salt from freezing in the steam gonerator a low melting point was also a requirement, On this besis a solution of sodium, lithium, beryllium fluorides (mole percentages of 30, 20 and 50% respectively) with a melting point of 527°F was selected., 20~ 1.4 Materials .Mbst fuséd salts are quite corrosive to the standard structural materials, However, it has been found that alloys containing large percentages of nickel offer the gdod corrosion resistance to the fused fluoride salts, BExtensive testing at ORNL under the ANP Project has shown that Inconel and the nickel- molybdenum alloys present the best combination of strength and corrosion resistance, DBecause the procurement and fabricability of Inconel are better defined at pregent it was seléeted for the basie désign although the corrosion resistance of the nickel-molybdenum alloys is much superior, Inconel was salso selectéd as the structural material for components in the steam system within the secondary shisld because of its superior resistance 0 chloride gtress eorrfisiono The complexity of mechanical design problems involved in a separate moderstor cooling system fiade it undesirable and must be weighed against the high temperature difficulties encountered with fuel cooling. A ceramic moderator appeared to offer a reasonable compromise from the temperature standpoint although most did not have adequate nuclear and/br physical pro- perties to bé aecepté‘ble° Beryllium oxide has the best overall characteristics at present as its fabrication and physical properties are reasonably well known and its satisfactory behavior under nuclear radiation had been demonstrated experimentally, 1,5 Heat gerg and Steam The primary heat exchanger is a once-through counter-flow type with the secondary salt on the tube side, There are 12 heat exchanger tube . bundles with each tube bundle made up of 6 subassemblies for ease of fabrication =21~ and inspection, The steanm genérator.and superheater are of conventional design fitilizing U-tubes to reduce the thermal stress problem. The high pressure water and steam are located on the tube side to minimize the component weight., Several other designs that offered potential weight decreages were considered but were not incorporated.because the design was not as well proven, A blender was placed in the secondary fluid upstream of the steam generator, This provides a means of maintaining the salt in the boiler at a lower temperature than that in the superheater by mixing a relatively low temperature salt for the exit of the steam generator with the high temperature sfiperheater salt, This cofisiderably reduced the thermal stress at this point and offered a weight saving over the use of a salt-to-salt regenerator, 1,6 Potential . The major objective of the atudy covered by this report was to design a power plant incorporating ideas and éomponents that could be substantiated by referencé to a reasonable amount of experimental development work and test programs. However, there exist many new facets of fused salt technology that appear to offer large potential but at presént are little more than qualified opinions plus a small amount of experimental verification., Because it was felt.that this potential was significantly greater than that existing with other type.of reactor systems, the study was extendsd to incorporate the most promising of these developments., Through the use of a new fuel that offers more self moderation, moderator materials that allow the core to operate at a higher power densities, and structural materials that offer improved corrosion resistance, the basic size =22 of the reactor itself decreased from 6.7 ft diameter by 6,7 £t high to approximafiely 5 £t diameter by 5.4 ft. high, To further decresse the size of the heat exchangers and steam generating equipment, sodium was used in the intermediate loops. This study indicated that it would be reasonable to expect that a specific welght reduction on the order of at least & lbs/@HP could be achieved in the future with fused salt reactor systems, 1.7 Gonclusions This design study of a circulating fuel, fused salt reéctor for é marine power plant has shown that such a.system is technically feasible at present, Reactor systems of this type not only allow overall perforggnce improvements over current systems, but allow reactor installations to”be considered for a lighter class ship, In addition, with the developmental and experimental work accomplished in thig'field at ORNL; the construction of this plant could proceed with a minimum of additional development work, Also considering the potential of this'systefi with developments that are novw in sight, it appears that considerable performance and weight improvements could be expected, The difficulties involved in handling the fluoride fuel and maintaining it above its melting point have been satisfactorily overcome and proven out in test loops and a reactor experiment, Algo, materials that will give adequate resistance to thé’high temperature corrosiveness of the salts have been found, although increased corrosion resistance would be desirable, Although the fuel inventory required is considerably higher than for other systems, this is partially offsat.by the elimination of the need for replacement ~23- cores and holdup for chemical reprocessing, When the many important advantages of this type of system are considered, they appear to more than offsget the above, These include: higher temperature gnd overall thermal efficiency, low weight and volume reéuirements, low pressure system, proven stability allowing the elimination of numerous integral cbntrol rods, continuous poison gas removal, fuel makeup as needed, etc, In the judgement of the authors the ciroulating fuel-fused salt reactor not only shows considerable performance potential over present and proposed marine installations but it offers the most promising system applicable %o a small ship installation, =2/~ 2,0 INTRODUCTION 2.1 Use for HPMR Atomic weapons were not only the forefather of atomic power reactors, but also the forerunner of a completely new concept of naval warfare, A small ship utilizing missiles with atomic warheads could have the destructive effectiveness of the largest warship of the preatomic era. If one could take such a small ship and build it in large numbers, give it a high speed along with an effectively infinite range, it clearly would present quite a formidable weapon. A small ship with no refueling problems would also have many other potential uses such as convoy and patrol duties in isolated areas, The purpose of this report is to determine the feasibility of a reactor systenm capable of being instslled aboard # small ship to give it the éffective infinite range mentioned above. Also once the feasibility of an improved high performance (lightweight) marine reactor is established for a small ship, it likewise holds promise for considerable weight savings on larger vessels and volume savings on submarines, | For the purpose of this report a 931 class destroyer of 3500-4000 tons displacement was selected for investigation, This ship is roughly half the displacement of the smallest current proposed reaefior installation (Sec, 2.4 and Ref, 8), but considerably over the minimum size felt necegsary to contain a crew for long durations. This ship contains two separate boiler and machinery rooms utilizing steam at 950°F and 1200 psig to produce a total of 70,000 SHP, These steam conditions fortunately fell into the rangé'considéred desirable for reactor installations of this type. With the machinery room fixed, the boilers could be simply replaced by a nuclear system without com- =25 promising the basic reactor design. This would considerably ease the redegign of a conventional 931 class destroyer to nmuclear power as well as offer the shipboard advantage of the crew being thoroughly familiar with the steam plant, In addition, the logistic and shore maintenance problems would be reduced because of the number of identical steam plants in service, 2,2 Ship Ingtgllation For the purpose of thig study it was felt most feasible to replace only one of the boiler rooms with a reactor system, This not only gives the advantage of having the proven reliability of a completely conventional system aboard ship, but would considerably reduce the total cost of the complete installation, 4lso the performance penalty paid for utilizing both the reactor and boiler systems would be very small if not negligible, The difference in speed of this class ship between operating on the reactor system along (35,000 SHP) and maximum power (70,000 SHP) ig approximately 4 knots, Obviously, this inefficient utilization of power is not warranted except under emergency conditions. In addition, structural design problems associated with vibration and noise along with their relsted detrimental effects on submarine and aircraft detection equipment does not make extended maximum speed operation appear feasible, As an additional point it should be noted that if an average fleet apeed of as high as 20 knots is assumed, this ship would be developing less than 1/3 of its potential reactor power and zero conventional, Therefore the conventional steam boiler system can be used to augmeht the reactor when emergency conditions exist apparently without fienalizing the overall ship operation and offer large savings from both the coat and reliability standpoint, s The total amount of fuel oil carried aboard is approximately 54% of its original value, Since this is to be used only under special conditions and not for crusing it is considered adequate. A typical combat problem was not available for analysis, but it is felt that the endurance of the nuclear- 0il fired ship combinatién at maximum power would be substantially increased over the conventional ghip, 20,3 Design Philosophy - Because of both the relatively short time available for this study and the limited experience in certain aspects of the field it should be realized that a thorough investigation of a1l phases was not possible. In instances where there appeared several feasible approaches, but with éach requiring a considerable design effort o evéluateg a somewhat arbitrary choice utilizing engineering judgement had to be made, These selections and the alternate possible choices are discussed throughout the report, The primary objective was to establish design feasibility for the small ship application, Accomplishment of this with the selected design, indicates that with additional study the possible alternates herein bypassed could either be incorporated with a subsequent design improvement or simply rejected, Many detailed problems concerning the steam system were not thoroughly investigated as it was félt,that solutions to these wers well established, .Major emphasis was placed on the really unusual problems concerning the reactor and intermediated salt systems to obtain plausible solubions, The basic design philosophy was to use materials, designs, and techniques that have been established as feaéible and backed up és much as pogsible by experimental verification, In cases where the restriction to present day w2 technology eliminated alternate approaches, they were briefly mentioned for possible future consideration., Attempts were made to fully utilize the experience, knowledge, and background of the personnel at the Osk Ridge National Laboratory and other industries, Several basic ground rules were established early in the design study, The first was that within the limits of the previous paragraph, the design optimization would be on the basis of obtaining the lightest weight on g 1b per shaft horsepower bésis. Another was to prohibit the use of a fuel or intermediaste fluids that were.not compatible with each other as well as steam plant and sea wafier6 This ground rule was believed to be basically desirable from the battle damage standpoint because of the severe punishment that ships of this type can receive and still be operable, Also this com- patibility offers obvious safety advantages in the steam generator design, Advéntage was taken of the conventionsl plant fuel oil left on board by using it for shiel@ing purposes. If a completely nuclear destroyer design | is required; it appears that the weight advantage may not be as acceptable as for the combined conventional and nuclear powered ship. However, because of the narrow beam of this class ship, the reactor compartment can be rearranged to utilize the salt and sea water at the sides to reduce the shielding requirements, Because of the decreasea volume and especially the height of the fused salt system it is possible to install the top shield deck of the . reactor compartment at the water line, Advantage could also be taken of putting the reactor compartments back-to-back to reduce the required shielding, Unfortunately limited time prohibited detailed consideration of these arrange- ments from being made for a completely.nuclear ship 'although an estimate was made on the added shielding weight required for the proposed installation. ~28-. 2.4 Reactor Comparisons and Selection In the process of 1nvestigating potential small ship applications and selscting g 931 olass destroyer for thie study, it became obvious that g nuclear power plant specific weight on the order of 60 1bs/SHP had to be realized, A brief review of current and proposed nuclear ship installations vas made and these all fell considerably short of meeting'this requirement, These values, summarized on Figure 2-1, are to'be_considered only approximate and neither the latest or the bost values, The lightest values found were 105 for the FIW and 90 lbg SHP for D1G, The FIW is a joint WAPDmBethlehem Steel effort in which g 1arge portion of the detail design has been firmed up. This design preduces approximately 80,000 SHP and is installed on g 14,000 ton ship which would normally be considered in the light to medium crusier class, The D1G program is a KAPLmBethlehem Steel venture that ig 81111 in the early preliminary design stage with the specific.weight given being only a design objective and not a design-substantiated valueo The design power is to be 60 000 SHP with a total ship weight of roughly 7500 tons. This size is in between what had been considered the destroyer class (2,5 = 4000 tons) and a eruigser class (12 - 18,000 tons) It is apparent that thege ”1nstallatzons do not offer much promise of g welght reduction to 60 1bs/SHP for a 931 class installation, In investigating the field in general for g lightweight reactor systemg the Aircraft Nuclear Propulsion (ANP) Projects appear to hold similar require- ments for low propulsion system specific weight, In addition, it seems substantially improved at a small enough increase in overall weight to make g ship application most feasibleo -29.. FIGURE- 2— | SECRET PARTIAL SUMMARY OF CURRENT AND PROPOSED NUCLEAR MARINE INSTALLATIONS SHIPS SHAFT HP PROPULSION SPECIFIC SYSTEM WGT* WGT™ (LONG TONS) (LBS / SH) SUBMARINE S NAUTILUS (S2W) 15,000 1100 160 S4wW : 6,600 690 230 SEA WOLF 15,000 1200 180 (SIR-52G) - TRITON (SAR- S4G) 34;000 1700 110 (2 REACTORS) SURFACE SHIPS FIW 80,000 3700 105 DIG 60,000 2900 - 90 93! cLAss DESTROYER 70, 000 1800 58 { NON NUCLEAR) {INC. FUEL) * NOTE: THESE VALUES ARE ONLY REPRESENTATIVE AND NOT THE LATEST OR BEST VALUES. ' T — i - =30- Two types of reactor systems were considered (1) the heterogeneous gas cycle using high temperature ceramic.fuel elements under development by General Electric at Bvendale and (2) the circulating fuel, fused salt L system being developed at ORNL, A pas cycle did not appear to be readily applicable at present for a ship installation because gas turbines of the size réquired”had not been developed, Alsgo, éven though bigh gas temperatures and correspondingly high turbine efficiencies could be achieved, material limitations could prohibitly limit the extended life required for a ship applicationo The fused salt reactor concept appeared to readily adapt itself to a steam generation application, The nominal reactor temperature could be decreased several hundred degrees (OF) from the ANP design values for an improvement of the corrosion pfoblemo This still wouid résult in an asmple temperature margin to provide steamlwith BmAOOOF of superheat at desired pressures, These factors coupled with.the "at. hand" availability of fused salt technology made this type of reactor appear to be most feasible at present, 2.5 Advantages and Disadvantages of Fused Salt Reactorg Like any complex system, a fused salt reactor installation exhibits | ‘ both strong and weak points, In any reactor comparison, a relative weighing of these must be made along with the determination that flo unsolvable weak points exist., However; in considering a 931 class ship installation such a comparison cannot easily be made because no othdr reactor configurations exist that can satisfy the strict weight requirements. Therefore if a need for a nuclear ship of this si axists, the fact that no unsolfiable problems apparently =31 exists in above sufficient reason to proceed with a detailed study., Faftunatelys & fused salt system offers many advantages over conventional reactor systems that could make it highly competitive even for large ship applications, There- fore the advantages as well as the problems of fused salt reactor systéfis are discussed to establish its potential over other reactor types for future comparisons, R.5.1 Advantages 1, High temperatures are obtainable which give rise to high overall plant thermal efficiencies, 2. Low pressure reactor system, Pressure required (< 100 psi) oniy to provide pressure differential for fluid flow and to prevent pump cavitation, 3. Inherent stability of this class reactors has been demonstrated, Multiple control rods and control drive mechanisms are not required at a conml siderable saving to cost, reliability and maintenance problems, A gingle control rod which may be required to compensate for fuel burn-up or to prdvide for désired temperature changés may be located outside of the reactor wvessel and subject to relatively easy maintenance, 4o With this type of reactor design it would be possible to overtemperature the reactor to obitain large increased in power output for emergency conditions, Undoubtedly this would be at a sacrifice in overall 1ife of the system but extreme conditions could warrant this use, 5. The fission product gases may be continuously removeds; thereby eliminating the Xenon override problem.and the excesgs reactivity requirements., 6. The basic reactor is generally much more symmetrical and smaller than other systems, thereby reducing the shiel&ifig problems as well as =32 the overall size and weight, 7. Inexpensive fuel preparation. The reactor core is of simple design and therelis nb fuel element fabrication and burnable poison costs, The handling of U-235 is simplified as "apiking" of the salt fuel carrier is required only aféer the readtor hag been filled, 8. Reloading o; refueling would generally not be required during the life of the reactor., Tuel additions may be made during reactor operation to compensate for fuel burn-ub and to override soluble fission product buildup, Because of thig and'(5) the excess reactivity requirements are considerably reduced and can lead to & reactor thét is inherently safe | from power excursions, . | | 9. Chemical stability - No radiation‘damaée or fuel decomposition problems. Explosive radiolytic gases are not formed, thereby eliminating problems such as the recombination of H2 and 02 in water reactor systems, 10, The combination of (7) and (9) make it possible to utilize high core power densities with the subsequent reduction in reacfior size, 1l, Chemical reprocessing is greatly simplified with & homogeneous system giving a corresponding cost decrease, 12, Although not diredtly applidablé to a marine installation, it should be noted that for breeding'purposes both thorium and uranium are soluble over a large range of concentrations, This is not true for either an aqueous homogeneous or a liqfiid nmetal éystem. Re542 Disgdvantgges 1, Gorrosion problems are more difficfilfilthan for an aqueous or sodium system but probably better than a homogeneous liquid metal reactor. <+ High melting point requires that careful attention be paid =33 to loading and Operational techni‘queso A molten state is required at all times; however the successful operation of a reactor experimant (Ref, 6) indicates that these problems may be SOlvedo 3, A high degree of leak tightness and reliability 1is required for the core vessel and primary heat exchangerso Careful and tight quality control and material inspection is requirefio 4; A high fuel investment in the reactor is required, Howéver, considering the elimination of replacement cores and the cooling period before chemical reprocessing the total investment may be comparable to heterogeneous, solid fueled systems, 5. Poorer neutron economy is obtained than with aqueéus homogenedus systems although newer types of fused salts offer improvements, \tions 2.6 A@ditignalhfizilig | Once the design feasibility of a fused salt reactof system is proven 1t offers considerable potential for both larger and smaller vesgels, If- specific weights on the order of 60 lbs/SHP ean be maintained for smalier power sizes many new opportunities are available for an even smaller ghip application. In going to a larger size ship an overall specific weight of 60 should be more easily attained, This would offer either a weight reduction or an increase in storage capacity of approximately 1600 tons for a ship the size of FIW or 1000 tons for D1G, A fused salt reactor also offers a considerable reduction in the size of an installation, While important for any ship, this is even more important for a submarine application, A preliminary comparison madé by KAPL in 1953 (Ref, 7) indicated the degign advantage of this type of system for a submarine _ym application, 4 current design would tend to give an even bigger advantags. Incidentally, a potential non-marine application not pertinent to this gtudy but of general-interest invélves_thé use of stationary fused salt reactor plants for treeding and electric power production (Ref, 5). _35 - 3.0 OVERALL POWER PLANT DESCRIPTION 3.1 Introduction The utilization of only a single reactor system aboard a 931 class destroyer offered gsome desirable flexibility as to the overall ship arrangg%- ment; However, the numerous basic considerations required to establish | 4 the dptimum shipboard installations were somewhat beyond the gcope of this report, With a cursory investigations, it at first seemed to be most -advantagecus to replace the fbreward boiler room with the reactor and steam generator equipmento This had the sizeable advantage of not requiring any layout considerations or secondary shield penetrations for passage of the propeller shafts from the other engine room, However, under detail design, the size of the reactor compartment was reduced below that originally contemplated and means of circumventing this problem became spparent, The af't compartfiént location also offered the advantages of easier accessibility and a better location of the fuel o1l tanks to maintain ship trim, The basic system upon whi&h the major design effort was placed consigted of the eirculating fuel, fused fluoride salt reactor incorporsting an integral heat exchanger unit fo remove heat from the fuel, A secondary fluid, another fluoride salt, is used to transfer the heat from the primary heat exchangers within the reactor vessel to the steanm generating equipment, The steam is then supplied to the conventional 931 class destroyer machinery room at a temperature of 950°F and a pressure of 1200 psig, A simplified schematic of this system is shown on Pigure 3-1, The detailed heat balance is discussed later in Section 10 and presented in Figure 10-1, IAIVA WINHL — @ ¥43ON38 — V - - }-¢ 3¥Nn9i4 dwnd —(d) 4.89% , . 7 o 1u/dl 01929 ¥3L1V3IH \\_ de204 H3ILVYMQ33d . | MOLVYINIO WVY3LS mumzuozoo\x ¢ + 4 1sdgez , & 4,226 . MWK 6°C6 HOLOV3Y ——h, , _ 431V 3H . | 4809 HS 000°€e § 1 3INISHUNL ¥43dns * /q) . 01X 612 /g1 Ol x 612 o 1 i} L\%/’. 1Isd 0021 —4,066 4, 8€11 I "W3ILSAS NOISTNdOHd — ¥01Ov3H 40 OJILVWIHOS d3I4I1dWIS .‘..37.. The mean reactor temperature is 1225°F with a power output of 95,9 MW required to supply 35,000 shaft horsepower to the ship's propellers. The temperatures across the various equipment as well as the steam and salt fldg- rates are given on Figure 3=1, The secondary salt syétemg which carries tfi;_ heat from the reactor to the steam generator; is broken up into two inter— ?