OAK RIDGE NATIONAL LABORATORY operated by UNION CARBIDE CORPORATION for the U.S. ATOMIC ENERGY COMMISSION ORNL- TM- 732 /73 MSRE DESIGN AND OPERATIONS REPORT Part V REACTOR SAFETY ANALYSIS REPORT . E. Bedll . N. Haubenreich . B. Lindauer . R. Tallackson — A 9w NOTICE This document contains information of a preliminary nature and was prepared primarily for internal use at the Oak Ridge National Laboratory. it is subject to revision or correction and therafore does not represent a final report. The information is not to be abstracted, reprinted or otherwise given public dis- semination without the approval of the ORNL patent branch, Legal and Infor- mation Control Department. LEGAL NOTICE This report was prepared as an account of Government sponsored work. Neither the United States, nor the Commission, nor any persen acting on behalf of the Commission: A. Makes any warranty or representation, expressed or implied, with respect to the accuracy, completeness, or usefulness of the information contained in this report, or that the use of any information, cpparatus, method, or process disclesed in this report may not infringe privately owned rights; or B. Assumes any liobilities with respect to the use of, or for damages resulting from the use of any information, apporatus, method, or process disclosed in this report, As used in the above, '‘person acting on behalf of the Commission”” includes any employes or contractor of the Commission, or employee of such contractor, to the extent that such employese or contractor of the Commission, or smployee of such contracter prepares, disseminates, or provides access to, any information pursuant to his employment or contract with the Commission, or his employment with such contractor. ORNL-TM-732 Contract No. W=-7405-eng-26 Reactor Division MSRE DESIGN AND OPERATIQNS REPORT Part V REACTOR SAFETY ANALYSIS REPCRT S. BE. Beall R. B. Lindauer P. N. Haubenreich J. R. Tallackson AUGUST 1964 OAK RIDGE NATTIONAL TABORATORY Oak Ridge, Tennessee operated by UNLON CARBIDE CORPORATION for the U.5. ATOMIC ENERGY COMMISSION iid PREFACE This report is one of a series that describes the design and opera- tion of the Molten-Salt Reactor Experiment. All the reports are listed below. ORNL-TM-"728 ORNL~TM-729 ORNL~-TM~730% ORNL-TM-731 ORNL-TM~732% ORNL-TM-'733 CRNL-TM-907%%* ORNL-TM-908%* ORNL-TM-209%* ORNL-TM-910%* *Issued. **¥These reports MSRE Desgsign and Operations Report, Part I, Degcription of Reactor Design, by R. C. Robertson MSRE Design and Operations Report, Part II, Nuclear and Process Instrumentation, by J. R. Tallacksocon MSRE Design and Operations Report, Part III, Nuclear Analysis, by P. N. Haubenreich and J. R. Engel, B. E. Prince, and H. C. Claiborne MSRE Design and Operations Report, Part 1V, Chemistry and Materials, by F. F. Blankenship and A, Taboada MSRE Design and Operations Report, Part V, Reactor Safety Analysis Report, by S. E. Beall, P. N, Haubenreich, R. B. Lindauer, and J. R, Tallackson MSRE Design and Operations Report, Part VI, Operating Limits, by 5. E. Beall and R. H. Guymon MSRE Desilgn and Operations Report, Part VII, Fuel Handling and Processing Plant, by R. B. Lindauer MSRE Design and Operations Report, Part VIILL, Operating Procedures, by R. H. Guymon MSRE Design and Operations Report, Part IX, Safety Procedures and Emergency Plans, by R. H. Guymon MSRE Design and Operations Report, Part X, Maintenance Equipment and Procedures, by E. C. Hise and R, Blumberg will be the last in the series to be published. iv ORNL-TM-911%** MSRE Design and Operations Report, Part XTI, Test Program, by R. H. Guymon and P. N. Haubenreich *¥ MSRE Desgsign and Operations Report, Part XIT, Lists: Drawings, Specifications, Line Schedules, Instrument Tabulations (Vol. 1 and 2) 1. 