} ‘connected ioopso The top loop supplies the superheater with a relatively | flot salt, This is required %o reduce the su?erheater size becausé of the low heat transfer coefficients characteristic of the steam side. The bottom loop maintaing the éalt at a lower temperature to reduce the thermal stress problems in the steam generator, This is desired here because the relatively high heat transfer coeffioiénts on both the water and salt side would give a large temperature drop across both the tube and header walls and hence a high thermal stress, Blenders are used to interconnect the two loops as indicated thereby allowing cold salt from the exit of the sfieam generator to be réciraulated to reduce the resactor inlet temperature to the desired value, -.It should be noted that ihe reactor and steam generators were basically designed to produce 125 MW, This overudesign.of approximately 30% was brought about by the time lag involved between when the basic reactor had to be selected and when the detailed steam conditions for the desired size ship could be obtained, A 10% over-design safety factor (used for other marine reactor applications) would be desirable but time did not permit a reiteration of the work to thié size, An approximation was made to allow for this over-design (Sec, 12) to .indicate the overly large weight penalty incurred. In replacing the boiler equipment with this reactor system, it appears (Section 11,2) that a substantial volume saving is also realized, While important for any small ship, it should further emphasize that in this application the volume is saved over the height of the boiler room (approximately 30 ft) making it also available for missile storage, 3.2 Alternate Approach The sodium component in the secondary fluid becomes highly activated as 1t passes through the primary heat exchanger due to both delayed neutrons from fuel in the exchanger and fést neutrons from the core, Because of the large amount of this secondary salt outside of the primary shield, it is necessary to incorporate a relatively thick secondary shield aboubt all of the steam generatlng equipment,, A more detailed design should consider the possibility of reducing this activation somewhat through the use of poison bearing materialsg:ioeog boron, in the_héat exchanger region, However, thig does not appear to offer a large reduction in tfie secondary salt activation because only approximately 10 = 15% of the activation results from thermal neutrons, the remainder occurring because of the high intermediafie energy flux and sodium resonance fieakso | An slternate approéch that was briefly gstudied used two intermediate fluids (both salt) in order to prdvide & non-redioactive salt in the steam equipment. Although a penalty was paid due to increased pumping power require- ments and superheater size, this was fiore than compensated for by a shield welght decrease due to a smaller enclosed volume, Direct access was also given to the steam generators and superheaters for maintenance° In addition, 1t now appeared feasible to keep_the secondary shield small enough to allow an aft boiler room installation, if desired, without the complication previously mentioned, A 2-¢ 3J4dnoid MW G2l HOLOVIAY 3INIMVIN 3ONVNHO443d HOIH ll8 -Ig > NOILO3IS 40d HO1LVH3IAOW g3aNOSiOd Saoy HOLVHIAOW— H0L03N434 TAMOIN — H3IONVYHOX 3 AUVWIEC g13iHS H38WVHO NOISNVdX 3 13Nnd a0y 04Y1INOD S3INVA 43SN43ia S \ 801.2 :’?:3 -\\\\‘ N Saoy QT3IHS \\ NOI93Y ¥IONVHOX3 LV3H \,g < SL3ININ0 HIONVHOX3F LV3H AdHVWIYd 13LNC dAnd QAIAISSVIONDN gyLee "OMO-U1-"INYO nNne 3,3 Reactor The overall éhape'of the baslie reactor 1s a cylinder approximately 80 inches high by 80 inches in diameter. Fuel is circulated up through a central critical region equivalent to a cylinder 75 cm éiameter by 80 cm high and then down through an annular downcomer around the periphery. (See Figure 3-2) This outer region contalns the primary heat exchangers for transferring the heat from the fuel to the secondary salt, Gylindrical rods of beryllium oxide suitably clad with Inconel are equispaced throughout the core 1o provide for moderation, The ends of these rods are loaded with a poisoned material to reduce.end leakage and fissioning ~in the entrance and exit plena. A single control fod éhannel, approximately . inches in diameter, extends through the center of the core region, The sides of the core are enclosed by a nickel reflector blanket 6 inches thick, This inelagtically scatlers some high energy leakage neutrons back into the core to improve the criticality as well as to offer both compact neutron and gamma shielding, To further reduce the neutron flux in the heat exchanger region the reflector is in turn surrounded by a 5-1/2 inch thick region of cylindrical rods containing a mixture of beryllium oxide plus boron-l0, Boron bearing Inconel rods are placed in the interstices of these cylinders for shielding purposes and to reduce the fuel located in this region, A thin slab of material essentially black to thermal neutrons, boron carbide in a copper matrix, then surrounds this region to completely absorb any neutrons that are thermalized in its outer periphery, Small passages are provided through the nickel reflector and the cylindrical BeO-BlO'rods to eirculate fuel for cooling purposes., An 1/8 inch annular gap is also located between the thermal shield and the reactor pressure e vegsel wall for cooling purposes. The lower temperature fuel from the heat exchanger axit circulates through here and then to the expansipn chamber in the reactor head., This minimizes both the head temperafiures brought about by decay heating and "snow" formation from the fuel (Ref. Sec. 4.1.2). Three removable centrifugal pumps which are located in the reactor head profiide for fuel flow and pressurization of the system. These pumps are also designed to facilitafie'removal of the gaseous fission pfoducts. Additional details of the reactor design are incorporated into Sections 6, 8, and 11, 3.4 General A major disadvantage of a reactor of this type is that provisions must be made tO'enSufe that temperatures are maintained above the fuel melting point at all times, Althdugh accomplishment of this has been proven feagible by both many loop tests and a reactor experiment; (Ref. 6), careful attention to operational procedures are required, Although a complete freeze up is not catastrophic from s nuclear sense, experience has shown that severe pro- blems exist from a stress standpoint upon remelting, Becauss of this, dump tanks are included between the double hull under the secondary shield com- partment for emergency use. The reactor and sgecondary salt pumps may all bé replaced through the secondary shield, Sufficlent room also exists above the secondary shield so that the pump drive and control drive mechanism motors may be accessibly located, The lightest and most flexible system for pump drives appeared to be the steam turbine, This offered the advantage of having variable speed characteristics and also did not require the addition of generator sets to the system, It o= was planned to back up each of these drives with a small AC, motor, Those would provide suitable'ciréulatibn under zero power standby conditions as well as offering the safety advantage of having two irdeperdent systems under emergency conditions. | Control of the reactor systefi could be accomplished by varying the sécondary salt fiow rate as demanded by the steam plant, This, as well as an alternate approach of by-passing sécondary salt around the reactor, is discussed in Section 11,3, The reactor as illustrated on Figure 3m2 indicates a possible method of unclamping the head to allow replaceménfi of heat exchangers and othér internal components. Two different methods of connecting the heat exchangers to the reactor vessel for disassembly purposes aré shown, Although shown to be feasible on this drawing,_it 1s very questionable as to whether or not the cost for this eage of disassembly 1s warranted from the overall maintenance standpoint, Additional discussion on two different concepts of maintenance is presented in Section 11.5, 3 . 5 Shieldin The basic shield is designed to limit the maximum allowable dose to 15 mr/hr on the outside of the secondary shield, This would allow access to the auxiliary engine room for 20 hours per week for maintenance on pump drives, deaerators, feed and boiler recirculating pumps, etc., At an average distance of 10 feet from the secondary shield, the limited access would be increaged to approximately 30 hours per week, Unlimited access would be allowed in the main engine room, The primary shield is of laminated structure containing the equivalent of 5 inches lead, 39 inches of water, and 1-1/2 inches of gstructural steel. The secondary shield is designed to attemuate the decay gammas from the activated sodium component of the secondary salt and gamma leskage from the primary shield tank, This shield makes use of the fuel oil fequired on board for the conventional system for shielding the forward, pért and starboard éidea and approximately 4 - 6=1/2 inches of lead for the top and aft seetiohs° An additional 1-1/2 inches of lead is located over the reactor fuel pumps to eliminate streaming through the crevices required to drive and replace these pumps, 3.6 Weight Comparison of Nuclear and Conventional System A weight Breakdown for a 931 clags destroyer with a conventional and a nuclear installation is given in Figure 3,3, To simplify the comparison, specific weight rather than the actual weight of the cqmponefits is presented. The actual shipboard weight (lbs) for the conventional system may be obtained by multiplying through by 70,000 SHP for the total weight or 35,000 SHP for the welght per engine room, This will also hold true for the fuel-oil weights listed. Using this information, it can be calculated that the conventional total ship power plant weight is 1123 tons (long) wet, with 728.5 tons of fuel oil, | System No, 1 is considered to be the basic design upon which most of the design effort was spent, It contains information a reactor and steam generating system over-designs of epproximately 10% and 30%., System No, 2 used the alternate épproach congisting of two intermediate fluids to allow placement of the steam generators,; etec, outside of the secondary shield, and a similar v vS €S 8¢ | 0'6S (%01) 909 269 (%0€) - 1viol = N9IS3AYIA0 mm L L v'g2 110 13nd 2 : - 02 02 032 S, S3IYOLS B QVO1 " 2’ 2¢ 2% 2l SWILSAS LN3AN3IJ3ANI r8H . 09 09 09 o' LNVId 01819313 98 4 92l S'v¥l 802 00 ONIGT3IHS NOILVIOVY 3 bl ovl €32l 08 AYINIHOVW LNVId ¥O1OV3¥ 0% 9 o 2Ll 2Ll 2Ll 9’61 AYINIHOVW NOISTNdONd WV3LS 878 V IVILNILOd QIIIGOW 0IS¥E TWNOILNIANOD | | AHO93LVO" TERENTY | H __ . ) dHS/gg7 NI SNOSIYVAWOO LH913M 0I1d4103dS = n (e 5 ¢'¢ 3YN9ld 1 3 45 approximation of both a 10% and a 30% overdesign safety factor, These designs utilized a portion of the fuel oil required for the conventionsl steam system, The 1,7 lbs/SHP for fuel oil as listed, is the fuel oil above 50% of the compietely non-miclear destroyer capacity that is required to maintain ship balance, A third design utilizing advanced material and reactor technology and eliminating the ground rule of required fluid compatability with water, achieved a further reduction in specific wéight to 56,7 1bs/SHP without any fuel?oilzrequired for shielding., A comparable value with the above utilizing fuel-oil would be 46,7 1lbs/SHP, It is Tealized that in many cases the welght of a reactor system goes up in proportion to the amount of design detail accomplished, However, this general tendency would be reduced in thig study because the entire steam and electric piant9 which accounts for approximately 1/3 of the total welght, has been actfially detailed and constructed, In addition, an attempt was made to apply conservative estimates to the various components to account for unknown growth factors, A detailed weight breakdown, including the estimates made, is presented in Section 12, 3.7 Hazard Fvaluation A hazard study for marine application of thig type of reactor was carried out by a pair of ORSORT students (Ref, 64). This evaluation indicated that basing the major destruction of both the ship and reactor vessel, this system was inherently as safe as any nuclear system, With a major catastrophe, however, a more widespread release of fission products would result, 46 4,0 FUEL AND SECONDARY FLUID L,1 Fuel b,1.1 Introdufition The chief'advantage of using s fused salt fuel is that high temperatures may be obiained at low pressures. BSuch a system ig also capable of high power density with accompanying small reactor size, and low shield weight. Also, gaseous fission products way be rémoved. No fuel element fabrication results in 1ohg life for core, and high fuel burnup. Fuel may be continuously or periodically added as it 1s burned. In addition, and by no means of least importance, fused salts do not react violently with water. | | For such a system, the fused salt fuel afid diluent must have g reasonably low welting point, low neutron capture cross section, stability at high temperaturés and in extended high fieutrdn, beta and gamma fluxes, In addition, it is essential that the fuel.system be sufficiently non- corrogsive to the confainer material that an acceptably long life and freedom from maintenance may be realized. The fused salt may or may not function as a moderator. In the reactor herein described, moderation of fast neutrons is accomplished largely by means of moderator rods dispersed throughoutmthe core, The design chosen and fuel selected resulis in an epithermal'br intermediate reactor, rather than a thermal reactor. A large amount of fundamental as well as engineering reéearch has been performed at ORNL toward development of fuels, and the selection of the fuel known hereinafter as Fuel 30 was based on the results of several years of 47~ phase dlagram reséarch, dynamic and static corrosion testing, and in-pile 160p tests, The choicé of this fuel permits the fise of technology already at hand, and does not require additional extensive fundamental research, In addition, critical experiment data and actual reactor operational data are availafile, where similar fuels were used or simulated, While it appears desirable for moderating efficiency that a.fuel be used which contains LiF and B§F2, the present technology of contaifiing such fuels is nqt considered adequate, However, it is expected fihatgfuture designs for_fused salt reactors will be possible as soon as research currently in progress has been completed. Such research is now leading toward development of very corrosion resistant nickel molybdenum alloys, which show.extremely good prospects for future use in fused salt reactors. k1.2 Cdmfiosition The approximate composition of Fuel 30, as modified by the criticality requirements of the particular configufation of the reactor, isl(expressed in mol percent) 49% NaF, U5% ZrF), 6% UF). Zirconium fluoride is made from hafnium free zirconium, Additional composition data are: Comgosition Mol % = Wt % NaF | 48,7 17.9 ZrF), | 45,2 65.7 UFy, 6.1 16.4 48 1200°F - 12000F Mol % Gus/Cm3 ) Atoms/Cm3 Atoms /Cm Sodium 15,49 335 . 8.76.x 10°% 2,57 x 1021 - Zirconium 14.2& 1,221 8.30 xv1021 | 2,43 x 102t Uranium 1,90 420 1.08 x 1022 .317 x 1021 Fluorine 68.36 1.43h _" 45,43 x 102t 13.32 x 10=1 Fiéure -1 is phase diagramlof the 3-component system, NaF-ZrF)-UF). Tt will be seen from Figure -1 that the composition selected 1s in the vicinity of fihe triple eutectic low melting composition, Also, if solid fuel concentrate is added in the form of'NaEUF6, only compositions having lower wmelting points than théfconcentrafe are formed as dissolution progresses, Figure L4-2 shows Zth.vapor pressure for various mol percentages of ZrF), as a function of temperature. It is apparent that this vapor pressure is dependent on both ZrF) concentration and on temperature. The formation of ZrF) acicular crystals ("snow")'has resulted from h%gh temperature treat- ments of ZrF) -bearing salt mixes; This segregati0n can become a problem if conditions are favorable for sfiow formation, According to our best information (Ref, L49) snow formation should not Qe 2 problem if the waximum fuel tem- perature is kept below 1350°F in the expansion chamber. The accumulation of snow-like ZrF) crystals is most undesirable and may lead to the plugging of passages or fouling of the expansion chamber, To further avoid this cold surfaces in tfie expansion chamber should be eliminated, It is clear thatthe use of a fuel devoid of ZrF), is desirable, Hofiever, corrosion considerations dictate the selection of Fuel 30 at the present time. -49- UNCLASSIFIED ORNL-LR-DWG 33984 NaF COMP. NO. 34 43 mol 9, ZrF4 ZrF4 Fig. 4-1 --Phase Diagram of the Three-Component NaF~ZrF4-UFrp System. GQ dduaasyax 8sg ‘aseyd .H_omm> 92Ul Ul 1sIXy VAIZ pue JeN LA[uoc ¥ey} wonydwnssy sy uo pesedq ¥ 417 Jo seanssaayg Ietlred -- g-3% ‘314 g 0T 0 §9¢ ORNL-LR=Dwg, =18170 UNCLASSIFIED -50 - Q09 T 2 ‘mmjpasdwaey Q0L 008 006 o001 QS0 i i [ [ wm_auw ‘0INI0ICWR L (09040120 109 AmdBw ww ‘einsseyy .,5]_- 4.1.3 .Corrosion 4.1.3.1 Introduction As;a design criterion, it was hypothesized that all design work should be predicated on the basis that the core vessel and all other parts of the systew, which were subject to activation or to radiocactive _ contamination, would be apeoified of such materials and thicknesses as to be able to withstand full power operation (125 MW), for a period of at least 10,000 hours without failure from corrosion by the fuel selected. Insofar as 1t is posgible to predict, from dynamic and static corrosion research at ORNL, this standard has been adhered to for the Fuel 30-Inconel-Secondary Fused Salt System described. Final metal thicknesses were selected on the basis of experimental results and personal experience {Refs. 39, h2,_hh, 45). Fuel 30 and the seoondary‘NaF-LiFnBeFé fused salt mix wefe selected because researohuand informed opinion showed that Inconel is a satisfactory container for them at tha temperatures of operation antioipatad. h.lf3.2 Corrosion Mechanism The most cfitical location, as far as corrosion is con- cerned, in this reactor is estimated to be the moderator cladding. The type of corrosion to be expected is chromium depletiop, by diffusion and dissolution, with hot leg-cold leg cycle accelerating mass transfer by solubility gradient. The'chemioal reaction is UF) + Cro = CrfFp + 2 UF3. Another possible source of trouble due to corrosion in fused salt Inconel systems i1s a mass transfer buildup, or depositionzo% chromium in the cold leg at a greater rate than inward diffusion can dispose of it. If such daposition were localized, clogging of small passages might result. This type of buildup was predicted for nearly all fuels tested., However, Fuel 30 was free from such -52- buildup after 1500 hours at 1500°F hot leg fiemperature in a thermal con- vection ioop. (Ref, 58) Dynamic hot leg-cold leg tests have shown that maximum initial attack is about 5 mils in first thousand hours operation, and will average 2 - 3 mils per thousand hours operafion at 1500°F. On this basis, 40 mils of Inconel moderator cladding is expected to be sufficient for 10,000 hours full power operation., It is tolhé hoted that the reaction which may be expected %o proceed if BeO méééiéfibf.directly contacts fuel is UF, + 2 BeOg=>2 BeF,+ U0, 2 This reaction would gradually concentrate the fuel on the surface of the moderator rods. With unclad BeO, this deposition of UO2 on its surface would greatly retard the reaction. The nature of the Inconél corrosion is sfich that the corroded layer is chromium poor, and characterized by unicellular voids. However, tests have sfiown that even helium cannot penetrate the corroded layer. The strength is greatiy lowered, but, bérring fracture and peeling of cladding, the UF), - BeO reaction rate, even when entire thickness of cladding is chromium depleted, 1s controlled by rate of solid state 'diffusion of Be through the cladding., No great difficulty is expected on fhis point, fhe corrosiofi rate in the heat exchanger tubes, based on extrafiol&tion of 1500°F dynamic corrosion data with a 3000F hot-cold difference (Refs, 36 and 42) to 1200°F and 1000F differences is estimated as 10-12 mils maximum pexr 10,000 hours .operation, on fuel side of tubes. Another favorable factor 1s that the fuel is already chromium rich (from contact with hot moderator cladding) when it enters heat-éxchanger. This would tend to reduce the corrosion t0 an even lower rate. 53 %.1.4 Physical and Thermal Properties The physical and thermal properties of the fuel, as determined by calculation, and by derivation from data contained in Ref., 4O are as follows: Density Solid at room temperature (gm/cc) 4,09 Liquid ( ¢ - gnfcc, T = °C) | ©= k.03 - ,00095T Liquid (63; Ts/f45, T = °F) Q- 253.0 - .0328T Mean fiolumetric coefficient of liguid expansion per ©C 2,83 x 10~1+ Liquidus Temggrature about 525°C (977°F) Ehthalpy, Heat Capacity Solid (340° - 500) Enthaipy (cal/gn) Ht - Ho'C= -12.6 + ,0215T Heat capacity (cal/gm °C) Cp = 0.22 Liquid (5400 - 8940C) Enthalpy (cal/gm) Ht - Ho®C=2,1+0,318T ~ 4,28 x 10-91° Heat capacity at 1200°F Cp=0.26k Heat of Fusion (cal/gm) Hl - Hs = 57 Thermal Conductivity x (BTU/hr £t F) 0.5 (s0lid siab) 1.3 {liquid) Viscosity °F 1b/ft-hr £t%/hr 1100 - 23.0 | 0.098 1200 18.0 0.08k4 1300 14.5 0.069 1500 9.7 0.0k7 5k~ Prandtl Number 4.4 at 1100°F, 3.3 at 1200°F, 2.5 at 1300°F Volume of Fuel in Core - : L.77 % 105 cm3 Total Volume of Fuel 12.7% x 105 cmd U°3? Content of Fuel | 605 kilograms 4.1.5 Nuclear Properties The use of Fuel 30 and Inconel cladding on beryllium oxide moderator rods results ifi a rather large fuel concentration. Absorption cross sections of the sodium atom is higher than is dgsi:able and #ery 1ittle moderation is accomplished in the fuél. When testing'and_dgvelOPment work on nickel molybdenum alloys éfid.fuels containing lithium.and beryllium has feen completed, it is expected that critical mass-and.fuél concentration may be materially reduced, For example, where use of Fuel 30 éictates.that 4O mils thickness of Inconel éladding 59 used around moderator rods, use of nickel molybdenum might permit a cladding thickness of perhaps 15 mils, with accompanying neutron econouy and reduced fuel concentration. Incorporation of Li and‘Be fluorides in the fuel would give shorter slowing down length and a smaller size for the core. However, Fuel 30 and Inconel is the only system whose technology is thoroughly tested and found satisfactory at this time. h,1,6 Availability and Cost Reactor grade NaF is commercially available at $0.20 per pound and hafnium free ZrF) can be obtained at & cost of $3.50 per pound. To prepares fuel mix.for the reactor, powdered salts are mixed and then treated with hydrogen and hydrogen fluoride at 1500°F. This reduces the corrosiveness by removing traces of sulfur, iron, nickel,-water, chlorides and other impurities, Mixed, treated, fused 52 NaF - 48% ZrF) can be produced at ORNL (Ref. 57) for a cost of'$7.50'per pound in thousand pound quantities, =55~ - It was estimated that 20,000 -~ 30,000 pound quantities might be available for $6.00 per pound , | 4.,1,7 TFuel Addition The uranium burnup is compensafed by periodic additions of (NaF), UF,. From the phase dlagram, Figure 4-1, it is noted that dissolubion of this makeup salt in Fuel 30 proceeds so that only constituents of consiétently'iower melting points result. (NaF)QUFh may be added as pellets or powder &ireétly to the reactor., It may be melted and injected directly, or it way beldissolved in a small quantity of fused salt solvent and injected as,neéded. The fuel concentration is dictated by the operational temperature and amount of poisoning material in the reactor. As concentration falls or as poisons bulld up, the reactor critical temperature decreasés. Fuel must be.added when adjustment of the control rod can no longer maintain the desired operating core temperature. 4.1.8 Fuel Reprocessing (Ref. 5) Xe13? wi11 be_continfiously removed from the reactor, along with a part of the 1135 precursor, and all stable xenon and krypton isotopes, It is expected that rare earth fission firoducts will accumulate in the salt mix; thelr solubility limits the problem to one of neutron polsoning. Ruthenium, rhodium, and palladium plate out on metal surfaces. Reprocessing of the fuel after several years operation will be required to recover U7 from the spent, poisoned fuel before discarding radioactive waste, The fluoride volatility process, which depends on the high vapor pressure of UFg, is expected to allow uranium recovery with a minimum of effort. Thig process is currently being perfected at ORNL, Figure 4-3 is a flow sheet for ~56- $S8001J LI12A009Y wmiuea) LIT[I}RICA @prIoanig-31res pasnyg -- ¢-% "S1d 06061 OSMT-YT-TINYO Ad313ISSVTIONN Pj17- 40N FLSYM D0.069-009 NOILYNIHON 14 1S ddsn4d e Yin-"447 - 40N - . (NOIld¥0S3q) 40N JLSYM | Sy3gH0SaY B 19N00¥d °Jn a X d34d a3d ; 40N 4DN dvyl v o X | dvdL | F— 1 309 | ny | L = YSOdSIa A Density at 1100°F - 1.97 Gm/Cmd Specific Heat‘(Cp) - 0.57 cal/efi Viscosity, Centipoises, estimated at 620°F - 390 TO0CF - 200 8656F - 70 1100°F - 22 Thermal Conductivity - 2.4 BTU/Hr-Ft-CF -5Q= Viscosity, density and conductivity are given as predicted by responsible ORNL personnel., Actual measurements are in process, but special eguipment' needed was not available in time to permit determination before publication of this report. Bagis of deductions is the three component BeF_-NaF-LiF phase diagraum, 2 Figure k-4, On this diagram, composition selected is noted by T275: that ig, ternary eubectic melting at 27500. %.2.3 Disadvantages of Fluid (1) A comparison of the heat e#changer volume needed tao transfer 125 MW using this salt and using sodium has been made. It was found that the reactor pressure vessel size could be reduced congiderably by using sodium, with a consequent shield weight reduction. | (2) Melting poinf of the fluid, 5270F is so high that a shutdown of the secondary system pumps would require that system be drained to prevent freeze up of salt. Considerable care must be exercised to assure that boiler feed water is preheated before introduction into boiler, or freeze up may résult. | (3) Pumping power is considerably greater with fused salt fluid than with sodium, because of greater viscosity. -60~ 693 b8 066 40N j i ; T 417 Qt e 0.G3 (6499 - 411 - 4ON) C6e¢ GG¢ 08t 498G -40N %409 - 41712 OrAY (2498¢ -317-40NGS)— gie3 2ee3 b1Z 08¢ \ 5 2508, 40N (2498 - 417) 9Gg3 0Le4d Do NI 34V SIUNLYHIJSWIL SNCINDH SISHHLINIYVL NI NMOHS SAONNOdWOD SNdINosans J1L03L1¥3d =d ONNOJdWOD AYVNY3L =021 2HS 2409 DI1D3LNT AHVNYIL =1 21L03LN3 =4 1SS OMO-YT-INFO IVILNIQIANOD wBlw 5.0 MATERIALS SELECTION 5.1 Structural Material Once Fuel 30 had been selected, the results of several years testing (Ref. 39) the dynamic and static corrosion resistance of structural materials made perfunctory the selection of Inconel as the primary structural material.. This included 1ts use for pressure vegsel, moderator cladding, primary heét exchanger structural material, pumps, and all other surfaces in direct contact with the fuel e#cept the nickel reflectors, As indicated in our discussion of the fuel, Sec. 4.1, developments in nickel molybdenum alloys now underway are expected td change the fuel and container materials picture in the foreseeable future. Nevertheless, this design study 1s based on present technology and already proven systems. (See Sec. 4,1.3 Corrosion) Some of.the results of corrosion research islpresanted as Jjustification for metal thicknesses énd materials chosen., Figure 5-1 shows stress'elongation and rupture curve for Inconel tested in Fuel 30 at 1300°F. Figure 5-2 shows temperature dependence of gtress rupture properties of Inconel in Fuel 30. Figure 5-3 shofis effect of section thickness on creep-rupture properties of Inconel tested in Fuel 30 at 1500°F at 3500 psi stress., Figures 5-1, 5-2, and 5=3 support choice of Inconel with Fuel 30, and thickness of tubing and cladding specified., Our specification of 40 mils wall thickness of primary heat exchanger tubes is based on research leading to Figure 5-3 and advice by informed ORNL personnel {(Ref, 42). Figure 5-3 indicates that creep resistance of Inconel immersed in Fuel 30 at elevated temperatures shows a remarkable improvement when section thickness reaches 4O mils. -62- 000°0F 0006 C9ESE IMC-YT-TINYO Q31dISSVIONN (worydeD yjrm 10a098) 4 L0081 38 08 "ON HeS pasnd Ul pajsa] [9UOOU] POATIVSY-~-SVY I0] 2AaIn) udiseg -- 1-¢ "8td (44) INIL 000¢ 000} 004G 00¢ OO0/ 0g 0¢ Ol G J¥NLdnNy | Y oL G o6 2 Yo b % G0 000 000¢ 0006 000 0} 000'02 (1sd) SS3IUIS -63- 000 0¥ 000% el OMO-41-TIN¥O Q31AISSVIONN Q002 (worided yjtm 38x09g) Qg "ON I[eS pesng pue uoday ur 4 (0S91 PUB 00ST ‘00%T }® PRISSL 19U0DU] PoATEO9Y -5V JO serlasdoad sanidny ssaxig ayi Jo uostaedwo) -- (44) IWLL 000} 00¢ 00¢ OO0} O¢ 0¢ O Z-¢ "314 000} 000¢ 000¢ 0000} 000'02 (18d) 8S34IS TIME (hr) 800 700 600 500 400 300 200 100 -84 - UNCLASSIFIED ORNL—LR—~DWG 17921 Fig., 5-3 -~ Effect of Section Thickness on Creep-Rupture Properties of As-Received Inconel Tested in Fused Salt No. 30 at 1500° F under 3500 psi Stress, 52 % 37 % — S, 0 /' RUPTURE 45%,/ | ey /V’dr’ 5 %o ® 159 f A~ L —— T 6‘70 2c70 A — 1 ] o 20 40 60 80 100 120 SPECIMEN THICKNESS (in.) {(Secret with Caption) 140 (x10™2) ~65- Insofar as 1t is possible to predict from dynamic and static corrosion data, Inconel thicknesses have been chosen so that, after design lifetime ‘has passed, sufficient sound void-free metal remains to provide stress registance édequate for the barfiicular use involved. | Inconel has also exhibited superior resistance against chloride stress corrosion over most conventional materials, Because of the severe prohlems that have been attributed to this in the steanm generating equipment of both mobile and stationary filants, it is recommended that it be used for both the steam and salt side of this equipment. 5.2 Moderator Since all surfacés in contact with the fuel were of necessity Inconel (except nickel) it was necesséry to choose a moderator of low neutron absorption which would permit the reactor to go critical with a reasonably small core volume, 'Consequently, after investigation (Ref. 43) BeO was selected as the leading prfiven moderétor which could withstand the temperatures ‘expected. A cladding of 40 mils vas considered necessary, as previously discussed in Sections %4.1,3 and 5.1, Whi1e this thickness of Inconel cladding does not make for neutron econony, or for low fuel loading, nuclear calculations indicated that the réactor could be expected to. operate satisfactorily. One inch diameter test pleces of BeO ceramic were exposed in the MTR and showed satisfactory thermal stress resistance (Ref. 45, 55), The diaméter of 3/4 inch selécted for this application was based on extrapolation of these results to the higher energy deposition rate expected. See Appendix 6.1. 66 5.3 Reflector Because of the poor moderating proPerties of the fuel and the somewhat high thermal neutron capture cross sectlon of the -core, due to Ne and Inconel, a8 fast neutron reflector constructed of pure nickel wasg choeen. Consequently, calculations indicate a large percentage of epithermal fissions. The only fair heat oonductenoe of nickel necessitates the circulation of a small portion of the fuel through the reflector to equeiize temperature and lower thermal stresses. ‘It is not considered necessary to ¢lad the reflector for corrosion resigtance, which is satisfactory unclad, 5.4 Poisoned Moderator Region An annular ring of boron beering, beryllium_oxide rods, clad with Inconel, with interstices filied with borOn bearing Inconel.rods forme the neutron shield, Calculations based on an everage thermal flux of_lo}o show that helium generation over a period of iQ;OOO full power hours is about .0l cm3 (STP) of helium per cmd of BeO, Since BeO may be about 6% of theoretical density, no significant pfeeeure will be generated, Between the beryllium-boron region and the heat exchanger region an Inconel clad, copper-BhC cermet layer is interposed as a thermal neutron abeorber, to prevent escape of thermal neutrons to the heat exohangere. The beryllium-boron regionm, although heavily poisoned, containg a eource ‘of thermal neutrons due to thermalization of fast neutrons from the core, Copper-BhC haslbeen satisfactorily fabricafed, cOntaining 25 volume percent of BjC, to a fheoretical density of 95%, by cold pressing and hot rolling (Ref, 36). -67- 5.5 Design Properties of Materials Appendix 5.1 shows the design properties of Inconel, beryllium oxide, and nickel used in this . study. ~68- 6.0 REACTOR AND PRIMARY HEAT EXCHANGER DESIGN 6.1 Introduction The basicbreactof design is conceived as being a pregsure tight cy}indrical vessel cofitéining a circuléting fluoride fuel., A primary objective of the design was to minimize its size and weilght in order %o redfice its contributibn to the overall system specific weight. In addition, & small reactor design is desirable because of the large effect it nay “have in turn on the.size of both the primary and secondary shield, The volume of the reactor is basically dependent upon; 1) the | nuclear properiies of the fuel as it affecfs both the.éritical size and limiting power densities, and 2) methods which can be devised to remove the_fission heat from the circulatihg fuel, The establishment of an -allowable critical size and fuel loading as well as other nuclear con- siderations are discussed in detail in Section 8.0, The methods of selection and optimizing a heat exchanger configuration are presented later in this section, (See 6.3),. Pogsible alternate solutions or approaches‘to the various problems are discussed in the appropriate sections along with tpe reasons (either | engineering or arbitrary because of time limitations) for the selections made, 6.2 Reactor The basic configuration, illustrated by Figure 3-2, is approximately 80 in. in diameter and 80 in, high. Its total net weight is calculated to be 69,700 1bs (Appendix 13.1). Centrifugal fuel pumps located in the -69- reactor head are used to circulate the molten fluoride fuel up through a central c:itical region, and then through an annular peripheral downcomer which contains the primary heat exchangers. Heat is removed in this region and the fuel is again circulateq up through the core. 6.2.1 Internal Arrangement Calculations for the central core region were based on it being qui#alent t0 a cylinder 75 cm in diameter by 80 cm high, This was modified for design purposes to an octagon shape for a more even moderator rod spacing and tapered ends to gain extra core volume. An optimum volume fraction of fuel for the core was found to be 50% (Section 8.0). Fuel cooled cylindrical beryllium oxide rods, clad with Inconel for corrosion resgistance, were used for moderation purposes, These were equispaced throughout the core on a triangular pitch under the. distance between centers being defined by the rod size and the desired volume fraction. Taper fittings were utilized at both ends of these rods to provide for the proper area and flow distribution. These rods would be held in ‘the bottom support plate by & bayonnet joint and left free to expand in an axial direction to eliminate thermal stresses. A hollow ring is attached to each rod at the end of the upper taper. This will maintain préper rod spacing and still.provide a suitable flow passage; These rings may be interlocked to prevent rotation and hence uncoupling of the bottom bayonnet Joint, but still allow free axial motion, The effective vertical boundaries of the core region are fixed by poison material located in the ends of the moderator rods., This poison material, beryllium oxide plus boron 10, also helps to reduce end leakage as well as to cut down on fissioning in the entrance and exit plena, ~70- A 40 mil cladding of Inconel is required around tfie beryllium oxide to provide propér corrosion resistance for the 10,000 hr design life (Ref. Section %4.1.3), Since this thickness is fixed and not a function of moderator rod size, it is of considerable nuclear importance to use fewer large rods rather than many small rods in order to reduce the total amount of Inconel poison within the core., However, the maxifium gize 1s limited not by the nuclear aspects such as self shielding of the fuel, but by thermal stresses due to heat generation within the woderator material, The feasibility of using beryllium oxide as a moderator material has been satisfactorily demonstrated under cyclic reactor conditions in the MTR (Ref. 54 and 55). Using this information, calculétions were made to limit the design stresses for the present system to that found to be allowable in the above tests (Appendix 6.1), This limited the moderator rod size, without cladding, to approximately 3/4 in? for the preseni. Because no indications were found in the MIR tests to indicate that higher power densities could be allowed, this minimum size could possibly be increased in the future when substantiated by additional test programs. Calculations of the temperature rise across the boundéry'layar (EOOF) and through the mcderfitor rod (143°F), also included in Appendix 6.1, indicate that a maximum centerline moderator temperatfire.of 1491°F is to be expected. This is well within the operational limits of this materiél and approximately equal to that of the MIR tesis. The moderator elements may be fabricated by inserting slugs of BeO B/h in. in dlameter by 2 in, long into Inconel cans of sultable wall thicke ness. In the MIR tests, improved heat transfer out of the moderator material was reallzed by utilizing helium in the small clearance gap required between -T1- the slug and cladding. Stress calculations indicate that a shrink fit of the cladding around the BeO could be used in conjunction with the above to obtain further improvements. A nickel blanket, approximately 6 in. thick, is incorporated around the cylindrical side of the reactor core to offer advantages both as a reflectaf and a shield, High energy leakage neutrons are inelastically reduced to a lower energy level and scattered back into the core %o improve the .core criticality and power distribution., Also because of its close proximity t§ the core 1t acts as an effective shleld, from a weight standpoint, for both:prompt gammas and neutrons. A detailed stress and heat generation analysis was not made on the reflector, However, because the reflector supports no load other than its own weight, it can be alloved to operate at high'temperatures and in the plastic region so that thermal sfiresses may be effectively annealed out. Fuel flow channels of approximately 2% by volume should be more than adequate for cooling the reflector. In order to minimize the dctivation of the secondary fluid, it is necessary to reduce the.neutron flux in the primaxry heat exchanger region as much as possible. To help accomplish 4his; a region.containing BeO to thermalize fast neutrpns and boron to capture the thermal neutrons is included outside of the reflector. This region contains closely packed 3/& in, cylinders suitably clad with Inconel and 1s approximately 5-1/2 in. thick. Small boron bearing Inconel rods are placed in the interstices of these cylinders for additional shielding and %o reduce the fuel and hence fissioning in this region., Sufficient flow areas will still exist within the interstices of fhe large and small rods to provide for cooling. To assure absorption of neutrons that are thermalized in the outer ~72- edge of the above region a thin layer of boron carbide in a copper matrix is then placed around the above region.. The feasibillty of using these materials are discussed in Section 5.4, A single control rod thimbie, approximately 4 in, in diameter, extends through the length of the.caré. A clearance gap of 0,1 in, on the radius is allowed between the thimble and the control rod to assure free operation. To facilitate fuel cooling of the poison rod this gap would be filled with either a salt or a liquid metal. A small reservolr éould be included in the reactor head in order to keep the'fihimble full as the controi rod is withdrawn. For the purpose of the control rod worth evaluation (Section 8.2.3) it was assumed that this gafi waé filled fiith sodium. Because of t he small quantity involved it was felt that this wéuld not be a serious hazard. | A low point drain hole is located at the bottom centerline of the reactor vessel to provide a place for both filling and locating an emergency dump or‘blcwout valve. This is incorporated into the bottom lateral support of the control rod fihimble. A thermal shield is located Just outside of the heat exchangers to reduce the gemma and neutron heat generation problem in the reactor vessel. A small gap is placed between the thermal shield and core vessel to provide . for cooling. Relatively cool fuel from the exit of the primary heat exchafiger fiill flow up through this gap and into the fuel expansion tank in the reactor head. This flow hag the additional advantages of providing increased circulation through the head to remove decay heating and to decrease the temperature in this region to help alleviate the snow problem (Section &.1.2), =13~ As mentioned previously the moderator rods are fixed only at the bottom in order %o allow free expansion and thereby reduce the thermal stress problem. For a similar reason, the remainder of the internal structure, that is, the reflector, control rod thimble, and the basgket supporting the poison rods, are suspended only from the reactor head. The only exception to this is the primary fieat_exchangers which run straight through the vessel. However as explained in Section 6.3.k4, the thermal stresses obtained were found to be tolerable, 6.2.2 Vessel Design One of the major advantages of a fused salt system is that due to the low vapor pressure of the fuel, it is necessary to contain dnly small pressures wlth the core vessel. With a minimum pressure of 30 psia required within tfie system to prevent pump cavitation and a pressure rise of approximately 35 psi requiréd across the pumps to provide fuel flow a normal design differential pressure of=on1y 50 psi is obtained, Basilcally this would require a wall thickness of less than one-half of an inch, However both because off design conditions would undoubtedly occur and navy requirements of meeting 20 to 30 .g shock loads are required, this thickness was increased to 1-1/2 in. using the ground rules proposed in Ref. 11, but adapted to Tnconel, “Although not shown in the reactor drawing, Figure 3-2, cooling coils must be included in the head design to také care of internal heat generation. The possibilities exist of using either the pressure drop across the fuel pumps to force the flow of a small amount of Ffuel through suitably con- structed cooling tubes or to circulate a small percentage of the secondary salt, 7l Two possibilities which exist as to the most economical method of maintaining reactor installations of this type are discussed in Section 11.5. Basically they affect the vessel design in two different ways: 1) the vessel should be designed so that 1t may be reasonably feasible to assemble and disassemble 1t seve?al times, or 2) the design should be simplified with the idea that only one assembly wbuld be required. Because these concepts were well beyond the scope of this study it was decided to present a reactor design that could satisfy both. To accomplish this only the feasibility of a system allowing disassembly had to be shown because this was the most complex. The alternate solution, being of simpler design, | was not illustrated as the Joints, flanges, etc., would just be changed to welded structure. In both concepts as deséribed, it was felt to be- desirable from both an economic and a Weight standpoint to remove the reactor from the ship for any maintenance. To facilitate easy removal, the head is simply butted against the dore véssel and held by the use of a fianged Joint. An omega type seal is welded across the Joint to provide proper leak tightness. This isg in turn backed up by a steel "0O" ring both as a safety precaution against possible fallure of the omega seal and to help prevent fuel from easily - flowing into the ring. If fuel settled in the seal fing, it would éddl to the decontamination problem upon reactor disassenmbly, Hofiever, corrosion would not be a problem because the chieflsource of corrosion with Inconel is8 with a dynamic system fiowing over a large temperature difference; Be- cause flow is prevented in this region, the seal weld will remain at con- stant temperature and corrosion would be limited to that caused by the initial chromium solubility. ;75_ The reactor may be disaésembled remotely by removing the hold down bolts and cutting the seal weld; Omega type seal welds of the type recommended should not present a problem as they have designed for use in pressurized vater reactors at pressures up to 2500 psi, Also mechanisms for remotely cutting and rewelding these types of owega seals have been developed for use in the marine PWR systems, In order to be able to remove the reactor head and replace the primary heat exchangers it is necessarylthat the secondary fluid inleit and exit“ pipes be detachable from the core vessel, Two suggested methods for doing this are illustrated in Figure 3-2. The one in the reactor head utilized a concentric tube-with the Jjoining weld being made approximately 12 in, off the reactor head for access purposes. This type of joint gives good vrigidity but has the disadvantage of allowing only a limited number of welds to be made. Also since it is a strength weld it wouid be more difficult to mfike remotely., The bottom connection is fashioned after a bridgeman closure which is used on many high pressure autoclaves. The closure provides the structural strength while an omega seal weld similar to that previously described is used to assure leak tightness. However, the rigidity of this type of connection under side loads and thermal cycling is not known. Methods for the head closure and secondary pipe attachment were not given detailed consideration bub are offered as one of many possible solutions. 6.2.3 Structural Arrangement Two possible solutions exist for supporting the basic structure, however, a detailed study would be required to determine the optimum., The 16— first of these as shown in Figure 9-3 simply rests the reactor on supporting structure allowing free vertical expansion. Side play would have to be. limited by guldes. A second approach which at first hand appears to be more advantageous would support the reactor through a beefed-up section Just . below the head flange. This would not only take the Welght of the head and most of the internal structure of f the reactor side walls but also simplify the ba51c installation. 6.2.4 Fuel Punips Three'centrifugai pumps are located in the reactor head to provide for fuel flow and to &id in the removal of the fisgion product gases. These pumps have common inlet and exit plena and are sufficiently overdesigned 50 as to allow almost full'poher'reactor operation in the event of a'single pump failure, To provide for.a lightweight and variable speed system (required to - compensate fér pump failure) a steam turbine driven motor was selected for tfiese pumps, A small_AC.eiéctfic_motor which could be clutched into the drive sgaft would also be incorporated to maintain circulation under zero power operation. Also becafise'if could be gwitched into the ship's emergency electric power system, it would serve as a safely device in case the steam flow to the turbines was interrupted.‘ A more detailed description of these pumps is given in Section 11.8, 6.2.5_ Pressurizer and Expansion Chamber 6.2.5.1 Pressurizer It is necessary to provide & pressure of at least_ 15 psig at the inlet of the fuel pumps to prevent cavitation., This pregsure is appiied by means of bottled helium gas at startup. After startup, a helium gas differential pressure of a few pounds is maintained at pump shaft over that in the expansién chamber, 1o prevent escape of fisgsion product gases along pump shafts. After initial filling, stable xenon and krypton generation can be used to maintain pressure. Off-gas systems to provide for poison gas removal are discussed in Section 11.6. 6.2.5.2 Expansion Chauwber -Thé pumps are so designed as to cause a swirling motion of fuel in the expansion chamber, so that equilibrium gas-liquid cdncentration' is quickly reached; A small stréam of 11750F fuel 1s brought up to the chamber through a passage between thermal shield and core vessel, ané circulated thrdugh the chamber to remove heat generated in fhe chamber by fisgion product decay and by fission, (See Appendix 11;2 for heating calculations). It is calculated that 150 kw is generated in gas, 157 kw is generated in liquid due to fission, and 93 kw is generated in liquid due to decay heat, It is obviously necessarfi that some heat removal system be incor- porated to cool off the expansién chamber roof due to this and internal heat generation as discussed in Section 602_02° Assuming that one-half of the gas heat is absorbed by the roof, a cooling rate of 75 kw, or about 250,000 Btu/hr would be expected at full power. Allowing a 50°F rise in temperature, this will require circulation through the head ofbabout 8800 1b of fused salt per hour, A stream flow of 50,600 1b of fuel per hour is required to provide cooling for liquid in the expansion chamber to prevent snow formation, Arrangements have been made to bleed off a stream of fuel from the cool region (1175OF) at the bottom of the reactor, so maximum fuel temperature 18- in expansion chamber should be the same as maximum temperature in resctor, that is, 1275°F with a conservatively estimated temperature rise of 1100°F, Thus, maximum temperature of liquid in expansion tank will be about 750F less than 1350°F,'maximum temperature at which'sndw problem may be neglected using fuel 30 (See Ref, 50). 6.3 Primary Heat Exchanger 6.3.1 Desgign Criteris The.design criteria for the?primary-heat'exchanger is the same as for the system as a whole; that is, obtaining the lowest specific ~weight for the overall power plant consistent with a life of ten thousand full-power hours, Méeting this goal required the 0ptimiéation of a com- bination of several quantities which fary with heat exchanger design. These are: heat exchanger weight, primary shield weight, pump weight, and pumping horsepower. Tfie variables of the heat exchangef design were placed, essentially, in two categories: 1) those which could be fixed early in the study dependent on the expefience of others doing similar work or due to the limitations imposed by the rest of the systefi, and 2) those which were varied in an extensive parameter study to determine the most favorable union of these quantities, 6.3.2 Basic Design The primary heat exchanger is of once-through, counterflow design, The heat transfer surfaée is providefi by straight Inconel tubes on a delta laftice which are contained.in an annulus surrounding the reactor core, ..79_ The headers are segments of tori which have an elliptical cross section. These headers circle the reactor core at the top and bottom of the heat exchanger, each segment having a nozzle which penetrates the pressure vessel and primary shield (éee Figure 3-2)}. To provide additional area on the header surface, the major axis of the ellipse is longer than the width. of the heat exchanger and is tilted'with respect to the horisontal, The'secondary coolant flows through the tubes, entering at the bottonm of the heat exchanger. The fuel flows on the outside of the tubes and enters at the top of the exchanger; The physical dimensions, flow rates, temperatures, temperature differences; and heét transfer coefficients for the final primary heat exchanger design are tabulated below. This heat exchanger would be capable of removing 125 megafiatts ofbheat from the reactor. See Appendix 6.2 for calculational details. Heat Exchanger Inner Diameter | 53.5 inches Heat Exchanger Outer Diafieter 73.7 inches Heat Exchanger Length 48 inches Tube Inner Diameter - | .120 inches Tube Outer Dianeter .200 inches Tube Spacing .030 inches Fue:l ¥Flow Rate 16.2 x 106 ibs/hr Secondary Coolant Flow Rate 7.48 x 10 1bs/hr Temp. of Fuel Entering Heat Exchanger lETSOF Temp., of Fuel Leaving Heat Exchanger 1175°F Tem§° of Coolant Entéring Heat Exchanger 10500F Temp. of Coolant Ieaving Heat Exchanger 11500F «80- Mean Temperature Difference from Fuel to Secondary Coolant | | 1250F Outside Heat Transfer Coefficient - 1836 BTU/Hr—OFnF’G2 Inside Heat Transfer Coefficient 914 BTU/Hr-CF-Ft° Overall Conductance | 37h BTU/Hr-"F-Ft° Straight tubes rather than U-tubes were incorporated in the primary heat exchanger because it would have been difficult to obtain as much : heat transfer area in a given volume with U-tubes. Also, inlet and outlet headers would have to be in the same end of the reactor, which would further complicate the space problem. The wain advantage of a U~tube exchanger would be the reduced 10ngitudinal thermal stresses, However, as will be indicated in a later section, longitudinal thermal siresses are not expected to be a major problem in this heat exchanger. | The heat exchanger inner diameter and effective length were detérmined by the reactor core design., It 15 necessary that the heat exchanger tubes be nested closely about the reactor, and be sbout the sane length as the reactor, in order to achleve fihe most compacf design, Tube wall thickness was fixed at .00 inches, primarily because of corrosion to be expected during ten thousand hours of operation. Although no corrosion data are available at the temperatures encountered in the heat exchanger, 1t has been predictéd that a maximum corrosion of twelve mils on each side of the tube could be expected {see Materials Section h.2)} This corrosion is of a penetrétive nature, with the maximum depth of cprrosion being given for a few scattered displacements., Since it is unlikely that penetrations on both side of the tube would line up, and because the dis- placements are not interconnected, a large safety margin is realized in the -81.~ twelve mil estimate. However, twenty-five mils were allowed for corrosion, with the remaining fifteen being sufficient to contain the pressure and thermal stresses. The upper fuel temperature was set at 12750F to keep the corrosion within acceptable limits. From examination of other proposed reactor systems of a similar nature, a mean temperature difference between the fuel and secondary coolant of 125°F was decided on. This is a compromise value which will give both reasonable heat transfer and rermisgible thermal stresses., Further investigation of this system should include an examination of the effects of changing the temperature difference. Many considerations were involved in the selection of a 1L00°F temperature drop across each fluid circuit, It is desirable to keep the temperature drop as large as-possiblg in order to reduce the flow rates, and hence, pumping requirements. Also, it is necessary to keep the temperature of the secondary fluid above the melting point of the fuel which is 970°F, Using a mean temperature difference between the two fluids of 1250F and a 100°F drop across each circuit, the lowest temperature encountered in the secondary coolant loop will be 1OSOOF, which should be safely above the fuel melting temperature. Since the heat ekchanger must be capable of removing 125 megawatts or 4,27 x 108 BTU/hr, choosing the temperature drops automatically sets the flow rates. In the final design, the tubes were spaced .030 inches apart on a delta lattice. The delta lattice was chosen over a square lattice because it permitted inserting more tubes of a particular size into a given space. The .030 inch spacing was established by a parameter study which will bve -82- demonstrated in a later paragraph. The tube spacing can be maintainéd by one of several wmethods, Most of the present small-fube high performance heat exchanger tests utilize flattened wire spacers, which are perpendicular to the tube axes. It was for spaceré of this type that heat transfer and pressure drop cal- culations on the fuel side of the heat exchanger were made. More recently, some work has been done with helical spacers, wrapped about each tube. Preliminafy results indicate thatlthis type of spacers will give about the same heat transfér with a ldwér pressure drop. To facilitate welding, it is necessary that the tubes be spaced at least .075 inch apart on the tube header (see Ref. 67), This requires that the headers have a surface area greater than the cross sectional area of the heat.exchanger, but preferably will fif into the same annulus. As described previously, this waé accomplished by méking the headers elliptical in cross section and tiliting the ellipse with respect to the horizontal. The heat exchangér will be fabricated in bundles of approximately six hundred tubes each, -This is approfiimately twicé the number of tubes per bundle presently contemplated for the more ‘complex ART fuel-NaK heat exchanger (see Ref, 36, Section 4.1). -Each bundle is to be tested individually in order to simplify inspection and preclude the necessity of scrapping an entire heat exchanger for a single tube-header Joint failure. 