2. PART REACTOR SYSTEM 1.1 Fuel and Primary System Materials Fuel and Coolant Balts ..., v, Structural Material — INOR=8 ... .iiiiiiinnnnnn 1.1.1 1.1.2 1.1.3 1.1.4 1.2 Bystem L.2.1 1.2. 1.2. 1.2. 1.2. 1.2. 1.2. 1.2. 1.2. 1.2.10 1.2.11 O 8 -3 &0 U M~ WM CONTROLS AND INSTRUMENTATION 2.1 Control Rods and Rod Drives 2.2 Safety 2.2.1 2.2.2 2.2.3 2.2.4 CONTENTS -------------------------------------------------------- --------------------------------------------------- 1. DESCRIPTION OF PLANT AND OPERATING PLAN Moderator Material — Graphite .................. Compatibility of Salt, Graphite, and INCR-8 .... COMPONENTS vttt ittt et cn st e it antnranennsnnanss Reactor Vessel ...ttt iiiiiiiinannnns Fuel and Coolant Pumps ..veeiiniienianennnacasns Primary Heat Exchanger .......c.iviiiiintineocanns Salt-to~Air Radiator ....... i, Drain and Storage Tanks .ot iiitienotssensennns Piping and Flanges ...u.eitiiieiiniieniiinanasarass Freeze Valves ... .ottt iiiiiinneanns Cover-Gas Supply and Disposal ... eeivinvinrnerss Sampler-FEnricher ... .ttt riteiecesstesssasns Electric Heaters ..i. vt ittt iassosnnnes Tigquid Waste System .. vv ettt ieiinnennans TN trumMe Nt a iy v vttt s s e vt ot oresennasseconossoss Nuclear Safety System ... en i Temperature Instrumentation for Safety System Inpuls ovivti ittt it i ittt e s Radiator Door Emergency Closure System ......... Reactor Fill and Drain System ........covveee... --------------------------------------------- ooooooooooooooooooooo ------------------------------- Page iii O ~3 ~1 W 10 12 15 16 19 23 25 28 33 33 34 39 41 43 45 47 55 58 68 70 71 vi 2.2.5 Helium Pressure Measurements in the Fuel = T P I T o 2.2.6 Afterheat Removal System .....ccuiiiitiiineniann 2.2.7 Containment System Instrumentation ............. 2.2.8 Health-Physics Radiation Monitoring ............ 2.3 Control Instrumentation .......... i, 2.3.1 Nuclear Instrumentation .......civivieeieinnnne, 2.3.2 Plant Conbrol vueirvirtiiiinnrinrrscnonnsnroanss 2.4 Neutron Source Considerationsg .....veveverierereeeeenen. 2.5 Electrical Power System ....vviiiii it rsensenscnons 2.6 Control Room and Plant Instrumentation Layout ......... 2.6.1 Main Control Area ......viiiiiiiirrerininnnenens 2.6.2 Auxiliary Control Area .....viieitereonennneeanns 2.6.3 Transmitter Room ......couiiiiiiiiniiieiininnn, 2.06.4 TField Panels tueeeiet oo nsnesnscsensnssssons 2.6,5 Interconnections .......c.iiiiiinieerinnnnncnnns 2.6.6 Data ROOM v.ivviir it inirieeentnsensneessnsnnens PLAN D LAY QU it ittt it ettt s enonnnessennsnnssesesensnsas 3.1 Fguipment Arrangement . ......c.eeueniernneeeseroneonnasens 3.2 Biological ShieldiIng . .vveiirinrinetnerorneaneenennnnns ST E FRATURE S ittt ittt ittt iie i tinnsrisnassonaananens ol LOCATION vttt i i e e et e e e e e 4.2 Population Density . .vviii ittt it i e 4.3 Geophysical FeatUres .uuiieiiiie it iieernenenennaneanns b.3.] MetEOrOlOEy vttt e tvat et e 3.2 Temperature .. ...ttt ittt ittt 4.3.3 Precipitation ...iieiiiiiiiiiiii it L O i o P 4.3.5 Atmospheric Diffusion Characteristics .......... 