8ix of these bundles will then be welded together and capped to make up one header segment. There will be twelve such segments, each one having a nozzle penetrating both the upper and lower heads, ~83= 6.3.3 Parameter Sfudz | In the pérameter study tha£ was made, the variables were tube outer diameter and tube spacing. For a given tube diameter and spacing, a heat exchanger outer dlameter which would give the required mean tem- perature difference of 125° fias determined by an iterative process. For é selected tfibe size and spacing, an assumed outer diameter of the heat exchanger was used to calculate film coefficients., Flow in the tufies vas at all times leminar and an empirical equation for film conductance during laminar flow (see page 232, Ref. 17), _ K 1/3 hi = 1,75 "EI was used. For flow outside the tubes an experimental correlation 0 (See Figure 7.6) O h = ell'T .u.ék_]g—. (Prf)oh‘(Re)o36 was used, which takes spacer effects into account. An expression was derived to give the weight of the heat exchanger plus primary shield - for each configuration. Pressure drops and pumping horsepower can be determined from flow rates and heat exchanger geometry., Pressure drop in the tubes was calculated from (See pages 45 and 50, Ref. 15) where -84 for laminar flow. For flow outside the tubes, an experimental expression for friction factor f = 5°T . (Re) 50 was used, This.expression includes the effect of spacers and is given in Ref', 13. The weights of the pumpé and drive motors were estimated at 25 1bs/FHP. The weight of the machinery and equipment not affected by heat exchanger design was determined, it was assumed that the steanm generation equipment could provide steam for 35,000 shaft horsepower, normal auxiliary équipment, and 600 pump horsepover, When the calqfilated pumping horsepower was less than this, the shaft.horsepower was increased by the diffgrence divided by the efficlency of the pumps, The total weight was then divided by.the adjusted shaffi'horsepower to, give the specifilc welght. Although this method will not give. the exact specific weight of the power plant, it will indicate the configuration vhich will give the lowest specific weight. Tube diameters were varied from ,1875 inches to .25 inches and spacing from .020 inches to ,04O inches. The results of this study are shofip in Figuré 6.1, 6.3.4k Stress Considerations Exténsive thermal stress calculations were not wmade for the primary heat exchanger. However, the thermal stresses due to the tem- perature drop across the tube walls were determined, and also the stresses which will be present due to tfie difference in longitudinal expansion of - the pressure vessel and_heat exchanger tubes for several extreme cases were calculated, (S3IHONI) ¥3IL3IWVIQ ¥3L1NO 28 08 8. 92 b2 22 02 * | \.\\\QWI/ 8°8G 3 = TN & b A s¢ / — 3 » 0°'6S % , . 5 \ —— : \ v \t\\\“m_\oqmm ._.__zy/ 2°6G b 66 SONIOVHS 38Nl SNOIYVA HO4 H3ILIWVIA H3ILNO HIONVHOX3 LVIH AYVWIHA SA OIlVY 43IMOd 7/ LH9I13M 40 31VWILS3 1’9 34NOI i -86“ L Due to the rather poor heat transfer characteristics of both fluids in the heat exchanger, most of the temperature drop is taken across the fluld films. With a small temperature drop across the tube wall, the thermal stresses.are also quite small. It was estimated that the contéinment vessel will be at an average temperature of approximately 12250F, The inside of the vessel will be cooled with fuel having a temperature of 1175°F and the average tem- perature will be some 50°F greater than this due %o heat generation in the vessel. The tube temperature can be thought of as being at an average between the mean wall temperatures or about iléSOF. This gives a‘temperature difference of 360F between the pressure vessel and heat exchanger tubes, It was assuned that this difference in thermal elongation would be taken up by mechanical elongation of the heat exchanger tubes only. This was figured for several extreme situations, one in which the tubes ran straight from one header to the other and were fixed at both ends. In this case, the stress in the tubes remained below the yield stréss. Another case was considered in which the énds of the tubes were bent at right angles and then fixed to the header. In this case, the difference in elongation was assumed to be taken up by deflection of the étub ends. Maximum stresses will again remain below the yield strenéth if the stub ends are at least .4 inch long. In the event that detailed thermal stress calculations prove thét straight tubes are untenable, the tubes could be wrapped partially around the reactof to provide flexibility in ofder to alleviate these stresses, The grestest pressure difference across the tube wall will be less than 100 psi even if the pumps on either circuit should fail, A cal- -87- culation was made, assuming that the pressure inside the tubes was 100 psl and the necessary wall thickness came out to be only .01 inch inches. For the headers, thermél stresses were not considered; however, pressure stress calculations were made, again assuming an internal pressure of 100 psi. For this condition, a wall thickness of .1 inch and end cap thickness of .375 inch were determined. -838- 7.0 STEAM GENERATING SYSTEM 7.1 Introduction One of-the wa jor problems in adapting nuclear power to naval vessels has been the development of a dependable steam generating system that will deliver steam at conditions that are compatible with ‘the requiréments for efficient steam turbine performance. In the design of the steam gefierators the group endeavored to duplicate the existing steam conditions of +the 931 class destroyer, Thesé conditions are 263,300 1b/hr per boiler room at‘950°F and 1200 psi. it was decided to replace one boiler room with & nuclear_reactor-steam generating sygtem.' Other design criteria were to keep the thermal stresses as low as possible, tb make the system as light and compact as practical, and to have a realistic and somewhat conservative system, | In the preliminary analysis of the system it was decided that the reactor power shquld be 125 megawatts, Therefore, the steam géneréting equipment was designed to remove 125 megawatts of heat. When the actual steam cycle data for the class 931 destroyer was received a complete heat balance revealed that only 95.9 megawafts of heat was necessary to supply the steam for the full power of 35,000 shp. This makes possible the operation of the steam generator at lower temperatures and lower.zst's throughout the system. Detailed calculations of the design speclfications are presented in Appendix 7.1. 7.2 Molten Salt Cycle Selection In selecting a cycle or rather a system for steam generation the -89- group was confronted with the problem of removing 125 megawatts of heat from a molten éalt coolant., The temperature of this salt as it leaves the primary heat exchangers is to be 11500F, and it is to reenter the primery heat exchangers after losing only 100°F in temperature. There- fore, to remove the full-power 125 megawatis 1t is necessary to circulate 6 1b/hr., One of the design criteria the molten salt at a rate of 7.49 x 10 is to keep the temperature drop across tube walls below 1O00°F and since the saturation temperature of water in the boiler is 572OF, then when allowance is made for the bolling water film température drop and for the molten salt £ilm temperature drop, it is found that the salt entrance temperature to"the boiler should not exceed 800°F. The superheater is to bring the steam temperature up from 572°F to 9750F, Because of the high temperature drop across the steam film an entrance temperature to the superheater of 1150°F and an exit temperature of 1126° would not exceed the 100°F drop across the tube walls, Therefore, the problem is to'lover the salt temperature from 1126°F to 800°F and then to raise it from 73M0F, the boiler exit temperature, to 1050°F in order to return it to the primary heat exchanger. The two methods considered for meeting this problem were to use either a regenerator heat exchanger between the boiler ang superheater, or blenderé° In the regenerator system the .1126°F salt would enter and the 1050°F sait would leave the hot end of the regenerator heat exchanger while 800°F salt would leave and 734° salt would enter the cold end. This meant that some 420 megawatts of heat would have to be exchanged in the regenerator. Due to the rather poor heat transfer characteristics of the molten salt and the low log mean temperature difference avallable, a tremendous heat ~50- transfer area would be required. This in turn led to a very large salt volume, high pumping power, and prohifiitive welght and size. Attention was then turned tb a blending arrangement as a means-of achieving the desired salf temperatures, It was found that blenders were being considered by a fused salt power reactor group at ORNL, (Ref. T72). After consultation it was decided that a blending system would be used thereby allowing the system to consist of a separate hot and cold loop {See Figure 3.1). The hot loop circulates the molien sali coolant from the primary heat exchanger to the Superheétef, from the superheater through the pump and blending apparatus, and then back to the primary heat exchanger. On the discharge side of the hot loop pump, molten salt at 1050°F is tapped off afid fed into the'éold loop, This flow can Be regulated by means of a trim valve and serves aé the heatl source for the cold loop. The cold loop‘contains the steam generator or' | boiler and a pump. The hot salt ié fed into the cold1100p on the suction side of the pumpvfrom'whence 1t trans#efses the stéam generator. From the cold side of the boller the requisife amount of salt is tapped off and fed into the suction side of the hot loop pump.. This completes the path of molten salt through the circuit. The salt flow rétes in the superheater and steam éenerator are to remain constant at 7.49 x 6 10° 1b/hr. The amount of salt bled from the hot loop to the cold loop at full power 125 mw 1s 1.74 x 106 1b/hr. Schematic layouts of this system are also shownlin Figures 7.3 and 10.1. The fluid horsepower necessary to circulate the molten salt was calculated to be 260 hp in the hot loop and 200 hp in the cold loop. The calculated pressure drop in the superheater was 15.2 psi, in the -91- primary heat exchanger 20.5 psi, and in the steam generator 38.7 pai, Since no finalized piping layout was attempted, the pressure drops due 0 line frictlon, bends, valves, entrances, etc. was estimated., It is reasonafle to assume that these losses would not be as significant as those in the primary heat exchanger, boiler, and superheater. There- fore, the molten salt pumping horsepower should not vary greatly from the above values, In order to determine the optimum salt line size a short parameter study was undertaken. Pipes with inner diameters from 7" to 17" fiere investigated. The pressure drop per foot of pipe length was calculated and from the resulting fluid horsepower a pump weight equivalent was obtained. (See Figure 7-7). This was combined with the weight of the salt per foot of pipe length and plotted against the various pipe diameters. The results showed that the optimum pipe 1.4. would be approximately 11". 7.3 Steam Generator T.3.1 Types Considered. In selecting a steam generator the genersl types considered were (1) the flash boiler, (2) the once through boiler, (3) the natural circulation, and (4) the forced circulation boiler. Each was given serious conslderation and the conclusions drawn about each type follows: (1) Flash Boiler: The only information found about flash boilers was contained in Reports EPS-X-265, EPS-X-270, and EPS-X-288 by the MIT Engineering Practice School at Oak Ridge. In these reports it vas pointed out that the main advantages of flash boilers are that they -92- are capable of responding rapidly to 163& dgmands because of the small amount of water contained therein and that it is probable that a high capacity boiler of reasonable size could be constructed. Tn the past the chief disadvantage of flash voilers has been tube burnout, but this problem is absent in nuclear reactor applications where the coolant is apt to be a moiten salt or metal. The chief reasons for not adapting this type hoiler were foxr the most part pcinted.out in the above reports. They were: (1) the need for high tube wall AT's in order to keep salt from freezing on tubes, (2) the lack of nozzles that would glve an adequate spray pattérn, (3) the need for very long tubes in order to ensure dry steam at high loads, (4) the need for a method to insulate the nozzle headérs from the heated tubes, and (5) the general feeling of the group that, although flash boilers éhow'great promise, much more developmental work is needed. | (2) Once-Through Boiler: Once-through boilers have been used in EuroPe for a number of_years and have recently come into their own_in this country with the installatidn of the supercritical units at Philo and Eddystone. They offer the advantage of boiling ang superheating in a continuous passage thereby eliminating the need for heavy steam drums. A once-through boiler also offers the advantages of rapid response to load changes, compactness, and ease of arrangement. The principle disadvantages are lack of water storage, the need for very high purity water, and, especially for nuclear reactor applications, the thermal stress problem. As was stated in Section 7.2 the stresses encountered when boiling water at 572°F with a salt at 1100°F introduces intolerable conditions. Therefore, it was necessary td separate the boiler and super- 93~ heater thus eliminating the chief advafitage of the once-thrgugh boiler, (3) Natural Circulation Boiler: The natural circulation boiler ils perhaps the conventional steam generator'for marine use. Unfortunately, here the problem of arrangement is encountered, It ig necessary to have a large drum at high elevation and a downcomer collector drum or drums, Also the problem of whether to put the molten salt in tubes or let it be on the shell side must be considered. Reports such as KAPL-1450, "Review of SIR Project Model Steam Generator Integrity”, seem to indi- cate that the best results for a liquid metal coolant such as sodium would be obtained by placing the coolant in the tubes with the water on the shell side. However, these experiments were all done uging stainless steel, In the steam generator proposed in this report Inconel ié to be used and with the obvious welght saving obtainable by filacing the high pressure steam-water mixture in the tubes it was concluded that the water-tube system was the more advaniageous, In order to find some compact method of arranging a natural-circulation water-tube boiler with a molten salt as a heat supplying wedium, a number of different configurations were considered. The most promising appeared t0 be a Lewis boiler which employs Fileld tubes. A Field tube is really 8 tube-within-a-tube. The inner .tube acts as an esgentially unheated downcomer. The bottom of the inner tube discharges into the sealed off end of the outer tube, The outer tube is heated and acts as thé riser. High recirculation rastios are'obtained with this type boiler. Also, since one end of the tube is free, there are little thermal expansion problenms, The Lewls type boiler bhas seversl other advantages but its chief dis- advantage would be the arrangement of a header sheet since it does have ~Qh_ the complication of a tube within-a-tube, The tubes must also be of the order of 12' to 15' and 1t was felt that poséibly some other arrangement would offer greater compactness. (4) Forced Circulation.Boiler: The forced circulation boiler was selected for the basic study because of its‘compactnéss and flexibility of arrangement, Use could be made of a nearly conventional steam drum, and ‘the tubes could be bent into a "U" shape to reduce the thermal expansion problem. The steam output could be controlled by the circulating vater pump., The forced circulation boiler is simple in design and principle and is well proven in marine applications. | 7T.3.2 Design of the Selected Steam Géfierator The collection of appropriate and adequate data for the steam generating sys%em proved to be a task df no small proportions. The molten salt was assumed to behave as a normal Newtonian fluid, Data is availsble from experiments performed at ORNL giving the heat transfer characteristics for heat exchangers, particularly delfia—array; This data was used for all salt side heat transfer coefficients (See Figure 7.6). The molten salt flow thréugh the steam generator is in the laminar region with Reynolds Numbers of 200 to 300. This is due to this particular salt's high viscosity in the temperature range to be used., The data for bolling water heat transfer characteristics was hardly as easy %o get. Wide variances are to be found in the literature for water boiling in tubes under pressure, After consulting with a group of industrial boile: designers it was decided to use a value of 6000 Btu/hr;ftg—oF for the heat transfer area calculations, A value of 2000 Btu/hr-f+°-OF vas assumed for scale deposits collectirg on the water side of the tubes. In the report, "Studies in Boiling Heat Transfer" «95 - (ucra Refiofit-Nb, C00-24), the conclusion was drawn that for water boiling in tubes in the'pressure-range from 1000 psi to 2500 psi, the difference between the tube fiall temperature and the water saturation temperature is independent of the heat flux. Accordihg to this data the value of b, - tsat that might be expected at 1250 psi was about 14.5°F, (See Appendix 7.1, Section 5A). ' In McAdams (Ref, 17, page 393) the equation, by = % 10959/A%l/h, indicates clearly that the boiling film temperature drop is P/90 heat flux and pressure dependent. These equations were used to determine sat = the temperature drop across the tube walls at points of maximum heat flux. The McAdams equatlon gave the lowest film drop with a value of 9.6°F, This in turn gave the waximum wall At of 85,20F at the maximum heat flux of 172,000 Btu/hr-rt2, It was decided to bring water into the steam generator at 5650F, seven degrees below the saturation temperature. This water would be a mixture of the recirculating water which is at the saturation temperature of 572°F and the fesdwater which is at 486°F. 'The water'éntrance velocity into | the tubes 1s 8 ft/sec., At the tube exit the dryness fraction is 0.11 which corresponds to a SBV of 65 percent., This is approximately the maximum steam by volume for this temperature and pressure that will still give good wetting of the tube walls (Ref. 21). A brief parameter study was undertaken to determine the most sultable tube size and tube pitch. It was concluded from this study that in balancing heat transferred against pumping power required, it should be possible to g0 to smaller tube size than usually used in oil fired boilers, Upon recommendation from ORNI pergonnel experienced wlth steam generators it vas decided that g 1/2"'i,do tube was the smallest suitable tube. The tube wall thickness compatible with the operating temperatures and pressures was calculated to be 1/16", and from salt pressure drop congiderations the closest tube pitech vas computed to be 3/4". Once the tube size was set (the recirculation ratio and steam flow rate are known), thé number of tubes necessérylto.carry the full power flow rate could be determined. The steam flow rate for the 125 megawatt design as calculated from a heat balance is 456,000 1b/hr. The number of tubes for ‘the steam generator is then 2336, The heat transfer area,fias’Qalcuiatéd_ih”two parts. The area necessary to raise the water temperature to the sa£firation point was calculated. The overall heat transfer coefficient in.the-watér heating region is 425 Btu/ hr-ft2~°F and the log mean temperature differencé 15.195°F; The heat trans- fer area for this region is 3050 fte. This_tdtal'area of 3854 £12 made necessary a tube length of 10.1 £+, | | It was decided that the tubes should bé bent into "U" shape for reduction of the thermal expansion problem, 'Calculations also showed that it would be bestlto split the 2336 tubes into 8 bfindles. This would keep the salt chketad vessels to a reasonable size and vall thickfiess,.and would reduce The header thickness and weight, | Of all the steam generatbr parts it was thought'that the headers would present the greatest problem: It was concluded ithat there was no reason why the tubes could not be run directly info the steam drum, The salt Jacket could be attached directly to the drum or by'expansion Jjoints. The tubes wouifi*%e "U" shaped and\"hfing" from the drum aé illustrated in Figure 7.1. A water header would be at the other end of the tube bundle. This header could be of éither flat head or dished head design. The hot (800°F) -97- salt would be introduced into the jJacket just under the drum and heat baffle would shield the drum from contact with the high temperature salt. The molten salt would leave the Jjacket just below the water inlet header, Four salt jackets containing tube bundles are attached to the bottom of each of two drums. The tubes serve the purpose of risers., They dis- charge their steam-water mixture into the drum where the steam is separated by mechanical separators and scrubbers. The water leaves the drum through the downcomers which are 1ocated‘on the bottom side of the drum along with the salt Jackets., The water in the downcomers is at the saturation tem- perature of 572°F. This water is blended with the 486°F feedwater and the resulting water temperature is 5650F. (The séturation pressure at this temperature is 1180 psia or approximately TO psi below the steam generator operating pressure), This water is forced back to the water inlet headers by the circulation pimp. The design capacity of the steam generator is 456,000 1b/hr. The water flow rate is 4,149,500 1b/hr and the salt temperature drop as it traverses the steanm generator is 76°F. At the 95,9 umw power load the full-power steam demand is 355,030 1b/hr and the salt temperature drop is 58,80F. At this o power the inlet the outlet tewperatures will be lowered to 761.80F.and 703 F. 7.4 The Superheater 1t was felt that the superheater design would be the most straight Torward of the steam generating system. The superheater is toc %ake the saturated steam at 572OF and heat it to 9500F. The 125 mw capacity of the superheater was to be 348,000 1b/hr but the capaclty necessary for 35,000 shp is 263,300 1b/hr. -98=- Again, investigation showed that tubes of the smallest practical diameter would give the best heat transfer characteristics, Tubes of 0.5" 0.D. and 0.4" I.D, were selected. A steam exit velocity of 100 £t/sec was chosen as the maximum practical velocity, To carry the flow at this velocity 722 tubes were necessary. It was decided to space these tubes at a pitch of 3/4". This gave a molten salt velocity of 11.55 ft/sec and a pressure drop of 1.33 psi/ft. The salt inlet temperature is 1150°F at 125 mev and 1138.3°F at 95.9 mev, The exit temperatures are 1126°F and 1120,4°F reépectively. The overéll heat transfer coefficient is 29 Btu/hr-£t°-F and the heat transfer area is 1070 P2, .This'gives-a tube length of 11.4% ft, The maximum heat: flux was calculated to be 231,000 Btu/hr-ft2 and the maximum tube wall At was 80°F. The superheater vessel ig "Uy" shaped with the headers at both ends, (See Figure 7 2) The tube bundle runs through the #essel with the tubes arranged in a delta—array. As is the case in the steam generator the headers were considered %o presefit the greatest actual problem. Numerous header arrangements can be deviged but the best seem %o be either a dished or flat head, 7.5 Auxiliary Equipment and Arrangement T7.5.1 The Steafi Drum and Desuperheater The steam drums are an integral part of the steam generator and contain the mechanical steam-water separators, the steam scrubbers, and the desuperheater tubes. It was decided to.use the two conventional drums of the class 931 destroyer boiler room with the attachment of the molten salt jackets to their undersides and the replacement of the conventional -99= risers with 5/8" o0.d. tubes as described in Section 7.3.1. The drum méterial is to be Inéonel but-the separators and scrubbers are to be of conventional design. The drum diameter is 52.2" and the bottom shell thickness is 4,8", The desuperheater will consist of tubes running through the saturated vater in the drum. It is necessary to desuperheat 5340 1b/hr of steam when running at 35,000 shp., Superheated steam at 950°F and 1235 psia will enter the drum in the tubes and be cooled to 650°F, The arrangement of the steam generating equipment around the reactor and within the secondary shield would be as shown in Figure 7.3. It is realized that the actual design of the steam and salt piping within the éecondary shield requires careful analysis, which 1s particularly necessary to keep stresses due to relative thexrmal expansion within reason. However, neither Time nor talent permitted such an analysis for this study, and therefore only a reasonable estimate would be made for the volume required for this pluwbing. As shown in Figure 7.3, there would be two identical galt and steam systems. It would be desirable to have all pump driver accessible from outside the secondary shield, It was therefore proposed that the secondary salt pumps be mounted ofi the top and drive through the secondary shield. Similarly, the water recirculation pumps could drive through the aft face of the secondary shield, T.5.2 Feedwater Heater and Other Components Although most of the equipment following the superheater in the steam generating system will remain unchanged, several components will no longgr be necessary when the two furnaces are replaced by & reactor and at least one new item must be added to the system. =101~ - to keep pump turbine weights down., If these pump turbines are assumed to be expanding the steam to the same extent as the main feed pfifip turbines, which also operate on superheated steam, the enthalpy contained in their exhaust would be more than sufficient to make up that lost by the exclusion | of the forced araft blovers and fuel ol pumps , In'ordérhté'maintain the deaerator saturation pressure at 18 psig at full load, a somewhat greater quantity of auxiliary turbine exhaust_must be bled back to the main condensers, This in turn wiil probably require the addition of several condenser tubes to maintain the former condenser .vacuum. Calculations show that at full power 12, 530 pounds per hour of auxiliary turbine exhaust must be bled to the condenser, if a deaerator pressure of 18 psig is to be maintained. 7.5.3 Feelwater Treatment Just how much feedwater treatment would be necessary to ansure long;term firouble-free service from the steam generators was one of the many problems that the group did not have time to investigate. The conventional destroyer supplies distilled wvater to the oil fired systeuw and,no attempt was made to answer the question of whether or not further treatment by ilon-exchangers would be necessary for the ;eactor heated steam generators. However, a heat transfer coefficient of 2000 Btu/hr~ft2—°F was included for scale. deposit;, which should be conservative. It should be pointed out that the replacement of the oil-fired furnace by_é ffised salt system makes the steam generation equipment relatively clean, and therefore the use of ilon exchangers to further reduce the water impurities way be Justified. Such a system should be a large lmprovement in cleanliness and require much less maintenance than the conventional oil-fired boiler, -102- 7.6 Part Load Operation The method of achieving a certain steam rate for loads which are some fraction of full power will depend upon the method by which the reactor is held at part load. Presumably,,thé'flow rate of the molten salt coolant and the sverage tempéiature of the reactor will remain the same, but the temperature rise of the molten salt coolant as it passes through the primary heat exchangers will vary according %o load.. Thus in the hot molten salt loop, which inecludes the superhéater, pfimary heat exchanger, one of the pumps and part of the-blending'abparatus;'the average " temperature and salt flow rate will remain cOnsfant and fhe inlet and outlet temperatures of the superheater will change apprppriatei& as will the amount of molten salt that.is bled off as the heat source for the cold or'steam generator loop. The'average'temperature of fihe steam.generat0r can be either railsed, lowered, or held the same according to the émoufit'of salt bled from the hot loop. It was fbund that in lowering the fiower frofi.las mw to 95.9 my the aversge temfiératureuof the éfeam génerator could be iowered 380 thus alleviating the thermal stress problem somewhat, ‘S 000°€9 - 1A 13M ‘183 "SE7 000°'CS - 1M ANG ‘1S3 L2681 -V3IYV HIISNVHL LV3IH UNCLASSIFIED ORNL=LR=-Dwg, =257L45 L1 -"a71 38nt .8/% -'0'0 380l 89i1 - §3an1 "ON HOLVHENIO WV3ILS Q3S0d0oyd "SHIAWOOINMOU 9 ' 3¥N91d -103- HOLVHINIO x«upmlxl\\ YOLVHINID ¥3d .wl\\fi 4d374N0 17VS Qo |0 , O NI H31VA NOILVINONMIDI Y -9 | 3744v8 1v3H HOLVHINGD ¥3d ¢ L37INI LTIVS ~ 1 o = ™ WNY¥Y Wv3als ns,2l 'S8 00€°1] = LM L13M 1S3 i u ‘'S87 0006 = LM AY¥Q 1S3 z1ld GES V3IYV WIJISNVHL LIV3IH ,004°0 =0’ 38nl ,00§°0 ='00 34anl I9€ — $38N1 40 ON P . 3 b6 | — _ § 1 u8l s3sni LITINI 1IVS ¢ . I S3IENL 13TLN0 LIVS | | D [ @& ol Q | mmmazqmpmllflhw Hu 0 i ln/h..._l. . S8 fl iz L AN og N A0 a5 Y O 'L 3J¥nold H31V3IHY3IdNS J3S0d0¥d A < o f GIFIHS ANVONOOIS 40 FA0YId MHINNI A GNY 8% ARYYd ONILSIX3 l}’ WY3FLE QILVIHNIANS waf— T T T WEZivIAmaIdns oL T WY3ILlS aaivIHE3disag ¥ [ e [ 8 -lm $ SHILYIH U3LVYMAIAd OL —e— : . 23 15 NI H3LVM Q334 — S i dWNd . 1 az_:.__._ofiomzl// ' _ Illllt.ll]cll‘ll!..!.".. + am4 4 MOT38 NMOHS NILSAS HWVILS S - - 3 dIHS JAOGY NAOHS WILBAS LIVS o + (] HOLOVIY a3 1HS X AN AHYWiHd \ i 3Ivos 0% MIIA NVd JYIHNTNE AONY LS . INYHAd SNILSIX3 13IHS ANVANOD3S NIHLIM . - JILYWIHOS ONIdid ANV LNIWIONVHNY oI1SVE Q3IS0dOo¥dd HOLVHANTAY nVv3Ils YILVYIHHILANS €1 34n9id -106- OENL=LR~Dwg ,=25748 - UNCLASS IFIED FIGURE 7-4 ESTIMATED SECONDARY SALT VISCOSITY | TEMPERATURE °F 700 T8O 800 880 900 950 1000 1100 1200 (300 1400 |800 i800 | R i i T 1T 500 oo\ N N i \ 0 o 100 a | - Z i o | >= }.... wn o O » > 20 N \ 350 400 500' 600 700 800 900 TEMPERATURE —°C 100 ORN L-LR-Dwg, =257L9 UNCLASSTF IED -107- FIGURE 7.5 FRICTION FACTOR CORRELATION EF. 25 TUBESIDE F. 25 SHELLSIOE - S N\ 1000 10,000 REYNOLDS NUMBER -108- ORNL~LB=Dwg,~25750 UNGLASS IFTED FIGURE 7.6 100 [ T T T T TT7 1 | - HEAT TRANSFER DATA CORRELATION Nu —Pr -4 : /| Nu _ 08 | 4 Pr.a 0.023 Nre ""\3’ % REF. MS ADAMS L7l | 1 /// Nu _ 0.91 Bra =0.006 Ngre \;/ 10 Nu 0.36 A "4 20.47 NRe - . Pr 4 B/ 6/' > *—REF J.L.. WANTLAND %\\ .0 . _ 100 1000 _ 10,000 REYNOLD'S NUMBER -109- FLUID HORSEPOWER - FHP FIGURE 7-7 EQUIVALENT FLUID HORSEPOWER,WEIGHT AND REACTOR RADIUS CHANGE FOR CONSTANT SPECIFIC WGT. 300 l ASSUMING : MNth=.25 STEAM PLANT EFFICIENCY - MGEN.=.95 / 250 }—————T MOTOR =.90 : N pumMpP =.80 b / / OVERALL PUMP weT, = 19103/ 1o y GENERATOR WGT = 101bs., o Y STEAM TURBINE PUMP = 51bs., , COMBINATION FH / / / 200 ' / 7 A*'/ / VERALL QQ\/ 0 | SPECIFIC WGT, <> / 150 LBS./SHD 100 d | / EQUIVALENT CORE RADIUS WEIGHT '/ BASED ON BASIC REAGTOR DIA, OF 50 | e L 72 INCHES & PRELIMINARY PRIMARY /fl SHIELD WEIGHTS ONLY, v, | 7, 0 0 4000 8000 12000 16,000 20,000 WEIGHT-LBS, 0 ' | ' 2 ' 3 . 4 REACTOR RADIUS CHANGE ~ AIN =110~ 8.0 REACTOR ANALYSIS 8.1 Nuclear Configuration Figure 3-2 schematically indicafies the physical picture of the core and reflector which will be described in detail below. General overall nuclesar conéepts of this high pérformance physically small system are modifications and combinations of advanced design ideas of ANP Technology under consideration at the Oak Ridge National Laboratory, Ref. 36, Physically, the core is a circulating, fused fluoride, uranium bearing salt flowing through a beryllium oxide moderating matrix and incbrporating an inelastic scattering metal reflectsf. Systeus of thesé-types feature high power density and relatively high operating temperatfires. Numerous nuclear advantages are manifested by these systems, The released energy is easily extracted from the core in that i£ is generated in and transferred out by the same fluid, Fluid fuel systems are, in general, self regulatifig under small perturbatiofis awvay from nominai operating conditions due to prompt vblume expansion within the fuel, Thirdly, . 'fiugh of the energy released in the fission process other than the kinetic energy of the fission fragments is retained and collected through cooling bmoderator, reflector and internal neutron and gamma ray shielding with the coolant-fuel, 'Other advantages are low Operating pressures and the relative ease of extfacfing volatile fission products, Disadvantages include, large fuel inventory required from excess fluid for component cdolihg, heat exchangers, pumps, core inlet and exit plena, A relatively serious hazard is present in this circulating fuel system, specifically, the containment of a corrosive, multicurie fiuid at high -11l- temperatfires. Requirements in maintaining the fused salt liquidous during shutdown also burden these system, Limitations were necessarily placed upon the nuclear designlto€meet the requirements_of metallurgy, heat transfer and nuclear design, and to narrow the breath of the étudy, With the choice of a fused fluoride fuel, no point in the system can be at a temperature less than its solidification value (in the order of 980°F) and no point should exceed & temperature ~ a% which rapid corrosion takeg place., All fluid surfaces should be Inconel clad to & thickness which will withstand 10,000 full power hours of operation. Power densities will be high, but not enough to induce. - dangerous thermal stresses in all materials. Resulting limitations regtricted the design to the following specifications: Mean Core Temperature 11500F to 12500F .Primary Fluid Surfaces 30 milslcladding Incone or greater : Maximum Power Density 1200 vatts/em> fuel Core Dimensions Riéht ¢ilrcular cylinder 70 ¢cm in diameter 80 cm in height 8.1.1 Moderator Matrix The moderating matrix consists of rodg comprised of beryllium oxide, three-quarters of an inch in dlameter, clad in 40 mils of Inconel, Radially, the rods will be close packed in a triangula?ly pitched array. The pltch is defined by the selected volume fraction of_moderator in core, Beryllium oxide "meat" extends the entire:length'of the core, and joining on each end of meat will be, one and a half inches of boron-10, BeO ceramic ~112« material to suppress fission in the core inlet and exit plena, Ends of the elements neck down fofming thé plena and joining the structural members. See Figure 3,2. 8.1.2 Reflector An inelastlc scattering reflector is utilized in the systenm. This choice has not been proven su?erior.to an elastic moderating material such as beryllium oxide, but it is believed to contribute definite advantages over BeO in this specific application, The choice of a nickel reflector is based_upon the fact that this material possesses excellent slowing down characteristics in the higher neutron energy range, which is fif considerable importance in this inter- mediate reactor. Secondly, relatively émall amounts of cooling will be required for the reflector and therefore it will retain its desirable nuclear properties to a large degree. Thié high atomic number material will attenuate core gamma rays very strongly and thus reduce the required gamma ray shieldifig. Also, fast nefitron leakage out of fhe reflector is within acceptable limits and is only of minor concern in fast neutron shielding, Time did not permit the detail investigation and comparison of systems incorporating elastic and inelastic reflectors; intuitive reasoning lead us to the nilckel reflector, Thé resulting reflector.is comprigsed of a 6 inch thick cylindrical shell 29,6 inches in inside diameter surrounding the core, Cooling annuli penetrate.the.nickel vertically through the reflector and coolant is supplied from the 16wer plenum,. Estimated coolant required will occupy 2 rercent of the reflector’s volume, | 8.1.3 Fuel Numerous types of fluoride salts are avallable,but in a large -113- ma jority, the existing data on their properties (corrosion, thermal and mechanical) ave limited. Therefore, it Wés necessary to make the basic criterion in the selection of the fuel depend upon the technology presently avallable. " The resulting selection contained ZrF, Nak and UFM° (See Section 4.1). Unfortunstely the nuclei constitubting this fuel lack some of the more desirable nuclear properties. Namely, it contains nuclei of high atomic number and thus has poor slowing down properties. Both zirconium and sodium have significant absorption cross sections in the intermediate energy range, although theilr thermal absorption cross sections are relatively small. Due to the large volume fraction of fuel in the core, any added neutron moderating material in the fuel will in general, reduce critical mass and the average energy of the neutron number density in the core, Although these changes will not be 1arge, they will be significant. Other possible cations which could replace zirconium or sodium are beryllium and lithium, Beryllium fluoride lacks corrosion compatibility with Inconel although it would contribute appreciably to the neutron moderation of the core., Lithium fiuworide also attacks Inconel and only isobopic lithium-7 could be considered due to high epithermal and thermal absorption cross section of elemental 1ithium, 8.2 Parametric Study An investigation into the nuclesr characteristics of the described core containing various ratios of beryllium oxide to fuel were deemed necessary in the selection of a feasiblé design. The principle objective of the study was to determine the critical U-235 concentration in the fused salt? and to minimize this value through varying moderator to fuel ratio. Secondly, the ~11h4- total fuel inventory was to be minimized through the selection of a critical fuel concentration and ffiel volume in the core, The range of investigation was limited between 0.4 to 0.6 volume percent fuel by the power density in the fuel at the lower limit and lack of sufficient neutron moderation in the core at the upper 1imit; o Group-diffusion methods were the means of analysis; specifically, a 3 group 3 region, one dimensional ORACLE diffusion code Ref. 60, Region allocation were to: (1) control rod thimble, 5 cm in radius, (2) core, cylindrical shell 32.5 cm thick and'ofitside radius at 37.5 cm and (3) the nickel reflector 15.2h cm thick., Thé cofistituents of the regions are as follows, measured in volume percent. Region 1; Ificonel 19%; Void 81%; Region 2: Inconel, BeO, and Fluoride Salt - variables Region 3; Nickel - 100%, 8.2.1 Cross Sections | | For the parametric study, the mean cére temperature was taken as 1200°F. This condition results in a mean'néutron energy of 0.0795 cu at thermal equilibrium with the core materials, assuming_no thermal spectrum hardening, Energy boundaries for the three groups were'éelected on the following basis; (1) Some existing data available for these_choéen boundaries, and (2) these boundaries were suggested by the spectral distribution of fission from multi- . group analyses of sifiilar reactors, Ref. 59. Table 8.2,1 Group Energy Range | Lethargy Range 1 10 Mev - 0.183 Mev 0 -4 2 | 0.183 Mev - 1.4k ev b - 15.75 3 | 1.y ev -0 - 15.75 - o -115= Cross sectlons for energy degradation from one group to the next lower group were defined by: i 1 | Op = Ogr, / AUL where 4 . = i i Oar jgaé * Clie These terus are defined as: 0} = average, transfer chss section from ith X 4 group to the (1 + 1)*" group cm gi = average elastic scattering cross section for group 1. e i g = 8average inelastic scatiering crogs section for the ith ie growp, cn? = averége slowing down cross section for the ith group. cn® _§° = mean log energy loss pé% eiéstic scattering event, . An spproximation in this method of Incorporating inelastic events in the slowing down cross section is the assumption that each inelastic event removes a neutron bne 1ethargy unit or the mean lethargy gain yper inelastic event is unity. | | Transport cross sections were evalusted in all groups as, O_rmi = O'e:L (1- K )+ 0‘}_& + O‘ai . where M =g%§r_ with the exception of Be0 in the thermal group where experimental datum was incorporated. Ref., 63. Chemical binding effects upon neutron scatiering were neglected and the assumption of free atom scattering was made throughout with the above noted exception. -116- Fast and intermediate group absorption cross sections were taken from various references. References 8 and 61. A large majority of these values result from analytical apptoximatiéns to the energy dependent cross section, Several are experimentally indicated values. Thermal absorptions cross sections were taken from Ref, Bh,corr¢Cted for temperature and averaged over the assumed Maxwellian spectrum, - - Bc/KT | | 512- F_a(kT) o (KD)l £/2 X ax o a (2200 M/sec) "~ Ec¢/KT -XI ‘I Xe © ax Ec is the upper thermal.group boundary f(KT) is the non 1/V c0rfed£ion. for Bc/KT = 18.1 | ‘Qzfigg%o W5y 0.50 £(KT) Group one and two fiséion Cross secfiions were averaged over each group from the values tabulated ifl Ref: 6l1. Group three fission gross section was taken as o £(XT) %15152) from Reference 3k, L+ct) A tabulation of all microscopic cross sections can be found in Appendix 8.1 along with their references. 8.2.2 Summary of Results Based upon the philosophy set forth as to the general concept of the over-all design study, the following criteria were utilized in the selection of a core design from the results of the parametric study, Of -117- ma jor importance is the power density within the fuel which must be main- tained below 1200 watts per cmd of fuel to insure the reliability and ‘integrity of the beryllium oxide moderator rods. Thermal stress induced in these rods through gamma ray and neutron energy deposition must be maintained within safe levels., Also, as stated previously, it is desired . to minimize the fuel concentration in the fused salt and to maximize the utilization of the uranium investment. In addltion, the reactor should be kept operabie with a maximum of thermal fissions, reducing both fuel investment and control problems. It was concluded on the bases of these data presented in Figures 8.1 and 8.2 and Table 8.22, case 2 (50.9 percent fluid volume in core) justify the above criteria most satisfactorily. Case 3, with the decreased fuel fraction, is eliminated automatically by its high“power density and wranium ¢oncentration in the fuel. Although case 1 (61 fiercent volume) indicates larger safety margins with respect to power density along with an impercepti- ble difference in fuel concentration compared with case 2, this system -exhibits 20 percent less thermal fissions. The twenty percent increase in thermai fissions with case 2, indicates . lower average energy of the neutron number dgnsity in its energy distribution, I% is believed this system will exhibit more control with an absorbing con- trol rod than the faster systenm, 8./2.3 Control Rod Study Reallzing the system under investigation would operate pre- dominately in the intermediate energy range, control of the system must be achieved through degradation in energy of fast and interwediate neutron to thermal energies and result as a loss to the system through absorption. -118. ORMLmLE=Dug ,=25752 UNCLASS IFIED FIGURE 8-I REACTIVITY vs MASS U?23® MEAN TEMP - I200°F 4 > ” / ~T / a 0 > _ 7 Yk / >~ / | / / -2 F—— CASE | 1 z h 7 = / :I) CASE 2/ u -4 / |, = / < — y / CASE 3 = / / & -6 ‘ / 0 / / a / . // / -10 / / 40 50 60 70 80 90 MASS U®**® IN CORE (KGM) ORNL~-LR=Dwg, =25753 UNCLASSIFIED -119- 0g’ 7304 WO/ . .N SWYYS St ot s¢ og NOILVYLINIONOD _,,N SA ALIAILOVIY ¢-8 JHN9IA ALIAILOVIY LIN3OH3d -120- TABLE 8,2.2 RESULTS OF PARAMETRIC STUDY ¥See Appendix 8.4 CASE NUMBER 1 2 3 Volume Fractions) Salt 0.6108 0.5090 0, 4072 ) BeO 0.3178 0.,4009 0.4840 In Core ) Inconel 0,0714 0.,0901. 0.1088 Mean Core Temperature 1200°F 12000F 1200°F Uranium-235 Mass (KgM) 90 72.5 70.0 Multiplication Constant (K) 1.00187 0.99317 1.00666 U-235 Concentrgtion in Fuel (gms/CM° Fuel) 0.k2khs 0.41030 0.49511 Core Fluid Volume (cm3) | 2,1204 x 105 L.767 x 105 1,513 x 107 Percent Fissions - Fasf _ 10.82 8.32 T.33 Percent Fisgions Intermediate 66.94 63.27 61.38 Percent Fissions Thermal 22,24 28.41 31.29 Average Powgr Density (Watts/cm” Fuel) 589.5 7071 884.3 Peak Power gepsity (Watts/Cm” Fuel) 831.2 990.4 1255,7 Prompt Temperatureo ' -5 -5 Coefficient §K/°F -2.63 x 10 2,19 x 10 -1.75 x 1077 ~ Prompt Neutron* -6 i ¢ Lifetime (Sec) 1.35 x 10 1.73 x 10 1.87 x 10" ’ 121~ This requires the control element to contain both moderating and absorbing materials, One centrally located rod was investigated and 5.4 percent reactivity control (see Table 8,.2.3) was obtained with the element con- taining the following materials: 19 percenfi by volume Inconel, Eh vercent BeQ, L0 percent nickel with 1 percent by wt Blo and 17 percent void for rod thimble clearance. It has been concluded that 5.4 percent control is adequate as a minimum value., (Ieave as a minimum, 3.7 percent shutdown marginc) There- fore, only one control element would be required., However, the above design was not considered adequate in that heat transfer across the clearance gap was insufficient, An alternate consideration immersed the rod in a bath of sodium, filling the clearance gap, Also, the rod thimble would at all times be filled with sodium and upon insertion, displaced sodium would be forced into a small reservolr located outside the pressure vessel. 6.1 percent control was achieved with this system and also 1 percent reactivity was added to the system with no rod penetration. See Table 8.2.3 below. TABLE 8.2.3 Void Filled Control Rod Thimble k = 0.98853 Sodium Filled Control Rod Thimble k = 0.99585 Rod at Maximum Penetration with Sodium in Gap k = 0,93786 The specified boron content was only a means of analysis, Due to the damaging metallurgical instabilities resulting from high irradiation exposures of boron in metal, it is recommended that the B-10 equivalence of a Europium Oxide dispersion in nickel be used as control rod material, -122- This oxide also exhibits more reliable control during operating life in that very large exposures arebrequired to obtain substantial burnup, hence only & small loss in control will be obgerved upon long exposures. The following date (Ref. 68) indicates thig phenomena ; 14000 e hr fi 152 13Y . H%ggfigb-E“ Bul53 fi20bb'EU15E,,£§¥%’)V 155 1E00b"‘ fu my'%5_IPE 5156 1sab 13,0000 & Natural Isotope Event Cross Section Per - Half Life Abundance ‘ | Event (at 2200 M/s) W77 R (n,y )Euto? 7200 b Stable 152 - | (n #)Eu 1400 b Stable Eul?? (h, y) 5000 b 13Y (p) Eulo? 8 9.3 hr 52.,23% . pylS3 (n, ¥) 420 b Stable Byt (n #) 1400 b 16Y () N 2 13,000 b 1.7 (g) Eyl56 B | 15 4 18.8% p'0 (n, «) 4020 . Stable In a high neutron field it takes 3.3 neutrons to be absorbed, on the average, in an Europium atom before it is lost to the system; only one is required in B-10. -123- 8.3 Nuclear Design Tabulated below are the resultant design conditions of the core, Table 8.3.1 Power 125 MW Core Volume 3.471 x lO5 cm3 Filler Volume Fraction o 0.5090 Be0 Volume Fraction - 0.4009 Inconel Volume Fraciion 0.0901 Mean Fuel Tempersture 1225°F Hot Clean K 1.0275 Critical Mass | T1.75 Kgu U235 in Core Excess Mass for 0. 4 2.75 percent 14,00 Kgm @35 in Core Startup U-235 Capcentration 0,48528 gms U-235/cm3 fuel Startup U-235 Inventory 605 Kgm U-235 Percent Fast Fissions - | 8.29 Percent Intermediate Fissions 63,87 Percent Thermal Fissions | 27.84 Prompt Temperature Coefficient -2.19 x 10 X/°F Prompt Neutron Lifetime 1.92 x 10'6 Sec Average Flux Over Core Fast 1.33 x 10%0 neutrons/sec cmd Intermediate 8.14 x 10lh " " Thermal 1.97 x 1083w " -12k- Average Power Density 708 Watts/cm3 Fuel Maximum Power Densify | 1040 Watts/cmS Fuel Total Control Rod Reactivity Worth 6.18 percent SK/K (_SM/M) ("8K/X) Core 7.1 Endurance | | 4000 Full Power Hours 0.5 percent Burnup ) K/K) System 50 8.3.1 Criticality Critical mass, and uranium concéfitration in fuel were obtained through a series of problems performéd ofi the ORACLE simular to thoge described in the fiarametric gtudy, Due to heat transfer considerations, the mean core temperature was increased to 1225°F, Results iqdicatéd a critical mass of T1.75 kegm U-235 under hdt, clean and unshielded con- - ditions, graphical results are presented in Figure 8.3, Radial flux and power spatial distributiofis are presented in Figures 8.4 and 8.5 respectively. Resulting flux distribution indicates the reflector savings in the thermal and intermediate groups through the gradient of the distribution near the reflector boundary, 8.3.2 Self Shielding Disadvéntage factorg were obtained by diffusion theory methods for the unit cell as defined in Figure 8,6. Both intermediste ang thermal group factors were considered, although the intermediate factors were insignificant, Thege effects were incorporated by defining effective crosg sections and expressing the effect as a&n excess reactivity to the unshielded Criticality calculations, A simple perturbation method was used to obtain ORNL~LR-Dwg. =2575) UNCLASSIFIED -125- (wbx) IYOO NI __.N-SSVW 002~ _ 00°1- \ T m 2 \ . o m 000 5 2 m D \ o ool = - - = 002 106070 - TANOONI 600%'0 - 028 060S°0 - 17VS L ¢WND Ol x SILP'E-TNNTOA 3HOD 00¢ 4. 6221 - dW3L NV3IW SNOILOVYd 3JANTOA 3HOD SSVIN 21 SA ALIAILOV3Y | ¢£-8 JYNOId RADIAL FLUX DISTRIBUTION RELATIVE TO AVERAGE CORE VALUE -126- ORNE~LR=Dwg, =25755 UNCLASSIFIED N | & \:»THERMAL AVERAGE CORE VALUES FAST ----1.33%10'® mrv ) INTERMEDIATE INTERMEDIATE -~ 8.14 X [0} v 1.4__...-’ ‘5_'_\ \\ 1.2 \*\\ ' N \\ N\ A\ 1.0 ‘\ 2 o w \ 0.8 & | QO \ REFLECTOR] & W N 063 . X i \\ - \ = \\ S \ 0.4 ° \ \\\ | CORE ___\ \\ \ 1\ 0.2) ¥ \\\\ : _ \ O 5 10 5 20 25 30 35 40 45 50 55 chA CORE RADIUS FIGURE 8-4 RADIAL FLUX DISTRIBUTION -127- ORN L~LR~ Dug, - 25756 ///////////// /// :: I W < 3 / S = /////////////// O @™ 5 / @ - ) = N ln . Q0 o / O Wl o © : ol o u_]m M W hao | 3 = o W | - > _ o E 80 < [ / gfl: ui | O w 2 <« g - < W oo — 1 > L o w < @ @ o O 15 / / // 7 /f/////f// O ALISNHG HBMOd -128- ORNL~IR=Dwg, »25757 UNCLASSIFIED CELL FLUX RELATIVE TO AVERAGE FIGURE 8-6 THERMAL FLUX DISTRIBUTION IN UNIT LATTICE CELL REGION AVERAGES BeO - 1.100 INCONEL~0.947 SALT - 0.833 1.4 .2 ] .\\ .0 \-\ \\ \ 0.8 ~ i) : r X 0.6 a = (1p] = J > W w 0.4 _J < 0 - O o > O = (4 < Ly T ° 0.2 @ 0 0 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 UNIT CELL RADIUS-CM -129- these_effects. The development of this perturbation technique is presented in Appendix 8,2.. 8.3.3 Burnup and Fission Product Poisons One advantageous feature of the large fuel inventory required 'for.this system is the extended endurance of fuel life. The large fuel volume is circulated continuously thropgh the active core and burnup is achieved homogeneously throughout the fuel, thus extending life by approximately a factor of six over a stagnant fluid systeu. Endurance in the order of 4000 full power hours is expected as the life of the initial core loading. This represents felatively swall burnup (0.5%) periodic additions of fuel are possible and would greatly enhance the l1ife- time. It is believed the system will operate for the 10,000 full pover hours of reactor life, with only minor additions of fuel %o the initial loading. In this analysis, non-volatile fission products were approximated as equivalent to lOObbarns of added absorption in the thermal group and 10 barns added absofption in the intermediate group per fisslon event, It is believed the above approximation results in an over approximation of the non-volatile fission product poiéons° Burnup and fission product poison effects upon reactivity were treated Jointly as reviewed in Appendix 8;3° Results are presented in Figure 8.7. 8.3.4 Prompt Temperature Coefficient Value of the prompt temperature coefficient as quoted (-2.19 x 107 §K/°F) contains the effect of the volume fuel expansion and the shift in the agsumed thermal Maxwellian spectral distribution with temperature, Effects of doppler broadening in resonance absorptions were | dNNYNE LN3IO¥3d G2 02 Sl o'l c'0 0 R <8 _ _ I _ 0 Ky 00002 000°s] oon_..o_ 0006 0 mm S¥NOH ¥3MOod 11Nnd / o'l i = = \ 5 \ oz Ve - m yd 2 o 0O€ m < -f mw p. _ Ob ® e m \ 3 /] 06 < \\\ 1 \\\\ | - \\ " \\\\\ _ . 0L —lo's SNOSIOd LO2Naodd NOISSId ANV dNNdNg oL 3Ina 3IWI1L3dITT OSNINYNA SSO7 ALIAILOV3Y LN3IOY¥3dd L-8 34¥N9ld -131- neglected but are expected to be negative also, Ref, 71, Method of calculating this coefficient.involved the investigation of the multiplication constant of identical systems at two temperatures using the 3G3R ORACLE code. Table 8.3.4 presents results. TABLE 8,3.4 Loading Temp Reactivity Kgm Op 72.5 . 