4,3.6 Environmmental Radicactivity ........ceiiiiiien... 4.3.7 Geology and HydroloZY cueeieeiveecorsosnnoassnnas e 3.8 el oMol ettt it et et et e 77 79 g1 92 96 96 104 117 118 119 119 119 123 124 124 125 128 128 132 138 138 138 146 146 146 147 149 157 158 158 164 vii CONSTRUCTION, STARTUP, AND OPERATION ... vivnernrvenneonnns 168 5.1 Construction ittt it it it et e 168 5.2 Flush-Salt Operation .v.eeieiiieeien ittt rnnnienannnnss 168 5.2.1 Critical Experiments ..v.ieeiiiiinreninnnennnanns 170 5.2.2 Power Operation .....ieieeintiennrnenesnaenannes 170 5.3 Operatlons Personnel cu.iieiiineersnrneesoersooesnsnsns 171 5.4 Maintenante .u.euireeier it iteneeeeetaoestassansesoesans 173 PART 2. GSAFETY ANALYSES O AT M T ettt tee et ennsseesasennereonnsosnnsanssnnnsesns 177 6.1 General Design Considerations ....c.oivieernerrnnnrnnnns 177 6.1.1 Reactor Cell Design ..v.vieriiteinreerenonnoneens 178 6.1.2 Drain Tank Cell DESigll .vv vt ietnnienereaneenns 182 6.1.3 Penetrations and Methods of Sealing ............ 184 6.1.4 Leak Testing voveirrrereerioanernoosonronanossnn 187 6.2 Vapor-Condensing System ... eiiiiertoeroennsoennnonas 187 6.3 Containment Ventilation System ....vven it iniennreas 191 DAMAGE TO PRIMARY CONTAINMENT SYSTEM ..vveerniiinriinnnnnnns 196 7.1 Nuclear Tncidents ...ttt innnsioneenns 196 7.1.1 General Considerations in Reactivity Incidents .. ittt it ittt et e 196 7.1.2 Uncontrolled Rod Withdrawal ......vivivennveann. 199 7.1.3 "Cold-Slug" Accident ....ieiiiiiiiiieiiiieaaenn, 203 7.1.4 Filling Accidents .o iiiiirninninienernsceranns 205 7.1.5 Fuel AdAitions ...c.iiiiiin i initneiennnnennes 213 7.1.6 U0y Precipitation ....coiiiiiiiinienineninnecas 214 7.1.7 Graphite Loss or Permeation ...........ccvu... 219 7.1.8 Loss Of FlOW ittt iiesiinnnnensnaesnonnnnnns 221 7.1.9 Loss OFf Load «uviivinnnrrnnnrentensanonasnnnsenns 225 7.1.10 Afterheat ... ittt iiiinierionnsarassennenns 226 7.1.11 Criticality in the Drain Tanks ....eceveensrens 230 7.2 Nonnuclear ITncidents ....iiieiiiiiriiniinriinionaereano. 231 7.2.1 Freeze-Valve Fallure ........o.uieiitennvennnnans 231 7.2.2 TFreeze-Flange Failure .......ceiueevenenonnne nn 231 7.2.3 Excessive Wall Temperatures and Stresses ....... 7.2.4 COrrOSIOn tuiiiriieeriereeeroeettonssneeeanennns 7.2.5 Material Surveillance Testing ...... et aeeaa 7.3 Detection of Salt Spillage ....... et cen 7.4 Most Probable Accident .............. et 8. DAMAGE TO THE SECONDARY CONTATINER . .vivivinninrnernnnnnnennn 8.1 Missile Damage .....veeeertrennrennonnenas et 8.2 EXCeSSIVE PressUrt v vivevrnrnnrnnrnearesneetoeeoneeenns 8.2.1 Salt Spillage .....iviiriinrnnnnnan e 8.2.2 0il Line Rupture ......... ..., C et 8.3 Acte of Hature ..ot et tie i .o B8.3.1 EBarthnguUaKe vttt et instneonsenanoeanonennas S T o o T S TN Loy v == 8.5 Corrosion from Spilled Salt v.iiiiieiirrnennnens e 8.6 Maximum Credible Accident ......ovvvvinnn... e e 8.7 Release of Radicactivity from Secondary Container ..... g.7.1 Rupture of Secondary Container ................ . 8.7.2 Release of Activity After Maximum Credible Accident ... e s i s 8.2 Release of Beryllium from Secondary Container ......... Appendix A. Calculations of Activity Levels ......eeiivinnne... Appendix B. Process Flowsheels . .i. ittt i ieenneennennens Appendix C. Component Develcpment Program in Support of The MORE ittt i i i i it et i e Appendix D. Calculaticn of Activity Concentrations Resulting from Most Likelyv Accident ...t ie ittt it inersnnn Appendlx E. Time Required for Pressure in Containment Vessel viii To Be Lowered to Atmospheric Pressure ............. 232 233 235 236 236 238 238 238 238 239 239 239 239 240 240 240 245 245 Fig. 1.1 Fig. 1.2 Fig. 1.3 Fig. 1.4 Fig. 1.5 Fig. 1.6 Fig. 1.7 Fig., 1.8 Fig. 1.9 Fig. 1.10 FPig. 1.11 Pig. 1.12 Fig. 1.13 Fig. 1.14 Fig. 1.15 Fig. 1.16 Fig. 1.1%7 Pig. 1.18 Fig. 2.1 Fig. 2.2 Fig. 2.3 Fig. 2.4 Fig. 2.5 Fig. 2.6 Fig. 2.7 Fig. 2.8 Fig. 2.9 Fig. 2.10 ix LIST OF FIGURES Fuel and Coolant Flow Diagram MSRE Graphite Showing Cracks Resulting from Impregna- tion and Baking Operations MSRE Layout Reactor Vessel and Access Nozzle Typical Graphite Stringer Arrangement Lattice Arrangement at Control Rods Fuel-Circulation Pump Primarj Heat Exchanger Salt-to-Air Radiator Fuel Salt Drain Tank Bayonet Cooling Thimble for Fuel Drain Tanks Freeze Flange Freeze Valve Cover-Gas System Flow Diagram Off'-Gas System Flow Diagram MSRE Sampler-Enricher Single-Line Diagram of MSRE Power System Simplified Flowsheet of MSRE Liquid Waste System Control Rod Drive Unit Installed in Reactor Diagram of Control Rod Electromechanical Diagram of Control Rod Drive Train Reactivity Worth of Control Rod as a Function of Depth of Insertion in Core Control Rod Height Versus Time During a Scram Control Red Shock Absorber Control Rod Thimble Prototype Control Rod Drive Assembly Functional Diagram of Safety System Block Disgram of Safety Instrumentation for Control Rod Scram Page 13 15 17 18 18 20 23 26 29 31 34 35 37 38 39 42 48 49 50 51 52 54 56 57 59 60 Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. . 2.20 Fig. Fig. Fig. Fig. Fig. Fig. Fig, Fig. Fig. Fig. . 2.31 Fig. Fig. Fig. Fig. Fig. Fig. 2.11 2.12 2.13 2.14 2.15 2.16 2.17 2.18 2.19 2.21 2.22 2.23 2 .24 2.25 2.206 2.27 2.28 2.29 2.30 2.32 2.33 2 .34 2.35 2.36 2.37 Typical Temperature-Measuring Channel Used in Safety System Radiator Door Emergency Closure System Pump-Speed Monitoring System Reactor Fill and Drain System Valving Typical Instrumentation for Measuring Helium Pressures in the Primary Loop Afterheat Removal System Helium Supply Block Valving Off-Gas System Instrumentation and Valving Lube 0il System Off-Gas Monitors In-Cell Liquid Waste and Instrument Air Block Valving Pressure Switch Matrix Used with Instrument Air Line Block Valves In-Cell Cooling Watér System Block Valving Reactor Building (7503) at 852-ft Elevation Showing Locations of Monitors Reasctor Building (7503) at 840-ft Elevation Showing Locations of Monitors Typical Low-Level BFs Counting Channel for Initial Critical Testus Locations of BF3 Chanbers Wide-Range Ccunting Channel Linear Power Channels and Automatic Rod Controller Nuclear Instrumentaticn Penetration Computer Diagran for Servo-Controller Simulation Results of Analog Simulatiocn of Sysftem Response to Step Changes in Power Demand with Reactor on Automa- tic Temperature Servo Control Results of Analog Simuistion of Reactor Response to Ramped Changes in Outlet Temperature Set Point with Reactor Under Automatic Control Simplified Master Plant Control Block Diagram Simplified Rod Control Block Diagram Regulating Rod Limit Switch Assembly Regulating Rod Contrcl Circuit Diagram of Safety System Bypassing with Jumper Board €9 72 73 74 78 g0 83 84 85 88 89 91 93 94 96 07 ) 99 190 103 105 105 106 109 110 111 116 Fig. Pig. Pig. Fig., Fig. Fig. Fig. Fig. Fig. 