1200°F 0.2550 72,5 1225°F 0.2003 8.3.5 Xenon Poison As discussed previously (Sec. 4 ), the removal of volatile flssion products can be achieved with relatively high efficiency in a high temperature, liquid fuel system. Provisions for periodic removal of the volatile matter are incorporated in the system desigg. of s most concern in estimating xenon polsoning is the efficiency of iodine and xenon removal which in = lafge extent is dependent upon their solubility in the fuel at these temperatures. If there were no removal of these elements, the steady siate poisoning is valued at -0,297 percent reactivity for an average thermal flux of 1.97 x 1013 neutrons per cme—sec, Assuming a high degree of removal, the poisoning is approximateiy as 10% of the steady state value with no removai. Poisoning worth of xenon is evaluated as 0,03 per- cent in reactivity. 8.3.6 Delay Neutron Loss Circulating fuel reactors suffer from a loss of delay neutrons in that a fraction of the delay neutron precursors undergo neutron emission -132- in the external fuel circuit. Loop time for the circulating is 1.127 seconds of which 0,287 seéonds are spent in the active core., For small loop times compared to the delay times of the neutrons, a valid approximation to the required excess reactivity required to compensate this loss is given by: (Eef° 69) * where o I Fraction of delay neutrons emitted per neutron emitted from the fisgion event. T, = Trangient time the fluid spends outside the active core. T, = Translent of complete fluid loop Resultant worth in percent reactivity has been evaluated as -0.56 §X/k. 8.3.7 Excess Reactivity Required Tabulated below in Table 8.3.7 are the excess reactivities estimated as required to waintain criticality for 4000 full power hours. Table 8.3.7 Condition SK/K Self Shielding 0.0051 Burnup and Fission 0.0158 Product Poison Xenon Polson G.0003 Delay Neutron Loss 0.0056 Temperature ( +32° AF) 0.0007 Potal 0,0275 ~133~ 14.0 kgm U-235 addition to the critical mass (71.75 kgms) are required to produce this excess, Total fuel inventory for the initial startup (k =1.0275) is 605 kem of uranium-235. 8.3.8 Control Requirements ~ Under nominal 0pération during initial startup (no burnup) the control rod must suppress 1.65% in reactivity (burnup, fission pro- ducts and temperature excess). This requirement places the control penetration into the core as 30 cm. Control curve in Figure 8.8 indicates core reactiv;ty s a function of rod penetration. In obtaining this curve, we have assumed the axiai.flux distribution as cosine in nature and the positional worth as a function of the flux-squared. JH0O N} NOILVHLI3INId dOYy 0¢S Ov o¢ 0z BN ORNL~LR=Digr 4 =25759 UNCLASSIFIED -13k- NOILIONOD NV3TD 10H NOILISOd Q0¥ TTOHLNOD sA >t>_.5*< ‘}é}n%-sgc O o N i io i 10 e o s 80 {20 Kapius -CM Core | Ecru..moL B.o-’ ch'r Exc.l-( Ilusut l e e '} SUPPO “THeRmAL sunr oy P- - ] o | -140- ORNL~LR=Diyz, =25761 1t 10 - H P ME Two GroupP lom NeutTroN FLux PLoT IN PrIMARY SHIELD TANK Fia. 9-1b IC;’ \F 1y L~ < $ FLux H/cm’- sec L~ / . 1o \\\ \ N\ N S N\ Zo: B0 WO 180 1eo ITO LEAD _ RoroTED WATER P -kt T des T e 1 Ao Tr . nry, .o 50 R wdte B Lo G f T 8o {20 200 2i0 220 280 240 e ,D,,l&TANCr_ Feowm % Core- Cm -141- A fast and thermal neutron flux. dlstribution was estimated through the north head into the primary shield plug based on the behavior of the ORACIE calculated horizontal flux through the heat exchanger. This vas used to determine thickness of gshielding needed above the reactor. 9.3 §§condary Salt Sodium Activation With the neutron flux situation estimated in the heat exchanger the secondary salt sodium activation was determined as the gamma source strength in designing the secondary shield. The average flux in the neat exchanger was calculated by numerically integréting the three group heat excfianger flux {as shown in Figure 9-1A) in a radial direction and then dividing by the heat exchangér thickness., No credit was taken for flux distribution in an axial direction as thils would have been at best a rough eétifiate. Neglecting axial flux distribution was the conservative approach. Sodium cross sectional data was taken from Reference 61 and was con- verted into three group averages as done in SBection 8.0, This was checked by independently three group averaging some earlier unpublished Curtiss=- Wright multigroup cross sections. Both averages accounted for the sodium regonance peak at 300 kev. Knowing the atomic density of'sédium in the secondary salt, and the three group averaged cross sectiofls and flux, the activation vas determined by their product: N = atomic density of sodium in secondary salt it (Wt % Na in salt) p_,. X .6023 x 1024 23 = 9.46 x 10° -142. A = activations/cc-sec :No‘a$ g;;;gi § | -GE(Barné) A Percent 1 3.52 x 102 2.1 x 107" 7.0 x 200 1.66 2 1,70 x 1012 2.1 x 1072 337.5. % 10° 81,14 3 - 3.0 x 10°° - 2,53.x 107t | 71.6 x 100 17.20 | | - b16.1 x 100 This gave a value of h300 curies total activation when multiplied by the volume of salt in the heat exchanger (13.5 cu £+) and divided by 3.7 x 1070, 9.4 Dofie Tolerance Levels The basis for the HPMR is a maximum dose rate of 300 mrem/week and 20 hour a week access time to spaces immediately adjacent to the reactor compartment (guxiliary engine room aft and abofié and thé storage compartment in old fuel oil deep tank forward of the reactor compartment), Ten percent of this is maximum allowed fast fieutron dose, Reduced to terms of mrem per hour, this is a total of 15 mrem/hr with not more than 1.5 mrem/hr in neutron dose, This set a fast neutron flux limit of 15 neutrofié per cm2 per sec taking the predominant neutron energy at 0.5 mev became the fast neutron source'is mainly from delayed fieutrons born in the heat exchanger. A flux of 10 n/cme-sec at 0.5 mev is taken as giving one mrem/hr (AEC Standards For Protection Against Radiation, Part 20 of Title 10 of the code of federal regulations, February 28, 1957). -143- 9.5 General Shield Arrangement At this point the steam genersting equipment sizes had been firmed up to the extent that a reactor compartment arrangement could be worked out. This was done with compactness and minimum shielding weight as the criteria with maximum use of the available fuel oil for shielding purposes. The arrangement chosen placed the'reactorlwith primary shield tank forward and steam generating équipment aft. This layout allowed a smaller primary shield tank, as the steam generators helped shadow shield the gamma and neutron leakage from the primary shield tank. Fuel oil attenuated this leakage out the forward, port, and starboard sides of the shield tank. The fast neutron dose determined the thickness of hydrogenous material reguired to attenuate to tolerah¢e dose 1evel (Section 9.4), Approximations of the gamma dose with simplified geometries and source energies determined the predominant rad;ations and gave estimates of lead thicknesses., With egtimates of secondary shiéld thicknesses the general shielding arrangement shown in Figfire 9=3 was laid out with a judgement estimate of the best pPro~ portion of shield material in primary shield to shield material in secondary shield, In light of a last minute alteration (Reference 73) in the shape of the fast neutron atitenuation curve in Watér {this change is incorporateé in Figure 9-1B) a foot of polyethyene should be packed around the after side of the shield tank at locations not shadow shielded by steam generating equipment as shown in Figure 9-3, This will have %o be fitted around existing piping as it was not allowed for in the original arrangement, NOILVWI1S3 LHO13M ¥Od aqa3Isn I = -t T HOL3MS OJILVA3HOS G713IHS AYVANOD3S INITAHLIAN0d S3HON!I O Ol ¢+ NOYdd G3¥3advl # ¢-6 34N9Id 0 GL'0 | #Ov r 0 SL'O | L0 i o ! ot H ) S0 g 9 0 51 S 4 0 Gl SS9 3 o2l ol G2 a 02l 01 TR 9 0-2l ol 09 g 08| 01 S99 v INITAHLIAI0d | NOMHI | av31 ] LNIOd SIHONI *SSINNDIHL ONIQ3LHS -145- 9.6 Primary Shield The primary shield deslgned for adequate fast neutron attenuation and with estimated gamma attenuation was then checked in more detail for adequate gamma attenuation, All gamma sources listed in Table 9-1 were considered . with the enmergy distribution indicated in the table and with & simplified source shape most closely approximating the actual source geometry. Form- ulations as given in Rockwell's shield design manual (Ref. 33) for lines, disks, cylinders and truncated cones with uniform and exponential source distributions were used, All dose vaiues below .0005 mr/hr outside the secondary shleld were neglected. The gamma dose from fission products in the heat exchanger, prompt gammas in core and heat exchanger, and water- lead capture gammas in the primary shield tank were éonsidered in more detail as described below. The fission products constitute.an important radiation source as they are rapidly circulated with a reactor cycle time of 1-2 seconds. This invalidates nuclear data on gamma energies and decay times, Therefore, the energy group breakdown presented in Reference Th was used which takes into acéount k.9 of the roughly 5.9 mev total available. This difference is con- sidered to decay before the fuel leaves the core. Saturation of long lived fisgion products is assumed which is conservative in this case. The pre- dominant enexrgy was found to be 3.2 mev for HPMR shield thicknesses. "The prompt fission gamma dose was calculated by an energy integration under the continuous fission spectrum from .1 to T.46 mev, A mean value for the HPMR shield was found to be 2.85 mev by running a series of energies assunming all proupt gammas at that energy. ~146- The energy situation for the lead and water is firm at 7.0 and 2.23 mev, However the source geometry becomes an important factor, especially in the case of water where the souice distribution (thermal flux) varies very rapidly and cannot be completely fitted with a simple sum of exponentials, Both radiations together contribute 80% to that dose outside the secondary shield which comes from gaumma radiatiOfi ieakage from the primafy shield tank. The geometry was handled by numerically integrating the dose con; tributions from unit line sources into contributions from unit cylindrical surfaces in the primary shield tank., These cylindrical surfaces of different radii wére then numerically integrated into the total dose contribution from the lead and water volumes. This method essentially gives an exact geometrical representation to within the accuracy of Simpson's rule for numerical integration. | Three directions from the reactor vessel to the outside of the secondary shield were considered. One horizontal shot out through the primary shield tank and secondary shield to the aft face of the after reactor compartment bulkhead, and a vertical computation through the north head, shield plug and top hat were done in some detail, Another horizonial calculation for- ward through the fuel oil shield tank was done for lead capture gammas in detail with estimates for water capture and heat exchanger fission product gammas, The resulis are tabulated in Table 9.1. Q9.7 Secondary Shield The activation of the sodium in the secondary salt required that a secondary shield be placed around the steam generating equipment., The overall dimension of this shield were established by the estimeted volume 147 requirements of the reactor and primary shield, the steam generators, and the superheaters., A plan view of the arrangement of this equipment within the secondary shield is shown in Figure 7.3, The resuliing shieid 1s box- gshaped with internal dimensions of 23' x 24! x 15' high, (Figure 9.2}, The thicknesses of shielding required were then calculated for a maximum dose of 15 milliroentgen per hour on the outer surface of the top and aft . faces of the shield., It was assumed that fuel on water would be used to aid in the attenuation of radiations from the forward and side faces of the shield, as described earlier, Except for direcily over the reactor, the amount of secondary shielding required was determined meinly by the secondary salt activity. The primary shield is relatively 5igh1y effective in shielding reactor sources, The total activity of 1300 curies introduced into the salt in the primary heat exchanger was assumed to be distributed in the ste§m generating equipment in proportion to the ratio of the volume of salt contained in any particular component to the total salt in the system, The individual volumeiric gource strengths were then obfained by dividing the curies of activity of the salt in a component by the volume of that component. Thus % of activity, ~activity location total salt curies decays/cmgsec superheaters 20,7 890 2.0l x 107 steam generators 36.0 1550 l.58lx 107 salt lines 36.0 1550 6.33 x 10 primary H X 7.3 310 (not contrivuting) ~148- ...... This assumption was recommended by ORNL personnel working with similar systems, and appears Jtsfified in view of the relatively long half-1ife of godium (15 hr) compared to the secondary salt cycle time (10 sec)., It was further assumeq thet the U-tube geometries of the superheaters and steam generator could be replaced by a straight_cylindrical gources of equivalent volume. The self-attenuation of these sources was determined by homogenizing the salt and tube bundles within each cylinder, and by computing mass attenuation coefficlents for the sodium gamwme ray decay energies of 1,38 and 2.76 mev. The approximation of:replacing the cylindrical sources by equivalent line sources was used, and Peeble's correction was applied to calculations involving slant penetration through the shield. Using these aspsuuptions, the thickness of shielding required wes calculated for eight points in the secondary shield and estimated for three more. It was attempted to select points which would give an indication of the shielding required for the areas receiving bath the largest and the smallest irradiations., Time dld not permit a Qbre extensive situdy. The - "hottest" points on the inner surface of the secondary shield were found to be (1) on the aft face of the shleld near the primary heat exchanger, (2) on the top face of the shield over the secondary salt pumps, where salt lines are near the surface, afid (3) directly over the reactor. The most lightly irradiated point appeared to be in the middle of the front face. Polyethylene was added to the aft face of the shield to attenuate fast neutron leakage wfiich could stream aft between the steam generators. In estimating the lead thicknesses required, it was first assumed that steel structure would be necessary to support the lead in the following 6%1 0 -b2 - 6 - b1 > _ 1l » | fi wC =l - fi 1 n €-6 3¥N9Id o _ . WOLL0OE 318Nn0Q ONIGIIHS = % ANVANOD3S any AHVYWINd P ONIMOHS " NOILO3S INITHILNID 7 7] 39VdS LNINdIND3 “ ONILVHINIO NV3ILS ININWLYEYEWOD \ ) T ORILYHIAS e WA HOS JONVIVE WVALS (3101034d { NopvaaNz ¢ : Lvshiono? WYl - - o 0L -~ "913 _. “ ky - in I _.,wxmmn; ot i x - Wi - : _ Hu/sgat o ® GNILY KAO 1 1 = - WY¥3LE ON[LNIWZNY o SHOLANPT MI¥ m| # HHS -ane.hl—fi “ “ ; ABYIIXDY T e la = I 0o HNILYHALD | “ HH/AT DIX ST =f o SHOLOAFY WY : 1 WH/EAT £8D . VA W = ¥ nHH ¥ wn zr Rn miMl | 3 ¥33NLE10 1 _ “ MO4 GANDISIQ MM G az% o . i 3 i EIRE LY DNILYM340 HOLOVIK «0Zik ) ! MH/EBT O 4= 0¥8 i, o] | Ly i e g | e D e ey 004t M3LY2HHZINE 34 (56T % | 440 N¥IT ANV gy 1y oNILYHIdO HILYIM ¥ L wwsen L AYNHALND * . uzoiuuzflw.wuuam L i g bk - - — - = . uzens L nLe 921 3= 9E9 DISd owil L1907 9% I.h.th._luo.._lxliov» —— ' 515 OF1 m... , v ="Ell { . POLSE COMI-YTI-TINNO an/nla 0iwi 4s 056 0154 0OT! —-— sa1 09L'M3 HH/5A7 Ol8Z . e — oiRd 0901 -156~ by superheated stean. The superheated steam requirements for the reactor system are summarized in Table 10.,1. TABLE 10.1 SUPERHEATED STEAM REQUIREMENTS T. Turbine and Turbo-generators = = = = = = = = = = = = = = = = 218,760 1b/hr II, Pumps: (1) Reactor Fuel - - - 150 P HP (2) Molten Salt = - = = = ~ = =~ 550 P HP (3) Recirculating Water =~ - - - 35 P HP Total - = - - = 735 P HP Steam Required 1.25.x oshs (BTU/hr) /hp x 735 hp 27,500 1b/hr (170 - 1328) BTU/1b TII. Feed Water Pumps = = = = = = = = = = = = = = = = = = === 11,700 1b/hr Total Superheated Steam - - = = = = = = - = = 257,960 1b/hr | Desuperheated steam is required in the galley, the air ejectors, feed booster puwmps, lube oil pumps and condensate pumps. Desuperheating is achieved in the steam drums by cooling superheated steam in tubes that pass through the saturated water in the drums. The steam is cooled from 950°F, 1200 psig to 625°F, 1165 psig. The desuperheated steam requirements are summarized in Table 10.2, Table 10.2 DESUPERHEATED STEAM REQUIREMENTS e Air ejectors, galley, leaks, efc, = = = = = = = - = === === 3,371 1b/hr Feed booster PUUP = = = = = =~ = = = = = = = = = = = = = = = == 895 Lube Oll pumps = = = = = = = = = = = = = = = = = = = = = === 300 Condensate pumps = = = =~ = = = = = = = = = = = = == - ===° 75 Potal = = = = = = = = = = = = 5,341 1b/hr ~157~- 10.3 Condensate and Exhaust Heat The deaerating feéd tank collects exhaust from some of the suxiliary equipment and it also receives the condensate. From the DAFT is drawn the feedwater which supplies the steam generating system. In order for the deaeration to be compiete, the pressure in the DAFT should not exceed 18 psig. The enthélpy of the saturated liquid at this ?ressure is 225 BTU/1Db, This is the enthalpy sssumed for the feedwater entering the steam generating gysten, The DAFT is unable to handle the exhausts at full power steam flow so it is necessary to run a portion of the exhaust directly to the condenser. This excess exhaust 1s 12,530 1b/hr at full power. This is slightly higher than the oil-fired systems 11,570 1b/hr; therefore, a small increase in condenser capacity way be necessary to handle this additional flow. A summary of the efihaust andlconéensate flows and their respective enthalples as they enter the DAFT ig given In Table 10.3. HEAT ENTERING THE DAFT From: - w(ib/nr) n({BTU/1b) Feed and Circulation Pumps - = - - - - 36,900 at 1,328 Feed Booster Pumps - - = = = = = = = 895 at 1,213 Lube O1l Pumps = = = = = = = = = = ~ 3b0 at 1,253 Condensate Pumps = = = = = = = = - - 775 at 1,218 Fresh Water Drain PUmps = = = = = = = 1,975 at 168 Condenser - - ; ----------- 220,1i3 at 102 Distillers « = = = = = = = = « - = - 2,300 at 148 HEAT LEAVING THE DAFT Boiler Feedwater - - = = = = = - - - 263,258 1v/hr at 225 BTU/1b ~-158- 10.4 Heat Addition in the Steanm Generafing System | | - The steam generating system must add sufficient heat to bring 257,960 1b/hr of water at an enthalpy of 225 BTU/lb up to steam at an enthalpy of 1470 BTU/1b plus 5,340 1b/hr of water at 225 BIU/1b to steam at 1263 BTU/1v. This total heat addition is 3.272 x 108 BTU/hr or 95,9 megawatts, In the reactor system a feedwater.héater is needed to do ‘the job that an economizer does in an oil fired system. Feedwater from the DAFT ai 18 psig is raised to a pressure of T00 psig by conventional boiler feed pumps and fed intolthe feedwater heater. Here, satuiatad sbeam at 1250 psia is mixed with the feedwater to produce yater at 486°F, It takes 91,730 1b/hr of saturated steam to achleve this. The 486°F water is now pumped to a preggure of 1500 psia and let down by thrbttling t0 the boiler pressure of 1250 psi. The feedwater heater forms an integral part of the steam generating system and with the saturated steam used for the heating forms a closed loop within the sysfem. As has been stated in previous sections, the heat addition to the steam generating system is by means of a molten salt. This salt drops 17.9°F in temperature in the superheater and 58.80F in the steam generator at a flow rate of T.49 x 106 1b/hr. These temperature drops and flow rates represent an input of 3.272 x 10° BTU/hr. 10.5 Comparison of Efficiencles No attempt to compare thermal cycle'efficiencies will be made here but only a simple calculation of the gross power-plant heat rate, For the reactor system: _]_59_ gross power plant heat rate For the conventional oil fired system: gross power plant heat rate Heat Input Shaft Horsepower 8 3,272 x 10 BTU/hr 35,000 shaft horsepower 9,34%0 BTU/shp-hr 3,108 x 108 BTU/hr 35,000 shaft horsepower 8,890 BTU/shp-hr The reactor system does require more heat input because the additional pumping power required for the molten salt and recirculating water is greater than the power required for the fuel oil pumps and forced drafi blowers. ™ ~160- 11.0 OVERALL POWER PIANT PARTICULARS 11,1 Introduction In order to determime the overall feasibility of a fused salt reactor installed in a particular class ship, it is necessary to consider all of the components in fihe complete system. A preliminary piping layout and drawings of the steam gen;rating équipment are included in Section 7.0, Rough sketches of the'primary:and secondary shields are also presented in' the shielding section (9.0). To completely determine the suitability of the resulting power plant, 1t is then'negessary to investigate the installation as to its effect on the ship's overall construction, balance, etc. In addition, it ié also necessary 1o consider operating problems such as control, emergency operation, and malntenance. 11,2 General Arrangement A brief study of the possible layout of steamn generating components within the reactor compartment and the location of the reactor comwpartment in the ship was made with minimum shield weight as the major consideration. No detailed optimization was attempted but rather judgement was used as to the relative sizes and -location of the primary and secondary shield, Giveh the decision of only replacing one oil fired plant with nuclear pover, arrangements were worked up using fuel oll as part of the shielding. As pointed out in the shielding section, fast neutrons are the primary radiation problem in this system. A hydrogenous liquid like fuel oll takes the place 0f polyethylene and serves double duty as fuel for the oil fired plant. -161- Arrangement One On first look, the best location for the reactor compartment seems to be the aft fire room where accessibilitj for removal of reactor com- ponents is done through the upper deck, However, preliminary estimates of steam generating equipment sizes indicated a larger reactor compartment than shown in the final design was required. To prevent propeller shaft penetration of the aft reactor compartment, the compartment would have had to move off centerline to such a degree that battle damage stability problems would arise, Therefore, arrangement one (see Figure 11-1) was worked out with the reactor compartment in the forward fire room, Provision for removal of the primary shield tank plug and reactor vesgel could be worked out through a side port, as there is twelve feet of clear height between the top of the secondary shield and the main deck. If removal through the main deck is dictated, the bridge super- structure would have to be removed. Tn locating the exact position of the reactor compartment, use of existing bulkheads and deep web frames were made, The forward boundary of the compartment is existing bulkhead 63 and the after boundary is deep web frame T5., Tying into existing main structural members winimizes additional support structure in the nuclear power conversion, A weight and moment study was made on arrangement one. The weights and moments of the boiler plant were replaced with the reactor system jincluding fuel oil shielding tanks, Fuel oll was distributed in the exist- ing after fuel oll and ballast tanks to balance the forward mowent pro- duced by the increased weight of reactor compartment over the boller com- -162- ponents. Results of this study showed that if Just the deep tanks aft of the after engine room were used, a resultant trim of 1.94' by the stern 1s produced as compared with a 1.63' trim by the stern for a completely conventional oil fired destroyer. This gives a total of 391 tons of fuel oil compared with 728.5/2 = 364,3 tons per one oil fired plant. As fuel is burned out of three after tafiks, the ship evens out, When the stern rises to.the point where propeller emergency or sea keeping ability becones a probiem, sea water ballasting will be needed in these empty after tanks. Even though the fused salt systém has a vertical center of gravity 4.5 feet below the boiler plant, the total nuclear powered ship C. G. gtays about the same due to empiying 173 tons from the relatively low fuel oil and ballast tanks forward. Since the total ship weight and free surféces stay about the same, the free surface corrected wmetacentric height (indication of ship stability) of about 3.2 feet stays about equal to the conventional DD931l., Moment calculations showed that the exact change in metacentric height was sensitive to more exact values of weilghts and centers of gravity than could be calculated for the miscellaneocus items in this feasibility study. Arrangement Two The finalized reactor compartment width was reduced to 23 feet. This reduction allows the possibility of locating the nuclear plant in the after engine room with onlylthree to four feet of off centerline required to avoid the forward propeller shaft. This is shown in arrangement two (see Figure 11-1). With fuel oil shield tanks on both gsides of the reactor com- partment, the dangér of serious list if damaged is lessened. A detailed £l IVEANIQIANG AT B IR | AUOALVId GNOT3IE MHTL wINIee s 1O "RENL <7 ik 40 3 u LD LAY fomainn o4 to3l TON JYUNOLL ¥IAAOHLERD SEVYTD 1864 QC HOLDVIY ANIMYN ZINVHHOIN3d RIIH o . MOON INIDNT | 0N 1'OKr "WAS CRHNL (I ] FON N3 'ORHAL ool ARV o ) AL aanie nOON AOY SO I W ) me XIS 37t4Q0Nd QHVOBHN! T ARIWIDNYUNY AMYITIXOY $ HOLOYIY ¥ ININAONTHYY ANYLUIXOY JHOLDYIH ANDRLNYA RDY NOLDYEY Tanmniwsenns ADLY TN 1 1 20vdil WOOM AMYIUXAY 4 AD v WOON TYLLNGOIANOD OLE 'OMI-uW]-THED 1 \ 1 — —— . ~164 - damage stability evaluation should be made before arrangement two can be recommended with certainty, but aside ryom damage contingencies, the arrangement offers the advantage of putting the heavy, concentrated weight of the reactor compariment near the C. G of the ship, thereby requiring jess fuel oil, 303 tons, to balance the moments to give egsential the con- yentional full load condition trim aft. The transverse stability gituation is better than arrangement ope in that the forward tanks have considerably less free surface than the after tanks which are =mpty under BPMR, arrange- ment two, full load conditions. The resultant free surface correction is .16 as compared %o 1.61 feet for arrangement one. These forward fuel oil and ballast tanks have a lower cente: of gravity than the after tanks used in arrangement one, but they hold less oil. Again as in arrangement one, exact values of metacentric height cannot be calculated with any confidence without a more detailed machinery arrangement, but 1t is indicated that arrangeument two has bvetter stability than arrangeuent one and has strong possibility for good improvement over an oil fired DD931. In su@mary, two general afrangements were worked on., Both give the big advanfiage of decreased space required. A detailed arrangement of the auxiliary room was not worked out, but a relatively large amount of the original fire room ig 1left both aft of and sbove the reactor com- partment, In arrangement one, fourteen longitudinal feet of deep tanks are freed for armament stowage or other use, Also a portion of the room left for auxiliary space could be used for stowage. Stability looks to be roughly about the same as & conventional DD931 with increased oil C.G.'s balancing a decrease in steam generating center of gravity caused by elimination of uptakes and stacks and the design of -165- a compact, low reactor vessel, steam generator and superheater, and sur- rounding shield. The free surface’correction can be éontrolled to some extent by keeping the fuel o1l shield tanks full and under slight pressure, As a feasibility project, the vertical moment study has indicated that a detailed design with an eye to the stability problem, especially in an arrangement of type two, cofild lead to an increase in metacentric height which could be gladly used by the armament people to add migsile launching and guldance systems topside; Perhaps the greatest restriction in these arrangements is a lack of flexibility in filling and emptying fuel oll and ballast tanks, Salt water nust be used for trimming purposes which brings up contamination problems, When fuel oll shield tanks are tapped, fuel oil or salt water must be pumped into the bottom Lo maintain the shield and =eliminate free surface. 