'ig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Fig. Pig. Fig. Fig. N~ MMM DD W W WD NNDNNN NN DD oo W W I O O O O O W oo e xi Main Floor Layout of Building 7503 at 852-ft Elevation Main and Auxiliary Control Areas Main Control Board Layout of Building 7503 at 840-ft Elevation Transmitter Room Layout of Data Room Typical Process Computer System (TRW-340) First Floor Plan of Regctor Building Elevation Drawing of Reactor Building Arrangement of Shielding Blocks on Top of Reactor Area Surrounding MSRE Site Contour Map of Area Surrounding MSRE Site Map of Oak Ridge Area beasonal Temperature Gradient Frequency Annval Frequency Distribution of Winds in the Vicinity of X-10 Aresa X-10 Ares Seasonal Wind Roses X-10 Area Seasonal Wind Roses Wind Roses at Knoxville and Nashville for Various Altitudes Reactor Division, Operations Department Organization for the MSRE Reactor Cell Model Drain Tank Cell Model Typical Flectric Lead Penetration of Reactor Cell Wall Vapor-Condensing System Power and Temperature Transients Produced by Uncon- trolled Rod Withdrawal in Reactor Operating with Fuel B Power and Temperature Transients Produced by Uncon- trolled Rod Withdrawal in Reactor Operating with Fuel C Effect of Dropping Two Control Rods at 15 Mw During Uncontrolled Rod Withdrawal in Reactor Operating with Fuel C Power and Temperature Transients Following InJjection of Fuel B at 900°F into Core at 1200°F; No Corrective Action Taken System Used in Filling Fuel Loop 120 120 121 122 123 126 127 122 130 134 139 140 141 148 150 152 153 156 172 179 183 185 190 201 202 204 206 209 Fig. Fig. Fig. Fig. Fig. Fig, Fig. Fig. Fig. Fig. 8 Fig. Fig. Pig., Fig. Fig. Fig. Fig. 3 .10 A1 .12 13 Xii Net Reactivity Addition During Most Severe Filling Accident Power and Temperature Transients Following Most Severe Filling Accident Effects of Deposited Uranium on Afterheat, Graphite Temperature, and Core Reactivity Power and Temperatures Following Fuel Pump Failure, with No Corrective Action Power and Temperastures Following Fuel Pump Power Failure. Radiator doors closed and control rods driven in after failure. Bffects of Afterheat in Reactor Vessel Filled with Fuel Salt After Operation for 1000 hr at 10 Mw Temperature Rise of Fuel in Drain Tank Beginning 15 min After Reactor Operation for 1000 hr at 10 Mw Heatlng of Reactor Vessel Lower Head by UOp Deposits with Power Level at 10 Mw Release of Fuel Through Severed Lines Relationship Between Cell Pressure and Weights of Fluids Equilibrated Radiation Level in Bullding Following Maximum Credible Accident (r/hr = 935 x upc/cm® x Mev) Activity Concentrations in Building Air Following Maxi- mum Credible Accident Noble Gas Activity After Maximum Credible Accildent as Function of Distance Downwind Change in Rave of Activity Release from Building with Time Total Integrated Doses Following Maximum Credible Accident Peak Todine and Solids Activities After Maximum Credible Acclident as Function of Distance Downwind Beryllium Contaminaticn After Maximum Credible Acci- dent as Functicon of Distance Downwind 211 212 218 222 223 225 227 229 2472 243 2417 248 249 251 252 MSRE DESIGN AND OPERATIONS REPORT Part V REACTOR SAFETY ANALYSIS REPORT S. E. Beall R. B. Lindauer P. N. Haubenreich J. R. Tallackson INTRODUCTION The Oak Ridge Natilonal Laboratory undertoock, in 1951, the