11,3 Power Plant Control 11.3.1 Introduction Due to its negafive temperature coefficient of reactivity, the fused salt circulating fuel reactor is self-regulating. That is; the pover produced in the core of tfie reactor follows the power demanded by the load with some characteristic time lag., The steady state mean temperature of the fuel in the core remains constant since, in the absence of control rod motion, burn up, and fission product build up, the reactor is critical only at one temperature, Of course, other temperatures throughout the system will véry with load, | Bven though the reactor is self-regulating, there are several reasons why a control system way be incorporated in the power plant design, First ~166- of all, the transient response of the system may be poor, Fér example, load changes ma& resuli in lafge temperature overshoots which, in turn, cauge intolerable thermal stresses, A properly designed control system can ifiprove transient response. A control system may also be used to set up some desirable pattern of steady state temperétures, pressures, and flow rates throughout the plant as functions of power output. Such a pattern is called the plant program. FPor the HPMR a constant steam temperature program is desirable. This requirement is dictated by the fact that steam turbines for marine power plants reguire esgentially constant steam condltions regardless of load. 11.3.2 Types of Control Systems Several types of control systems seem to be possibilities for establishing the constant steam temperature program. For example, control rod position in the core may be varied as a function of output steam tempe;ature. With é negative temperature coefficient of reactivity, con- trol rod poéition determines the mean fuel temperature in the core and thus fixed the level of temperatures throughout the system. Thus, ii seems possible that steam temperature could be maintained at a constant level by such a system, Another system which strongly suggests itgelf is controlled by varying the flow rate of'the inert salt in the intermediste heat transfer loop. The rate at which heat is carried away from the primary heat exchanger depends on the sglt flow rate and the difference between inlet and outlet salt temperatures. Thus, if flow rate is varied with power output, the steam temperature can be mainiained constanfi. This system has the distinct advantage that the pumping power required decreases with decreasing load, -167- o There is & resulting gain in efficlency which is lacking in the other control systems, There 1s one other factor which should be considered here. In any system in which the flow rate is varied, the possibility exists for tra;sitions from turbulent to laminar flow and vice versa. Such transitions usually result in large thermal shocks and are highly undesirable. In the HPMR power!plant salt flow in the primary heat exchanger and steam genergtor is laminar at design power and is well into the turbulent region in the superheater., Thus, the flow rate can be varied over a wide enough range to make control by this method feasible. | A third possible control system involves & by-pass line across the salt side of the primary heat exchanger, As the load is decreased a valve ° in the by-pass line i1s opened allowing a larger percentage of salt to by-pass the primary heat exchanger. Thusg, returning cold salt is mixed with the hot salt from the heat exchanger with the result that salt tem- peratures throughout the'rest of the system can be adjusted to hold steam temperature constant, A study to determine the optimum control system was not attempted due to time limitations, 11.3.3 Simulation It was decided to set ué an analog simulation study of the reactor and power plant on the Reactor Controls Computer (Reference 30) at the Osk Ridge National Laboratory. The study had two main objectives: l. To determine the transient response and stability of the reactor and power plant when subjected to changes in load, changes in reactivity, and other perturbations, 2. To determine the ability of one particular type of control system to maintain constant steam tempersture, -168- For details of the simulation, c¢ircuits used, etc., see Appendix 1l.1l. A schematic diagram df the system which was simulated is shown in Figure 11.2, Two heat transfer circuits are shown; each handles 62.5 megavatts at full power. Due to the limited number of amplifiers avail- able on the computer Only'cirauit 1 was simulated in detail, In circuilt 2 as shown in Figure 11,2, the superheater and steam generétor vere approximated by a single heat exchanger. Circuit 1 represents the arrange- ment of components as visualized when the study was set up, It is not markedly different from the arrangement finglly decided upon, The control system chosen for simulation was the by-pass line across' the primary heat exchanger;. This system waé chosen because it was relatively easy to simulate and because it offeved good possibilities for control. Not enough eguipment was availlable to simulate control by varying salt flow rate. No control system was simulated in circuit 2. 11.3.4 Results - Since the details of the_simulaticn are presented in the appendix only the results will be indicated here. In order to study the transient behavior of the reactor and powexr plant a nuwber of runs were made with the control system inoperative. The result of the first such run is shown in Figure 11.3 With the reactor operating in steady state ét full power, the load demand was reduced linearly %o one-half power (62.5 megavatis) over a period of 15 seconds. As can be seen from Figure 11.3, reactor power followed the load demand and stabilized at half-power with no undershoot. The mean fuel temperature in the core reached a peak of about 1236°F and then returned to its steady o state value of 1225 F also without oscillation. ~169~ The results of the above test seemed to indicate that the transient response of the system was completely satisfactory., As further verification, it was decided to subject the plant to a wmore severe load change. In this test the load demand was increased instantaneously from 10% pover (12,5 megawatts) to full power, The results are shown in Figure 11.L, Reactor pofier and temperatures throughout the system leveled out at steady state values without_oscillation,' A number of other runs involving load demand changes were made inciuding cases involving 25% and 50% overload. In all cases the reactor and power plant éppeared tc behave as a critically damped system; that is, reactor power followed load demand without oscillation and temperature swings throughout the system were very mild. Several runs were made to investigate the effect of gtep changes in reactivity. The results 6f one-éuch test are shown in Figure 11.5 At t= 0, a step change in reactivity of +0.2% was introduced. At t=70 seconds, a step change of ~0.2% was introduced. All of the tests described above seemed to indicate that the transient responge of the reactor and power plant was satisfactory. Therefore, phase two of the simulation was devoted to the study of the control system, As stated previously the purpose of the control system is to establish a constant steam temperature program. The system which was simulated is a by-pass line across the sal®t side of the primary heat exchanger. The amount of salt flow through this line is determined by the steam temperature by means of an elementary servo system of the on-off type. The salt flow which could be by-passed through this line was limited, in one case, to 75% (1570 pounds/second) of the total flow and in another case to 90% -170- (1880 pounds/second) of the total flow. Figure 11.6 shows steady state steam temperature as a function of load for these two cases as well as the case where the control system is inoperative, Steam temperature is held constant over the range of 60% power to 100% power for the 25% flow. cutoff and over the range of 30% power to 100% power for the 10% flow cutoff. The amount of salt which may be safely by-passed is probably limited by the temperature difference across the salt in the primary heat exchanger. A%t any given power level this temperature difference will increage as the salt flow rate through the exchanger is decreased. No study was attempted to determine the maximum tolerable temperature difference, | Also of interest is the transient response of the power plant and, in particular, the steam temperature during load changes., Figure 1l1.7 shows the resulfs of a run in which power demand was decreased from full power to half power in 15 seéonds. The steam temperature sitabilized at its design point value (975°Ff* afterlafiout 100 seconds. The maximum excursion of the steam temperature was almost 100°F, It should be noted that.little attempt was made to optimize the control system., An optimum system would undoubtedly improve this trangient response. It is interest- ing to note that reactor power undershbots 1ts steady state value after the load change. This is in contrast to its behavior with the control system inoperative. Temperatures throughout the system also oscillate slightly. *mmis value was changed to 950°F in the final design. ~171- 11.3.5 Conclugions The results of the simulation indicate that the kinetic behavior of the reactor and power plant is completely satisfactory. These regults alsoc demonstrate the feagibility of a control system to maintain a constant steam temperature prograu, A more detalled study of all of the possible céntrol systems is required to determine which is optimum, Because of the higher efficiency obtalnable, control by varying salt flow rate appears most atiractive at this tinme. 11.4 Emergency Operation It is extremely important that any reactor installation subject to battle damage be as inherently safe as possible. The demonstrated stability of reactor systems of this type (Ref. 6) along with the elimination of numerous integral control rods mekes it basically very desirable. In addition, however, consideration has to be given to emergency conditions, both major and minor, not only to establish an overall safe system but to maintain operation if possible and to prevent damage to the reacior. A partial list of such considerations as applied to this system is given below: (1) Primary fuel pumps are over-designed sO that high power operation can be maintained in the event of partial failure, (2) Primary and secondary pumps are d:iven by steam (available from both the reactor and conventional system) and backed up'b& an electric motor which can be operated from emergency service. -172- ORNL=LR=Dwg , =25766 UNCLASSIFIED avon 133HS MO1d HOLVHEINIO WVY3lS = . " ¥IONVYHOX3 YW 1V3H , AHVAINd 3405 M .n@ -u.@ .u-.m ua?fi $ _:W . :»@ 0 .pwm. Jw = X ¢ = - /= avot “e [s *m vy |0 g d31v3HY3dNS ¥ 1IN0¥HI0 2 L1InJO¥HID Y3I4SNVYL LV3H ¢l 3¥N9ld 434SNVYL LV3H 03S — 3NWIL 06 08 0L 09 oS ot 0§ 02 ol 0 [ D T~ e L e = * ~—— MM ~< 1 S mc ../il = ~ & ™ o /. N £ GNVW3IG ¥3IMOd NI // 1 JONVHO dWVH V HOd4 3WIL SA 4N3IL ONV H3IMOd HOLlOV3IY _cozi| dN3L 1304 NVIN —-— // YIMOd HOLOVIY =~ ——— 0S121 OGNVYHW3AG ¥3MOd \ \ 05232 —= N\ ....../ \ 0o GE3! — - o p—" 0S¥ 2l ¢l 34N9id 06¢ Gl 00l Gel MW - d43IM0d 06 08 0L 09 oS '03s — JNWIL o+ o¢ 02 ol 0 2 ;. B A ANVN3Q ¥3IMO0d NI / LE ———39NVHO d31S V H04 IWIL SA 7 mw dW3IL OGNV H3IMOd HOL1oV3Y 32 = 25 m & W ‘ - - / m \‘ © / M 7 z / . / T % / 2 c Vv 2 S I -~ - e \ 9 ...ll.....\..\\ Ifl_ R .I.l..llll...l.ll.ll <31 | — dW3L 1309 NVIN — - . HIMOd HMOLOVIYH————m— - P / §02/— anvw3a ¥amod T \ l....\ll_l_lll |I.._..|\....l| / 22| e i St 2l Y 11-34NOid G2 0G w0 ~ MWN — 43MOd 00l G2l UNCLASSIFIED ORN L~LRe Dutgo =25769 «175=- '93S — 3IWIL oS! 0zl oll 06 oL oS o¢ ol 0 o yad 7 -~ ~ | \.\\ : - —— G2l = .1...././ * = ~ Z N Il N n c \ _ m \ — \ = | = ¥3IMOd ¥OLOVIY——- 062 3 JUNLVEIdWNIL 13Nd NVIW c22i — 052 // Slélinas0L=} 1V % 20- ANV O=4 1V %20+ 005 | ALIAILOV3Y \ | NI S3IONVHO d31S ¥O04 INIL SA \ JHNLVHIdNIL ANV mmzom_mozqum | . G2t ! _ St J4NOI4 MWN — H3IM0d HOL1OV 3y OKNL=LR~Dwg. =25770 UNCLASSIFIED -176- 09l VOl ¥3MOd 11nd 40 LN3dD¥3d 00l 08 09 ov o< ovlI 02l /.l- /]Ir.... nnumew“;;;;; o N\ /// \ ..// /...r ;;;:;f::,; 4///1 440100 %01 —WILSAS TOHINO) ====== 4401N9 %GS2 —WILSAS “TOHLNOD WILSAS TOHINOD N e o | avol SA FUNLVYHYI4dWNIL WY3LS LNdLNO 9l 3IYN9IL 008 006 N Ilm m ) - oool * _ —{ m = 0 o 00I1 2 nl.— c X m oozl | 2 00¢ | ORNL~LR-Dwg, =25771 =177~ (do) 3¥NLVYIIWIL WV3ILS 03S— 3NIL UNCLASSIFIED oLl 0S| o \ o oyt - L o ...l..l....l.l.ll.llll-l.....lll._ . - 13INOONI QIVINNY [ — & o8l 4 X n 3 i S ogl 0’9 m— ......r/ \ x N T~ % Obl = — 02 O — a37704 po: f,/ = R m m ss INQuvH TTINIEE — og 08 X— < > > ] 091 3 ——0'6~N— =] < 0Ll »——0"0} > — NOISNVdX3 40 LN310144309 @ o, 08l Ol ~» -n Viva N9IS30d T3INOOINI -GSV JHNOIL 1 -205 - ORNL-IR“DWE . "2 57 ?3 UNCLASSIFIED FIGURE A-5-2 | | : INCONEL DESIGN DATA 0.1é I | CIFIC HEAT BTU/# °F o > .15 T tSF’E o o F-FT pd ) » o THERMAL CONDUCTIVITY P / / o o /<_SP£ CIFIC HEAT e o o v @ o // / L~ » o A THERMAL GONDUCTIVITY BTU/HR- > 2 400 800 1200 1600 TEMPERATURE, °F 2000 -206~ ORVL~LR=Dwg. =2577h UNCLASSIFIED FELONGATION (% IN 2") TENSILE MODULUS (10° PSI) 120 100 60 TENSILE STRENGTH (103 PSI) 4of 30 YIELD STRENGTH (103 pPsl) 257t 20} 10t FIGURE A-5.3 INCONEL DESIGN DATA ™~ ] X/TENSH.E STRENGTH / N\ \ / \ / \ ELONGATION N — W —~. S YIELD STRENGTH N (0.2% OFFSET) \\\ . MODULUS/'/)\ N\ “ " - ELASTICITY N 000 1200 1400 6 00 (1800 TEMPERATURE (°F) -207- APPENDIX 6,1 JUSTIFICATION OF MODERATOR MATERIAL Allowable Moderator Rod Size The best way to justify a material selection is by 1ts satisfactory performance under actual operating conditions. Beryllium-oxide has suitably'withstood a preliminary evaluvation in the MIR under both high temperature radiation énd cyclic operatiofi (Reference 54 and 55). While the specimen. size and test conditions were not idenfiical to that proposed for this study it is é reasonable first approach to calculate the critical stresses that were withstood in this test and then apply these %o the current case. The lifiiting stress in the case of internal heat generation, especially for a ceramic, is its tensile, The specimens from the MIR test were 1l in, in diametef and were satisfactorily exposed to & power generation of approximately 15 Watts/cm3 at a temperature of 1500°F, Utilizing the development given in Ref. 10, Bquation LVIII, on the working curves of Ref. 3, the maximum tensile stress is found to be 3000 psi. This will now be used as.the deslgn basis for the selection of a maximum allowable moderator size for this study. The energy release rate is approximately 197 mev/fission which may be broken up roughly as follows: Local deposition - 165 mev/fission - Fission product kinetic energy 5 mev/fission - Fission product decay beta particles -208- Non-local - 6 mev/fission_ ~ Fisgion product decay gaumas 5 mev/figsion - Prompt fission gammas 5 mev/fission - Prompt fission neutrons 11 mev/fission To determine the energy deposition rate in the woderator, it is felt quite conservative to assume that all of the neutron energy is absorbed in the moderator, On the basis of this it was felt juStifiable to increase the moderator rod size to 0.75 inches diameter although additional testing would be required to obtain definite verification, Moderator Rod Temperature Digtribution - By making the reasonably valid sssumptions of steady state, negligible axial heat flow, and non-variance of the materials thermal properties with temperature, the teuperature distribution through'a clad cylinder due to internal heat generation can be found from the basic equation: oo gtk = b (e ) r By applying the proper interface and boundary conditions this reduces %o: In the cladding - r, = r LT, 2 ‘bc(l‘) ’b(rc)-}-. ——-——nc—-—-—— 1 = (....?r__) (q LYy _ C r T Neutrinos (deposited at o) qcili) 1 ~209- On the moderator - 0&1r < r, m b1t p 2 6 (r) = t(x,) +_f—%—m—— [1 - (.%.)2:‘ where: the subscripts ¢ and m refer to the cladding and moderator respectively t(r) = temperature at point r g''' = heat generation rate per unit volume k = thermal conductivity Ty = outer radius of the moderator r_ = outer radius of the cladding Applying these results to the case of a 0,75 in diameter cylindrical rod of BeO with %0 mil Inconel cladding (k = 14,5 and 16 BTU/hr-ft-°F at approximately 1500°F) the temperature distribution given on Figure A6.1 was obtained, It should be noted that a temperature drop across the interface 6f the moderator and cladding is not included at this point.and that the other non-local energies were lost from the system, The possible gamma energy absorption was not included because the relatively light welght BeO gives poor gamme attenuation making this effect within the conservation of the previous statement. Maximum energy deposition in the moderator is then 2.% (l =55 ) of that in the fuel, Considering the power distribution fofind acrogs the core of 1.4 (Section 8.2.2) and a probable axial distribution. of the same order gives an overall peak to average power of approximately 2.0, The maximum heat generation in the moderator rods can then be calculated by: ~210- Power density in the fuel = 700 watts/cm3 Fraction fuel volume = .50 Fraction moderator volume = 4O Genefation rate in BeO = 2,.9% x 2.0 x 700 x L%gu = 50Iwatts/cm3 ?he critical location for the moderator'materialois approximately at the central region where the powef'density is & mgximum. Although the tem- peraturelof_the fuel increases toward‘the exit of the core, the power density decreases at a ra#e stfficient to cause a net reduction in the moderatorltémperatfire.i It is estimateé later in this appendix that the waximum surface temperature (location of maximum tensile stress) of the moderator is 1410°F, C;omparingvthis with the MTR test information at 1500°F and the basic strength characferistics of BeO_With temperature (Ref. 9), it is appareht that‘no correction in the allowable stress should be made. Utilizing the allowable stresé of 3000 psi and a maximum uniform heat generation fafie of 50 watts/cmB, the limiting-cylindrical rod diameter based on test data is calculated to be 0,6 inches (Reference 3). - However, the folldwing considerations should be made before the rod size is limited to this value: | 1. The above calculations were fel£ to be conservative, 2. Actual tensile strengths for BeO of 9000 psi were measured (Ref. 9) 3. MIR tests which ran the ténsile strengths to 3000 psi gave éatiSfactory performance, Applying these results to the care of a 0,75 in. diameter cylindrical rod of BeO with 40 mil Inconel cladding (k = 1L4.5 and 16 BTU/hr-ft- F at approximately 15000F) the temperature distribution givefi on figure A6.,1 was obtained. It should be noted that a temperature drop across the inter- -211- face of the moderator and cladding is not included at this point, Pemperature Rise Across Fuel Boundary layex The temperature rise across'fhe bcfindary layer between the fuel and the moderaior-rods waé caléulated by means of Reff lland 2, It is shown that the overall temperature rise is geparable into the sum of two temperature differences (l) Due to the temperatuxe drop required to remove the heat generated within the moderator material and (2) Due to the temperature rise through the boundary layer duse to decreased velocity and thus higher power density. | The followiné constants were found for.the reactor core operating at 125 MH with a temperature rise of 100°F and a mean temperature of 1200°F, Flow Area | - =2,38 ft? Hydraulic Diameter - 0546 £t Velocity _ -g.1 ft/sec, Reynolds Number | =2 X 10lL Prandtl Number " =3.6 fuel Thermal Conductivity = 1.3 BIU/hr-ft-OF Using Referxence 1, with an equivalent cylinder gives a Nugsault numbe$ of 100 for the (1) solution., (Use of the hydraulic diemeter analogy with this method is indicated in Ref. 66). Nu = 100= hd _ a = - 9 KAt T Tk For a rod of .830 inches (.75 + .08 cladding) a/A = g, x Volume Assuming heat generation rate in — moderator and cladding are equal. -212- at q"’m v d which upon substitution reduces to U8 T T 100A ' At =701 q"'m where g'" = moderator heat generation rate in vatts/cm3 Total temperature rise across the boundary layer at the center of the core 1s then AT = 701 q"& + .01l q;‘ q"s = 700 x 2.0 watts/cm> where 2.0 is the peak to average power q"! = 50 Watts/me as discussed previously then: .gfiT = 5OOF — temperature increase of outside cladding above uwean fuel temperature. Temperature Rise Across Cladding - Moderafor Interface Becauge this interxrface gap 18 quite small compared to the rod diameter even at operating tewperatures (Sec. 6.2.1) it may be treated quite accurately by an approximation to a flat plate as: q/A = k AT 1 q/A _ qn!m Volume _ qll!:;:i a _ .750 qil!m Area T = 1510 ¢"' where q"' = wafits/cm3 Because & shrink fit of the cladding around the BeO appears 1o be feasible a maximum of one mil clearance should be realistic. If the shrink is made in a helium atmosphere k is estimated to be .1k BTU/hr-£4-CF (Ref, 20) Ar _ 1510 @'y (,001) = M5°F for q"' . 50 watts/cmd (.14) 12 ~213- Total Temperature From Figure A-6,1 the temperature use through the cladding can be estimated as AT/q"' = .3k AT — 17°F Similarl& the temperature rise to the centerline of the rod (neglecting the interface) is found o be: | 4T/q"£} ¥ 1.96 AT = 98°F Combining the temperature rige across the boundary layer, cladding and interface with an assumed maximum fuel temperature at the center plane of 1300°F {ave. temp.=§'1225°F) provides a maximum temperature at the BeO éurface of 1412°F. Similafly the maximum temperature at the moderator centerline is found to be 1493°F. ONIgavo dOLVHIQON 2 v7,, | 0’1l 860 $6°0 60 80 90 v 0O ¢ 0 0 // - _ / ol IONVHD 31voS 310N — ] | ORNL~LR=Dwg, »25775 UNCLASSIFIED 21k O (3} -0y ¢Wo/ SLIVA ALISN3Q ¥3MOd OSNIgavo =2 wib; 1w = 2 gWo/ S1LIVM= ALISNIQ mob R 43M0d HOLVY3IAOW = ,; . 02 «SIP =SNIAYY ¥3LNO = 2 2 "b/1V [ dvid T3NOONI ObO . 038 "ViQd ‘NI 06/ NOILNEGIYLSIO 3JYNLVHI4WIL QOH MNOLVNIQOW I'9-V 2dn9id -215- APPENDIX 6.2 CALCULATIONS FOR FINAL DESIGN OF PRIMARY HEAT EXCHANGER 6.2,1 Basis of Design The primary heat exchanger is designed to transfer 125 megawatts or 4,27 x 108 Btu/Hr from the fluld fuel to the secondary coolant. Many gquantities which would ordinerily be considered design parameters were, due 1o the short time allowed for this study, given what seem to be reasonable values and held invariant throuhgout the calculations. The following calculations represent the final iteration of the most promising combination of tube diameter and spacing as indlcated by Figure 6.1, 6.2,2 Properties of Fuel, Secondary Coolant and Inconel As given in Section 5.0. - 6.2.3 Quantities Determined Before Employing Iterative Procedure Heat Bxchanger Inner Diameter = 53.5 in. Heat Exchanger Length = 48 in, Tube Wall Thickness - .00 in. Temperatures : Fuel Entering = 1275°F Fuel Ieaving - 1175°F Coolant Entering == 1050°F Coolant Leaving = 11500F -216- 6.2.4 Quantities Determined by Iterative Process Tube Outer Diameter = ,200 in, Tube Spacing - 030 in, Reactor Outer Diameter ~ T13.7 in, Number of Tubes - k43,420 6.2.5 Flow Rates | | | . a, Fuel Loop Flow = .C A?T _ _h-QT X lgt Btu/Hr x — - P . 264 TH5-OF 100°F _ 16.2 x 100 B - R 4,27 x 108 ifi“ b, Coolant Loop Flow = —g g 7 - . 4 Biu o, P 57 THE-OF 100°F = T.49 X 106 LB : Hr 6.2.6 Flow Velocities a. TFor specific values of tube dismeter, tube spacing, and reactor outer diameter, the approximate aumber of tubes contained in “the heat exchanger are determined as follows: -217- let 4 = tube 0 outer let S = SPace thickness diameter & = Dp - Dy 2 Number- of vertical rows of tubes P-— [9:3 (do+s)} (D + Dy) N = Z "3 (do+ 8) Number of horizontal rows of tubes Q:—Wfl—dofs.i- 1 Dy - D, - 58 - 3 do X = 2 De - Dl - 5 = 3 do + 1 W:-_. 2{do +« s) Total number of tubes 'W(DQ + D])} [ n = N??: 3 Ido-+ sj D, = Dy - 58 - 3do 2 (do + s) 5 Y ~218- w(Dy + Dy) . TEVI @+ a2| (P2 Py -3 -0 Using the results of the final iteration, nfi[wfi&7+5&wfi 2 Y3 (.200 + *'030)2jl 131 = 53.5 - 050 - .200 = 43,420 tubes b. Coolant Velocity } zc 7,49 x 10° 1b/nr _ . ¢ ffe E.23 1b/ft3] [43,11-20 "“1-&?24;)-080) Ft?l] = 17,830 ft/hr = 1,95 ft/sec ¢, TFuel Velocity v We 16.2 x 10° 1b/hr toephe (208 w/e83][ I (73.7 - 53.5% - 43420 x .2%) | 14 x 144 = 17,150 £t/hr = 4,76 f£t/sec 6.2.7 Fuel Side Hydraulic Diameter d . 4 x cross sectional area h = wetted perimeter 2 w [T5mr (7377 - 53.57 - 43h20 x .22] “n 13~ (73.7+ 53.5 + 43,420 x ,2) .00788 £t I -219- 6.2,8 Reynold's Numbers a. TFuel Side Reynold's Number 5 _ % of T Fe _ (208 1n/£47) (00788 £1) (17150 £1/nr) 18 1b/hr £t = 1560 b. Secondary Coolant Side Reynold's Number R - (Ocdivc ec Mo (123 lb/ft3)(;9%§9 £4) (17830 £t/hr) 53,2 1b/1t nr = 412 6.2.9 Prandtl Numbers a, Fuel Side Pr_ - Cpf’flf _ (.264 BTU/1b °F)(18 1b/ft sec) £ ki 1.3 BTU/hr-£4-CF - 3.66 b, Coolant Side -220~ 6.2,10 Filw Coefficient a. TFuel Side (See Figure 7.6) 0}+ - o = W (prg)” (Re,) ° Q - o G 6ot s O = 1836 BTU/hr-ft°- F b. Coolant Side (Ref. 17, page 232) . a. hy = 1.62 XC (Re Pr __t )1/3 94 L Ft 0 1 ~ 1.60 (2:h BTgégr £8-F) (410 x 12.6 x 2220 ) /3 (55 Ft) e o BTU = 9 G ree-op 6.2.11 Overall Heat Transfer Coefficlent L U _ — 0= _ L+ - e in Yo R~ In_ 2k L _ 1 1 + .200 in 1836 BTU/hr-££°-°F (,120 in)(915 BTU/hr-f+°-CF .200 Ft) in '(,200 ) 12 . 120 (2 x 14 BTU/hr-ft-"F) ~221- 1 ' BTU > ~ 100182 + .000545 + .0030%) Hy-Fte-OF - 37% —BIU He-Ft2-"F 6.2.12 Total Heat Transfer Area A=nwd L = 43420 x v x (.:..f...g_o £1) x (.;&.g. £t) = 9090 Ft2 6.2.13 Mean Temperature Difference Required to Transfer 125 Megawatts of - Heat 8 AT - 9 (b.27 x 10 BTU/Hr) ® AU (9090 F7)(37h BTU/Hr-Ft -OF ) = 125 4°F 6.2.14 Temperature Drops Across Surface Films and Tube Wall a, Fuel Side Film 8 Ar . 