develop- ment of a molten-salt-fueled reactor for the Aircraft Nuclear Propulsion Program, and the Aircraft Reactor Experiment (ARE) was successfully op- eragted, in 1954, to demonstrate the nuclear feasibility of a system in which molten-salt fuel was circulated.? Development work on molten-salt- fueled systems has been continued, with the ultimate objective of design- ing a molten-salt reactor capable of breeding. The Molten-Salt Reactor Experiment (MSRE)?,? is presently being as- sembled as an engineering demonstration of a single-region, nonbreeding salt-fueled reactor. The goals in the development of this system have been to show that this advanced-concept reactor is practical with present- day knowledge, that the system can be operated safely and reliably, and that is can be serviced without unusual difficulty. It is the purpose of this report to describe the MSRE (Part I) and to evaluate the hazards associated with this reactor (Part II). W. B. Cottrell et al., "Operation of the Aircraft Reactor Experi- ment," USAEC Report ORNL-1845, Oak Ridge National Laboratory, August 1955, ?0ak Ridge National Laboratory, "Molten Salt Reactor Experiment Preliminary Hazards Report,'" USAEC Report ORNL CF 61-2-46, Feb. 28, 1961; Addendum, Aug. 14, 1961; Addendum No. 2, May 8, 1962. 3A. L. Boch, E. S. Bettis, and W. B. McDonald, ."The Molten Salt Re- actor Experiment," paper prepared for presentation at the International Atomic Fnergy Agency Symposium, Vienna, Austria, October 23-27, 1961. PART 1. DESCRIPTION OF PLANT AND OPERATING PLAN 1. REACTOR SYSTEM The Molten-5alt Reactor Experiment comprises a circulating-fuel re- actor designéd for a heat generation rate of 10 Mw and the auxiliary equipment necessary for 1ts operation. The fuel circuit and the heat re- moval, or coolant, circuit are illustrated schematically in Fig. 1.1,% which also gives some standard operating conditions for the circuits. The fuel salt, a mixture of LiF, BeF., ZrF,, and UF,, is pumped through a cylindrical reactor vessel filled with graphite blocks. At the 10-Mw power level, the fuel enters the vessel at 1175°F and is heated to 1225°F. It then flows to a 1200-gpm sump-type centrifugal pump, which discharges through the shell side of the fuel-to-coolant heat exchanger. The fuel returns from the heat exchanger to the reactor inlet. The coolant salt, a mixture of LiF and BeF,;, is heated from 1025 to 1100°F as it passes through the tubes of the fuel-to-coolant heat ex- changer. Following dissipation of the heat in an air-cocled radiator, the coolant salt is returned to the heat exchanger at a flow rate of 850 gpm by a second sump-type centrifugal pump. Electric heaters on the piping and equipment keep the salt mixtures above their melting points. In order to assure that the salt mixtures remain molten, the normal electric utility supply is augmented by an emergency supply from the diesel generators. The molten salts may be drained from the reactor system within a few minutes. The fission gases xenon and krypton are stripped from the fuel salt continuously in the gas space above the liquid surface within the fuel pump bowl. A continuous flow (~3 liters/min) of helium removes the fis- sion gases from the pump bowl and carries them to activated-carbon beds outside the reactor cell. Helium is also used as the cover gas through- out the fuel and coclant systems. The components of the fuel and coolant systems are constructed of INOR-8, a nickel-molybdenum~chromium alloy developed especially to con- tain molten fluoride salts. In the fuel system the piping to the *Detailed flowsheets for all the reactor systems are presented in Appendix B. UNCLASSIFIED ORNL-LR-DWG. 56870R1 COOLANT PUMP 850 gpm ] I II FREEZE ” FLANGE H “ H 1200 gpm ” J COOEEEI ]] I ’ “ REACTOR Il J‘ I I VESSEL I i |[ REACTOR CELL “ RADIATOR H ! w H e = e — e — — — | 300°F n AIR LT . - —— 200,000 cfm I oraIN - 7[ 100°F I i | , I I ) | | ) COOLANT DRAIN TANK (44 cu f1) SPARE FILL AND FLUSH FILL AND DRAIN TANK TANK DRAIN TANK (73 cu ft) {73 cu ft) (73 cu ft) Fig. 1.1. Fuel and Coolant Flow Diagram. components is connected by "freeze" flanges which utilize frozen salt as the sealant. These flanges are also provided with gas-buffered ring- Jjoint gaskets. The use of freeze-flange joints facilitates removal and replacement of radiocactive equipment after power operation. Joints of this type are not needed in the coclant system, because the radiocactivity there is low enough to permit direct maintenance within a few minutes after shutdown. No valves of the ordinary type are used in contact with salt. Flow in lines connecting the reactor vessel and the drain tanks is prevented by freezing salt in designated sections of pipe. The "freeze valves" thus formed can be thawed by stopping the flow of cooling alr and heating the pipe. By this means, salt can be drained from the reactor vessel,. In addition to the fuel and coolant circuits, there are auxiliary components, such as controls and instrumentation, salt-sampling equipment, facilities for handling radioactive liquid wastes, a chemical processing cell for purification of the salts, and remote-maintenance equipment. Details of the reactor components, auxiliary equipment, and instrumenta- tion are discussed in the following sections. The control problems of a molten-salt reactor differ in many respects from those of solid fuel reactors‘(see sec. 1.3). The reactor core is provided with three control rods that will be used principally to main- tain the fuel salt temperature within the operating limits. The reactor power level will be controlled by regulating the rate of heat removal at the radiator. 1.1 Fuel and Primary System Materials 1.1.1 Fuel and Coclant Salts The MSRE fuel will be a solution of U??°F, and ZrF, in a molten Li7F-BeF, solvent. Thorium tetrafluoride may be added as a fertile material. Both Li’F and BeFy have relatively good neutron cross sections; the ratio of these materials was chosen to provide the optimum compromise among freezing point, fluid flow properties, and heat transfer character- istics of the final mixtures in which UF, and ZrF, concentrations are firmly fixed. The ZrF, is included in the fuel mixture because oxygenated species (i.e., Ho0) capable of yielding oxide ions on contact with the fuel precipitate ZrO, from LiF-BelFj,-~-ZrF,-UF, solutions whose Zrt ato-Utt ratio exceeds 3; U0y precipitates if oxide ion is admitted to solutions in which the ratio is appreciably below that value. To prevent precipi- tation of U0z upon inadvertent contamination of the reactor system, the Zr**-to-U** ratio is fixed at a conservative value (>5:1). Moreover, since precipitation of ZrO; is also undesirable, the fuel mixture is pro- tected at all times from gases and vapors bearing oxygenated