9 _ b.27 x 30" BTU/Rr o hA (1836 BTU/HreFt--°F) (9090 Ft°) = 25.6°F b. Coolant Side Film d AT - 9 h.27x 10° BTU/Ar , | 1o A, (915 BTU/HEr-Ft2-OF) (2320 x 4 x 7 x 43420 Ft ‘ 12 = 85.5°F c. Tube Wall AT = 125.k - 25.6 - 85.5 = 14.3°%F -222~ 6.2.15 Approximate Average Surface Temperature of -Tube a, Fuel BSide +3 i Inlet Surface Temperature + Exit Surface Temperature _ (1275 - 25.6) + (1175 = 25.6) : 2 = 1200°F b. Coolant Side Twc:rlnlet Surface Tewperature + Exit Surface Tempersture - 2 _ (2050 + 85.5) -+ (1150 + 85.5) 2 1;8605 6.2,16 Approximate Average Tube Wall Temperature + T 0 Tye Y Tue 1200 + 1186 T e — wa 2 2 = 1193°F 6.2,17 Friction Factors For Both Sides of Heat Exchanger and Reactor Core a, Heat Exchanger, Fuel Side F {Re ) (Ref. 13) o 2 b. Heat Exchanger, Coolant Side f c (Ref. 15, page 50) h jo\ a|&= : C 6h = g5 = 155 6.2,18 223~ c¢. Reactor Core f < .02k (Ref, 15, Figure 2-21b) o See also Appendix 6.2.18 for Reynolds Number in core Pressure Drops for Both S8ides of Heat Exchanger and Reactor Core a. Heat Exchanger, Fuel Side - T e . AP = fF o —F%) Ftc (Ref'. 15, page 45) af 2(g 5 ) Bec 5 ~ I (4.76) = 0921 .00788) 2(32.2 - = 16,6 Ft _ (208 1n/FE2)(16.6 Ft) 2 1l in_g Ft . = 24,0 psi b. Heat Exchanger, Coolant Side ap-+s B _(Ye) F 4. 2g 1 i1 55 (b FE) (4.95°7%° 120 2 x 32,2 Ft 5 Ft ey sec = 23,7 Ft (123 Lb/FtB)(23,7 Ft) - = 20.3 psi (144 In® ) 2 e -22h - ¢. Reactor Core ' 2 1. Flow Area = A_ = .50 L ( 75 cm ) * 4 12 38 4 5 gy <2 Ft 7 In 2 = 2,38 Ft 2. Wetted Perimeter = P_= fi’I cm , 10cm 12 2 2,54 S0 cu FE Mmoo Sk + 749 x .75 In) = 156 Ft. See Bection 3.3 for dimensions of reactor core. 3., Hydraulic Diameter = dhr - hox A, - - 4 x 2,38 P 1 - 2 = 0610 Ft ' 6 “. Velocity . Wy _ _16.2 x 10 Iv/Hr Cp * Ay 208 Ib/Ft> x 2.38 Ft° = 32,700 Ft/Hr = 9.08 Ft/sec 5. Reynolds Number = Rer _ Pfdhrvr (208 Lb/Ft3)(.0610 Ft) (32,700 Ft/Hr) - }Jf o 18 LB/Ft-Hr = 23,1 3,100 5 r r ( d ) 2g : hr Lo, 2 ~ ook (3.9% Ft) (9.08 Ft7) , L0610 F% 2 x 32.2 Ft/Sec 6.2.19 ~225- 3 = 2.0 Ft = 208 Lb/Ft2 x 2,0 F% In 1k - Ft == 2.89 psi Pumping Power Requirements for Both Sides of Heat Exchangexr and Reactor Core : a, Heat Exchanger, Fuel Side b ] (W, Ib/Hr) (AP, F ) £ (1.98 x 10° %;%) FHP (Ref. 15, page 80) 6 Lb (16.2 x 10° = )(16.6 Ft) Ft-Lb Hp-Hr 1.98 x 100 = 136 HP b, Heat Exchanger, Coolant Side prp o (1.9 X 10 Lb/Hr)(23.7 Ft) 1.98 x 10° Ft-Lb/Hp-Hr = 96.0 HP ¢. Reactor Core FHP = (16.2 X lOé Lb/Hr) (2.0 Ft) ¥ 1.98 x 10 Ft-Lb/Hp-Hr = 16.3 HP -226- APPENDIX 7.1 STEAM GENERATING SYSTEM i.- Heat-Transfer Calculations for Steam Generator A, 'Incofiel Tube Data: (1) size; 5/8 in, 0.D., 1/2 in, I.D. (2)I Pitch; 3/4 in. delta array (3) Thermal Conductivity = 11.3 Btu/hr-ft-oF (h) - 8pecific heat = 0.124 Btu/lb;oF (5) Density = 510 1b/ft3 | B, Steafi Generator Inlet Conditiofis: (1) Molten Salt: (8) T = 761.8%F (95.9 MW); T = B0O°F (125 MW) (b) w= T.49 x 105 1b/hr (c) cp, 0.57 Btu/1b-"F (8} p =130 cp (95.9 MW); p= 126 cp (125 MW) (e) © = 127 1b/td (£) k= 2.4 Btu/hr-£t-"F (2) Water: (a) T~ 564°F (b) P - 1250 psia (c) h = 567 Btu/1b (d) w = 3,230,000 1b/hr (95.9 MW); w = 14,149,500 (125 MW) (e) wvel, = 6.24 fé/sec (95.9 Mi); vel. = 8 ft/sec (125 MW) (£) spec. vol. = 0.0221 ££3/1b -227- Steam Generator Outlet Conditions: (1) Molten Salt: (a) T = 703°F (95.9 M{); T = T24°F (125 MW) 6 lb‘/hr (b} w=7.49 x 10 (c) 'cpa-_ 0.57 Btu/1b-°F (d) p = 195 centipoises; ML 175 centipoises (125 MW) (2) Waters (a) T= 572°F (b) P = 1250 psia (c). h = 579 Btu/1b () w= 2,874,970 o/br (95.9 M); w= 3,689,000 (125 MW) (a) T 572°F (b) P = 1250 psia (c) h= 1181 Btu/hr | (d) w< 355,030 1b/hr; w = 456,000 1b/hr (125 MW) Water Flow Ares and Number of Tubes: (1) Area. ¥ X (spec. vol.)_ (3.23 x 10° 1b/nr) x (00221 £43/1p) =3.18 £t vel, 3.6 x 103 sec/nr x 6.24 £t/sec 3 (2) Number of tubese Lobt8l Flow Area 3.184 ftz - = 2336 Flow Area per Tube 66136) £42/tube Salt Flow Characteristics: (1) Flow area/Tube = 0,867 (Pitch)2 - _’Lf_ (cz)2 = 5 V@ 625 12 0.867 (‘—E‘) _g.. (_122_) — 0.00126 ft° -228~ Total Flow Area = 2336 tubes x 0.,00126 fte/tube.: 2., a0 W 7.49 x 10° 1b/br (3) Hydraulic Diameter, De = hA a 4 (0.00126 £t ) Td (_“égi.ffi) » (4) Re veDe 2. 58 ft/sec x 127 lb/ft3 x 0.0308 ft H 170 ¢p x 6.72 x 10” -l (1b-sec/ft)/cp | Nu . | (5) From Fig. 7.6; *;—573 = 2,15 (95.9 W) = 2.7 (125 r 127 1b/ft3 x 2.94 fta x 3.6 % 105 sec/hr 2.4 ol £t° = 0,0308 % =191 (95.9 MW) 258 (125 MW) It MW) | ‘ ). b= 2,15 x 2 [;C Flo- 235 (2.4) [50.57)(2.h2) 170:k 0.k | K d 0,0308 = 1015 Btu/hr-fta-oF = 1170 {125 MW) 1 =;_-—-—-_-—"20| 0 6 Rsalt 1015 00098 Inconel Tube Wall Regilstance: 3125 0.3125 ) Lot (ro/ri <; o 550 (1) R _ = = 11 3 — 0.000485 wall Boiling Water Film and Scale Reslstance: (1) Ren® (b 4 - T h h i scale boilin | 1 1 | = 1,25 (—5553 + 30—0(')—>= 0.000833 = 5,58 £t/sec -229- Heat~Transgfer Coefficlent for Weter in the Tubes: e (6 24(2:2) 1. D (1) Res o 5 oos )(36;;10)3188::10 (95.9 M) - = 2.4 x 10° (125 MW) (2) h=0.023 & K pr0-4 g 0.023 O 1.35 (0.225)|°"" (1.88 x 10°) = G.297 OU X . 20 = 2700 Btu/hr~ft ~ F (95.9 MH) = 3290 (125 MW) r R _ e fr -, 1 (3) water = p, h + h ;) L water scal 1 1 =1.25 3755 + 2000) = 0,001087 Overall Heat Transfer Coefficient for Water-Heating Area: (1) RTO tal™ 0.000986 + 0,000485 + 0.001087 = 0,002558 (2) U=_L_ _ m@%g = 390 Btu/hr-£t°- F (95.9 M) Total | = 425 (125 MW) Over-all Heat Transfer Coefficient for Boiling Area: (1) R = 0.000833 4+ 0.000k85 1 0.00Q986:= 0.002304 Total (2) U = 430 Btu/hr-£t°-OF (95.9 M) = 460 (125 M) -230- K. Heat-Transfer Area for Water-Heating Region: 8 (1) apep = ¥ab = 355,030 1b/hr (579 - ¥72) Btu/lb= 0.373 x 10 Btu/hr 8 q 0.373 x 10 Btu/hr Atsalt'“ 3 = . = 8.65°F We T 7.49 x 10° Ib/hr x 0.57 Btu/hr-1o-"F ? (2) at_ = 138.8°F (95.9 M) - 163°F (125 MW) ' 8 __a_ _ 0.373x10 o el (3) A= gat = BE(00y = 00 (5.9 W) 2 Area = 80k £t (125 MW) L., Heat-Transfer Area for Boiling Region: (1) gq= wah = 355,030 (602) = 2.1k x 108 Btu/hr (2) At (761.8 - 572) - (T2L.7 - 572) 1m 1 (189,8 = 1630F (95.9 M) "\139.7 L - = 195 F {125 MW) (3) A= 3 o 218 x‘;bB Btu/hf _ 3080 £ ( i) Uat 430 Btu/hr—ftdnoF x 163°%F 92.9 2 3050 £+ (125 MW) M. Total Steam Generator Heat Transfer Surface: ~ Area = 3040 + 690 = 3730 £t (95.9 MV) A= 3050 + 80k = 3854 ft2 (125 MW) 5 | The area of 3854 £t 1s the design erea and gives a tube length of 10.08 ft, —,231— II. Heat-Transfer Calculations for the Superheater A, Inconel Tube Data: (1) size: 0.5 in. 0.D,; 0.4 in, I.D, (2) Pitch: 0.75 in. delta array (3) Thermal Conductivity = 13.5 Btu/hr-ft-CF (4) Specific heat = 0.133 Btu/lb-F (5) Density = 507 1b/et B. Superheater ‘Inlet Conditions: (1) Molten Salt: (a) T = 1138.1+°F'(95.9 Md): T= 1150°F (125 MW) (b) w= 7.49 x 106 1b/hr (c) c = o'.57 Btu/lb-dF (d) H = 18.5 centipoise (e) e = 123 1b/red i (f) k =24 Btu/hr-ft-oF (2) Stean: | (a) 7= 572F . (b) P = 1250 psia I (¢) h = 1181 Btu/ib (d) w= 263,300 1b/br (95.9 MH); w = 348,000 lb/hr (125 M) C. Superheater Outlet Conditions: (1) Molten Salt: (a) 7= 1120.5°F (95.9 Mi); T = 1126°F (125 MW) (2) Steanm: (a) T 950F (b) P= 1235 psia ~232- (¢) h = 1470 Btu/1v (d) w= 263,300 1b/hr (95.9 MW); w = 348,000 1b/hr (125 MW) (e) vel. = 75.7 ft/sec (95.9 MW); vel, = 100 ft/sec (125 MW) (£) spec. vol.= 0.650 £t°/1b | Required Number of Tubes: (1) Number of Tubes . ¥ X (spec. vol) vel, x area/tube (263,300 1b/hr) x (0.650 £15/1b) (2.72 x 107 f£t/hr) x (8.72 x 10~ £4° /tube ) = T22 Salt Flow Characteristics: 2 2 _ 0.75 _ 0.50 (1) Fléw area/tube = 0.867 35 nip. G_Téé- ' 2 = 0,00203 't (2) Totsl Flow Area = 722 x 0.00203 = 1,465 £2 . » Cvel. = o 7.49 x 10 1b/hr . 11.55 £t ) CA (123 16/£t3) x (1.465 £4°) x 3.6 x 10° sec/hr ?7 e (4) Hydraulic Radius, De = _l% = B__(g%gg:)_) = 0.0622 £% L) ,F(_;i) 12 (5) Re=Y£De _ 1L.55 x o.ofizz x 123 710 6.72 (1077) 18.5 (6) From Figure .7.6; %g = 21 x 1‘-‘r0°1‘L ' 0.k L 2.4 | 2.2 (0.57) (18.5)'] x 21 " 0,622 2.5 . 2 O = 2050 Btu/hr-ft - F R = 0.000487 gsalt -233- ¥, Steam Heat Transfer Coefficient: (1) Average spec. vol., = 0.500 ft3/lb v _x spec, vol. (2,63 x 10 1b/hr) x 0.500 ft3/1b G = Average Vel.= Area = D 722 tubes x (8.72 x 107 £+°/tube G = 211,000 £t/br 0.0266 0.8 0.20 (2) From Ref. 21,}3x x G Xc XM . _(6/12)0'2 4o For these conditions, 0.20 CP XF —_— Ooll'o 0.0266 o.2 0.4 %) 0.8 h= x (211,000) x 0.40 =382 Btu/nr-£t°-F (95.9 Mi) = 478 (125 M) T o 1\ O.E 1 R steam = X (fi')~ 0. (382 ) = 0.00313 G, Heat Transfer Resigétance Through the Tube Wall: v, 1n (zo/7;) _(0,25‘) ( 2 x 32.2 ft/sec? ) = i 1.1 £t l} C. From Appendix 6.2, the head loss in the primary heat exchangers is 23,7 f1. D, Assume the total Kl due to bends, entrances, valves, etc, is approximately 6.0: K Q%é;) 6( (12.} £t/sec)? 2) 2 x 32.2 £t/sec h H i I} b b £t w237 - E. Total Head Losses: H = 18.6 + L.k 4 23.7 + 1h.b - 58.1 £% F. Pumping Power Required: p_ ___ 14.55 1b/br x 58.1 £t (550 ft-1b/sec)/hp x 3.6 x 105 sec/nr = 130 hp V. Temperature Drops énd Maximum Heat Fluxes A, In the Steanm Generator: -g8lt water (1) Atmax== tmax - tsat = 800°F - 5720F = 228°F (2) In "Studies in Boiling Heat Transfer" document No, C00-2k (UCLA 1951) we find the empirical equation for water boiling tubes, (ti - tsat;) 1 loc 7. 123 = 35 logyy (P, ) i} 123 - 35 10g10_(1250 psi)_ 14,5°F (3) In McAdams (ref, 17) equation 1L-7 for water boiling in T T EaéLialélifi w ' Tsat T T p/900 (4) The results from (2) we used as a starting point to get a local tubes is, heat-transfer coefficient for boiling, The over-all local Bo ~238- heat-transfer coefficient vas computed at BOOOF. The coefficiefit, h, for the salt was 1290 Btu/hr-£t-OF and the wall resistance, R, is 0.000497, By a trial and error method thé heat flux, temperatu:e dropy and over- all.heat-transfer coefficient (local) was 752 Btu/hr-fte-oF. Then: (q/A)ma = UAt = (752 Btu/hr-r+2-CF) x 228% X 2 = 172,000 Btu/hr-ft At . 1.9 (172,000) /% 9.6% sat = 12507500 = . . ) | O At 11 = (172,000)(0.000497) = 85.2°F 172,000 o Atsalt = 1290 = 133.2°F Total At = 228°F In the Superheater: (1) Aoy = 1126°F - 572 F = 554°F (2) Steamn: ~ Q.0266 ' X GO,B % }10.2 X c loc T (d/12)0.2 D = 795 Btu/hr-r42-OF (3) R, ;4= 0.000487 (%) R g11 = 0.000343 (5) v 1 propt 2} Uloe = = = 417 Btu/hr-ft - F Total =239~ (6) (q/A)max = UAt - 231,000 (Btu/hr)/fta ‘ | Q (7) At 4. = 231,000 (0.000487) = 112°F O (8) A% o11= 231,000 (0.000343) = BO.F 231,000 (9) &t - = —3~%§-5——— = 36207 (10) Total At = 554°F APPENDIX 8.1 ~240- 3 Group Cross Sections 1200°F BE(KT) = 0.0795 ev 110 Mev —— 0,183 mev Eiiiini Cog 5ie 5a - otr Beryllium 0;768(61) See Ref 1 0.037u(61) :3.517(61) Oxygen 0.322'%%) 0 0.00(61) 2,576(0%) Fluorine o.3h7(6;) Neglected 0.00173 % 3.26]_(61) Sodium o,é61(61) Neglected 0.00021 %% 2.998'61) Nickel =0.1é&(61) 0.743(62) 0,05(61) '3.607(61) Zirconifim 0.135(6$) 0.765(62) 0.00017(63.) 6.15h(61) traniun-235 0.745(61) meglectea 1430 1.207(6Y) 7. 055(61) Boron-10 Neglected Neglected 0,986 %) A chromtn 0.131(%Y) 0.625(2) 0,050(6%) 3.373'6%) 1ron 0,104 0.665(52) 0,050(61) 2.819¢6Y) ML en | EBMr MY Beryitium 1.2200%%) 0.000198(6%) 50616 Oxygen 247'% Negloctea 3.573¢%) Fluorine 37711 " 0,00364' %) 3.5u0(6) Sodiun a97(81) 0,0220(61) 5. 6a9(61) Mickel 580¢6L) o, 1066(61) 16.356(61) Zirconium 1481 0.060(8) 7.08(62) Uranium-235 Neglected 25f19(61) 16.36¢81) o.9p08(61) Boron-10 Neglected 91.9(61) 3.238(61) Chromiun 0.22(6?) Same as Ni 5, 77(62) Iron 0.32%%) sane as mi 8.93(62) Croup 3 = Thermal -2 - KT = 0.0795 ev Atom or Molecule oca op | o Beryllium Oxide 0.0102(34) 9_6(5) Fluorine 0.005(3%) 5.8(84) Nickel 2,3(3“) 19.6(3%) Zirconium 0,09(3“) 6.3(3“) Uranium-235 351.1(3%) 296.5(3) 357,1(3%) Boron-10 2005(34)‘ 2005(34) Chromium 1.45(3h) h.h(Bh) Iron 1_265(3h) 12.2(3h) APPENDIX 8,2 Perturbation Technique The perturbation technique developed below is a'vefy simplified approach in obtaining reactivity changes incurred through small perturbations in cross sections, away from the critical parameters. Upon making diffusion theory approximation to the current J and assuming a2 solution for a bare one region three group system of 2 2 the form (7 + B )é = 0, one obtains the following steady state equations. (DlBl2 + 2g1 *Zrl)‘bl -V (Zpry + Zephy + 2 p5f5) = © 2 . (DB, + Zap *2rp)02 = 2198y = © 2 (D3B3 + Za3)¢l “51'2(#2 = O -2h2- The steady state solution is Zey (_Zra ) ) 1, 2e 21 ] - Za3 is the fraction destroyed by fission. i=3 i=3 T _E ,<0“f> o E: , \ 0% :>>*' Y1 O /. 0 Yi i=1 1 i=1 | where Lpi = fraction of fissions in energy Group 1. 0"f> 0a i 1. <0“f/0“a> i Y4 Ratio of fission to absorption cross section for U-235 averaged over energy group 1. i 1 0.8988 .083 2 0.6495 .639 3 0.8445 .278 o < £ > = 0.7243 o & gpectrum i 21 Therefore one gram burnup of U-235 requires on the average 1.856 x 10 fissions, and one full power hour of reactor operation at 125 MW is -2hh - equivalent to T7.515 grams U-235 destroyed., Inventory of U-235 within the reactor at any time T is expressed as " M(T) = M(o) - 0.007515 ¥ M is kgms T in full power hours., Also the concentration of‘U-235 per cm3 of fuel can be expressed relative to initial concentration as e (T) - E,(O) 1 - 0.00;%i% T ~{ All cross section involving the fuel are written as functions of @ for example: Sa3® {0.00165 + 0.45539} | ° Figsion Product Polsons The additional sbsorption resuliing from non volatile fission products are approximated by the following assumptions. 1 fission = 100 bérns equivalence of thermal polsons 1 fission = 10 barns equivalence of intermediate poisons, Then the added wacroscopic absorption.cross section for the core region are given as: Core AT, o(T) = 566 x 1001 cmt Core - _ ' - dZaE(T) . 566 x10 T — . T in full power hours, The worth in terms of reactivity are calculated as function of T by the perturbation method described in Appendix 8.2, -245- APPENDIX 8.l Prompt Neutron Lifetime The following analysis is & relatively simple method of estimat- ing prompt neutron lifetimes from multigroup constants for an unreflected system. Method in part 1s similar to that presented in Ref. 70, Define: Ti = average time a neutron spends in the'ith energy group. Ni Yzi = Relative number of neutrons existing in the 1 4h group o Ni = Fraction of neutrons born in the 1P group J1i = Average neutron speed of the 3 group, Then: . _ Ti = i Vi 2y < | 2 = + D B, Where :Eti ;Eai'* zgxi i 1 {j:i—ll . Ni = Ni+ i (2x) ; p=1 N3 = 2 ti . J=l Then the prompt neutron lifetlme over all energy groups, k in nunber is: izk ’ %Sm >0 Gl \I_I_ _m.s._.._.. ¢S 1S. #S _ O Mo MM / ool ~ i . H304003Y 001—-O .«\«@ oNILL3S A OOl -M NIVO | Otigee >oo_.+M -V 3¥N9I1d Instantaneous gammas and capbure gammas constitute about 6% more. As a part of the delayed heating is in the gaseous phase, and an exact cal- culation of gamma heating in liquid phase of expansion chamber is not within the scope of this project; an estimate of the liquid phase heating due to beta and gawma energy absorption in the expansion chamber is taken to be 5% of the total Fission heat, divided by the total volume in the expansion chambef.' Thus, 125 x 100 watts x .867 ft3 x .05 = 120 KW 45 £43 In the absence.of flux data for the expanglon chamber, a figure of 130 XKW is assumed for fission heating in the chamber. The ball park estimate is founded ofi information obtained from Mr. Lackey of ORNL, and is derlved from an estimate for the ART, Therefore estimated heat rate in expan31on chamber 1s 150 + 120 + 130 2’ 400 KW, Assuming that fuel is bought ihto chanmber at 12500F and experiences 8 100°F temperature rise before being expelled, 400 KW x 3415 BTU/hr kw = 1,366,000 BTU/hr, 1,366,000 BTU/hr = (.27 BTU/1b °F x 100°F) = 50,600 lb/hr, fuel flow required to remove heat, | If a temperature rise of T5°F is experienced, 67,500 1b/hr will be required. Assuming that inlet tempersture is llTBOF, expansion tank exit tem- perature would be 12750F and average temperature 1225°F with 100°F rise. Use of a helium purge for the system would result in reduction of ans5i0R, chamber by no more than 30%- A, w255 - APPENDIX 13.1 BREAKDOWN OF BASIC REACTOR POWERED SYSTEM COMPONENT WEIGHTS Category A and B (Steam Propulsion Machinery) L. 2. = W oo ~I &N \Wn 10. 11. 12, 13. 1k, - 15. 16, Main propelling units Main shafting Main shaft bearing Iubricating oil system Main condenser and ailr ejector Circulating, condenser, and booster pump Propellers Steam and exhaust piping Water and service piping Insulation and logging Floors, gratings, and adjuncts Auxiliaries Fittings and gears Ligquids Total weight Specific weight Category C and D (Reactor Plant Machinery) 1, Reactor Proper Pressure shell Thermal shield Fuel 130,150 1b 86,480 14,810 19,650 36,040 13,435 18,280 69,580 72,110 21,530 22,400 2,200 12,500 42,510 601,675 17.19 1b/SHP 13,172 4,100 11,365 ~256~ Coolant Heat exchanger Structure and headers Moderator rods and cladding Moderator support structure Control rod -Poison rods and cladding | Nickel shield Miscellaneous (5%.total reactor weight) Total reactor weight Steam Generating System Dry boiler, 2 at 55,000 15 each Salt holdup in boiler, hOOO 1b each Water in boiler, 4000 1b each - Dry superheater, 2 at 9000 1b each Salt in superheater, 2300 1b each Secondary salt plumbing, total Salt in secondary plumbifig, total Steam and salt in lines Salt pfimps, L at 4000 1b each Boiler recirculating pumps, 2 at 6000 1b each Additibnal feed water heating Thermal insulation Total | Additional structural support at 25% of total Total 3,430 12,550 2,185 5,330 80 L, h52 9,510 3,482 69,656 110,000 8,000 8,000 18,000 4,600 3,000 8,000 4,000 16,000 12,000 8,000 4,000 199,600 0,000 2kg, 600 T L. 2, 3. b, -257- Dump Tanks (Primary and Secondary) . Fuel Pumps, 3 at 4000 1b each Miscellaneous (Instruments, additional lines, etc) Total Weight Specific Weight, 431,260/35,000 Category E (Radiation Shielding) Primary Shield Tank inner wall Lead Tank outer wall Water Shield plug Total weight of primary shield Secondary Shield Aft-face Top face Top hat Side faces Forward face Superheater shadow shields Total weight of secondary shield Total Shielding Weight Specific Weight of Shield 30,000 1b 12,000 70,000 431,260 12.32 1b/SHP 13,800 1b 51,000 14,200 117,000 28,140 22k ,1h0 142,620 248,320 13,380 78,900 10,000 10,000 503,320 727,460 20.78 1b/sHP ~258- Category F and G (Blectric Plant) Total weight 210,000 1b Specific weight - - 6.00 1v/sHP Category H and J {Independent Systems) Total weight | | 182,000 1v Specific weight 5.2 1b/SHP Category L (Tools, Equipment, and Spare Parts) Total weight 70,000 1b ~Specific weight o 2,00 1b/SHP Fuel 0il " Total weight [ 0 b Specific weight | | 0 1b/SHP < Total Systefi Weight 2,281,000 1b Specific Weight of Entire Plant ' 63.5 1b/SHP R Ref, No. 10 11 12 13 =259~ BIBLIOGRAPHY Poppendiek, H, F., and Palwer, L, D,, "Forced Convection Heat Transfer Between Parallel Plates and in Annuli with Volume Heat Bources Within the Fluids", ORNL-1701, May 11, 1954, Unclassified. Poppendiek, H. F. and Palmer, L. D., "Forced Convection Heat Transfer in Plpes with Volume Heat Sources Within the Fluids", ORNL 1395, December 17, 1954, Unclassified. Field, F, A., "Temperature Gradient and Thermal Stresses in Heat Generating Bodies", ORNL-CF-54-5-196, May 24, 1954, Confidential. Fraas, A, P., "In Pile Testing of High Temperature Moderating Materials", ORNL CF-56-9-99, September 2k, 1956, Secret, MacPherson, H, G., et al, "A Preliminary Study of Molten Salt Power Reactors", ORNL CF-57-4-27, April 29, 1957, Secret. Weinberg, A. M., et al, "Molten Fluoride Reactors", ORNL CF- 57-6-69, June 1957, Unclassified, Davidson, J. K., Aebert, R. J., Morecroft, B. T., "A Fused- Fluoride Homogeneous Reasctor System for Submarine Propulsion (SAR Phase III Study)", KAPL-992, September 15, 1953, Secret, "DIG Project Progress Report', KAPL-1800~1, February-March, 1957, Secret, Smith, C, O0,, "Notes on Materials Engineering for ORSORT Students", First BEdition, ORNL, 1956-57, Unclassified, Smith, C. 0., "Stress Analysis for ORSORT Students", Second Edition, ORNL, October, 1956, Unclassified. Cooper, W, E., "Modified Structural Design Basis, SAR Reactor Components-I11", KAPL-N1-WEC-9, Rev. 1, May 1, 1957, Unclassified, Fraas, A, P., and Laverne, M, E,, "Heat Exchanger Design Charts", ORNL-1330, December 7, 1952, Secret, Cohen, S, I.,, and Jones, T, N., "Measurement of the Friction Characteristics for Flow in the ART", ORNL-57-3-95, March 19, 1957, Secret. Ref, No, 14 15 16 17 18 19 20 21 22 .23 2k, 25 26 -260- Platus, D. L., and Greenstreet, B, L., "Deflection Equations for Various Loadings of Circular -Arc Curved Beams", ORNL-57,4~96, April 22, 1957, Unclassified, pPotter, P. J., "Steam Power Plants", The Ronald Press Co., New York, 1949, Wllner, B. M., and Stumpf, H., J., "Intermediate Heat Exchanger Test Results", ORNL-CF- 5h 1~155, January 29, 1954, Secret. McAdems, W, H,, "Heat Transmission", 3rd Edition, McGraw- Hill, New York, 1954, Unclassified, ASME "Boiler and Pressure Vessel Code", New York, 1952, Unclassified. Lewié, W. Y., and Robertson, S. A., "The Circulation of Water and Steam in Water Tube Boilers, and the Rational Simplification of Boiler Design', Institution of Mechanical Englneers, Proceedings of March 19&0, London, Unclassified, Glasstone, 8., "Principles of Nuclear Reactor Engineering", Princeton, 1955, Unclassified. The Babcock and Wilcox Co., "Steam, Its Generation and Use", New York, 1955, Unclassified, Davies, R, W,, et al., "ORSORT Reactor Design and Feasibility Study 600 ww Fused Salt Homogeneous Reactor Power Plant", ORNL CF-56-8-208, Oak Ridge, 1956, Secret. Kays, W. M, and Londom, A, L,, "Compact Heat Exchangers", National Press, 1955, Unclassified. Peak, R, D, and Cooper, M. H., Report of Experiment No, 7405-1-1, "Heat Exchanger Evaluation, Type IHE-2, ORNL-1l and 2", 12-15-55, Becret. Peak, R. D., and Cooper, M. H., Report of Experiment No, 7405~1-2, "Heat Transfer and Pressure Drop Correlations, Internediate Heat Exchanger Type IHE-3", ORNL-l and 2, 2-1-56, Secret, Enstice, L. R,, and Hopkins, H. C,, Report of Experiments No, 7405-1-3, "Heat Transfer and Pressure Drop, Intermediate Heat Exchanger Type IHE-3", ORNL-1 and 2, 6~1-56, Secret. 3 Ref, No. e 28 29 30 31 32 33 34 35 36 37 38 39 40 ~261.- Cohen, 8, I., and Jones, T:-N., "Measurement of the Friction Characteristics for Flow in the ART Fuel-to-NsK Heat Exchanger"”, ORNL CF-57-3-95, 3-19-57, Secret, Wantland, J. L., "Thermal Characteristics of the ART Fuel- to-NaK Heat Exchanger", ORNL CF-55-12-120, 12-22-55, Secret. Wantland, J, L,, "Transverse Pressure Difference Across Staggered and Inclined Spacers in the ART Fuel-to-NasK Heat Exchanger", ORNL CF-56-6-143, 6-29-56, Secret. Stone, J. J., and Mann, E. R., "Oak Ridge National Laboratory Reactor Controls Computer”, ORNL 1632, March 1, 1956, Unclasgified, Schultz, M. A., "Control of Nuclear Reactors and Power Plants", - McGraw-Hill, 1955, Unclassified. Nuclear Power Branch of Central Technical Department, "Shielding Notes", Shipbuilding Division, Bethlehem Steel Company, Unclasgified, Rockwell, Theodore III, "Reactor Shielding Manual', TID-T00L, March 1956, Unclassified. Hughes, D, J., and Harvey, J. A,, "Neutron Cross Sections”, BNL-325, July 1, 1955, Unclassified. Blizard, E. P., "Nuclear Radiation Shielding", ORNL, September 17, 1956, Unclassified. Quarterly Progress Report for Period Ending December 31, 1956, "Aircraft Nuclear Propulsion Project", ORNL-2221, March 12, 1957, Secret, Stevenson, R,, "Introduction to Nuclear Engineering", McGraw- Hill, 1954, "Report of the 1953 Summer Shielding Sessions", ORNL-1575, June 1, 195k, Secret. Weir, J, R, Jr., et al, "Inconel as a Structural Material for a High-Temperature Fused Salt Reactor", ORNL 2264, June k4, 1957, Secret,’ Cohen, 8. I,, et al, "A Physical Property Summary for ANP Fluoride Mixes", ORNL-2150, August 23, 1957, Secret, Ref, No. 41 2 43 4y 45 k6 b7 48 k9 50 51 52 53 54 55 56 57 58 29 60 ~262- Cohen, g8, L., ORNL, Personal Communication, July, 1957, DeVan, J. H., ORNL, Personal'Communication, July, 1957, Doney, 1. M., ORNL, Personal Communication, July, 1857, Douglas, D, A., ORNI, Personal Communication, July, 1957, Grimes, W, R., ORNL, Personal Communication, July, 1957, Hikido, 7., ORNL, Personal Communication, July, 1957, Inouye,.H., ORNL, Personsl Communication, July, 1957, Hoffman, E, E., ORNL, Personsl Communication, July, 1957, Lackey, M, E,, ORNL, Personal Communication, July, 1957, Samuels, G., ORNL, Personal Communication, July, 1957, Thowa, R, E., ORNIL, Personal Communication, July, 1957, "Liquid Metals Handbook", Navexos P-733 Rev,, June, 1952, "Metals Reference Book", Interscience Publishers, 1952, Manthos, E, J,, e4 al, "Disassembly ang Examination of the BeO Moderatop Capsule Test, ORNL-lO-h", ORNI, CF-57-3-53, March 13, 1957, Secret, Bolta, C, ¢,, "In.pile Test of Moderstor (BeO) Material ORNL-10.4 " ORNL~CF~57-2-123, February 15, 1957, Secret. Thoma, R, E,, "Reactor Coolant Mixtures", ORNL Intra- Labora tory Memorandum to A, p, Fraas, Dated June T, 1957, Nessle, g, Joy ORNL, Persongl Communication, July 1957, Quarterly Progress Report for Perieg Ending September 10, 1956, "Aircraft Nucleap Propulsion Project”, ORyL 2157, November "Zebra Project Quarterly Project Report", Curtisg Wright Research Department, CWR-470, March 31, 1957, Secret, Bote, R, R., Einstein, I, T., Kinney, w, E., "Description and Operation Manual for the Three Group Three Region Reactor Code for ORACLE", ORNL CF-55-1-76, January 13, 1955, Unclassifie&, Ref, No, 61 62 63 6L 65 66 67 68 69 70 71 72 13 s -263~ "Neutron Cross Sections for Multigroup Reactor Calculations”, Curtiss Wright Research Department, CWR-413, September, 1955, Secret, Deutsch, R, W., "Calculation of the Neutron Age in Hydrogeneous Mixtures", KAPL Memo RWD-13, R. Ramanna, et al, "On the Determination of Diffusion ang Slowing Down Constants of Ordinary Water and Beryllium Oxide Using a Fused Neutron Source", P/872. Dunning, F. 5. and LeDoux, J, C., "Bazard Consideration for a 100 MW Fused Salt Reactor", ORNL CF-57-8-8, to be published, Secret, Sense, Karl A., et al, "Vapor Pressures of the Sodium Fluoride-Zirconium Fluoride System and Derived Information"”, BMI-106k, Unclassified. Palmer, L. D., "A Preliminary Analysis of the Temperature Structure Within a Solid Moderator Rod, Cylindrical Reactor", ORNL CF-57-4-138, April, 1957, Confidential. ' Quarterly Progress Report for Period Ending June 30, 1957, "Aircraft Nuclear Propulsion Project", ORNL~2340, To be published, Secret, ' Stehn, R. J., and Clancy, E. F., "General Electric Chart of the Nuclides", April, 1956. ' Walker, C. 8., "Reactor Control", ORNL CF-57-1-1, January 5, 1957. Holmes, D, K., "Calculation of Average Lifetime of Neutrons Using the Results of Multigroup Calculation”, Y-F10-15, October 2, 1950, Geortzel, G., "An Estimation of Doppler Effect in Intermediate and Fast Nuclear Reactors”, P/613. Kinyon, B, W., ORNL, Personal Communication, July, 1957. Schroeder, J, H., et al., "A Preliminary Design of a Unit Shield", ORSORT Study Group, ORNL CF-58-813, To be published, Secret, Dee, J. R. and Woodsun, H, C., "An Analysis of F, P. -Ray Spectrum”, Volume 4 and Document No, NARF-56-41T, FZK-9-109, 1956,