species by using helium as a blanket gas. The compositions and properties of three fuel salts are listed in Table 1.1. EFach of these salt mixtures is expected to be employed at some stage of the experiment. Salt C, the partially enriched salt, will be loaded for the first series of experiments because its chemical char- acteristics have been studied more thoroughly than those of salts A and B. Salt B, the highly enriched salt, is the composition which is pro- posed for the core of s large two-region breeder reactor (or a U23° burner). It will be used after a long period of operation with the par- tially enriched salt. Salt A contains thorium and has been proposed for single-region molten-salt breeder reactors. It will be tested in a third series of MSRE experiments. The coolant salt is a mixture of Li”F and BeF, whose composition and properties are included in Table 1.1. The coolant mixture can withstand higher oxide contaminant levels than the fuel before an oxide (BeOQ in this case) precipitates, but the same practice of blanketing the mixture with helium will be folilowed. Another salt mixture of essentially the same composition ag tThe coolant salt, but called the flush salt, will be used to rinse the fuel Teble 1.1. Compositions and Physical Properties of the Fuel, Flusk, and Coolant Salts uel Salt B, Fuel Salt C, F?el Salu-A, with 02% with 35% Flush and with Thorium . . Coolant and Uraniunm Enrlghed Enrlghed Salt Uranium Uranium Composition, mole % LiF (99.99+% Li") 70 66. 8 65 66 BeF», R2. 6 29 29.1 34 ZrFy 5 %4 5 0 ThF, 1 0 0 0 UF, 0.4 0.2 0.9 0 Physical properties at 1200°F Density, 1b/ft? 140 130 13 120 Viscosity, 1b/ft-hr 18 17 20 248 Heat capacity, Btu/lb.°F 0.45 0.4 0.47 0.53 Thermal conductivity, 3.2 3.2 3.2 3.58 Btu/hr- 2 (*F/ft) Ligquidus temperature, °F 840 840 840 850 %At 1060°F. system of fuel salt and fission-product residues after shutdown and before maintenance is begun. If the maintenance operations allow the inner sur- faces of the fuel system to be exposed to air, the flush salt will also be circulated after maintenance as an 0p-HpO cleanup measure. This should protect the fuel salt from oxygen contamination. However, if any of the salt mixtures become excessively contaminated with oxygen, they will be purified by treatment with an Hy-HF mixture (see sec. 2). 1.1.2 Structural Material — INCR-8 The principal material of construction for the reactor system is INOR-8, a nickel-molybdenum-chromium alloy developed at the Oak Ridge National Laboratory for use with fluoride salts at high temperature.l The composition and properties of INOR-8 are listed in Tables 1.2, 1.3, Table 1.2. Composition of INCR-8 . Quantity . Quantity Constituent (vt %) Constituent (vt %) i 66~71. Mn 1.0 (max) Mo 15-18 Si 1.0 {(max) Cr 6-8 Cu 0.35 (max) Fe 5 (max) B 0.010 (max) C 0.04-0.08 W 0.50 (max) Ti + Al 0.50 (max) P 0.015 (max) S 0.02 (max) Co 0.20 (max) and 1l.4. When INOR-8 is corrosively attacked, chromium is leached from it and tiny subsurface voids are formed. The rate of attack is governed by the rate of diffusion of chromium in the alloy. Measured rates of attack in typical fuel and coolant salts for many thousands of hours have been less than 1 mil/yr. Furthermore, no greater attack has been observed in several thousand hours of in-pile tests (see sec. 7.2.4). The perfor- mance of INOR-8 in the MSRE will be followed by removing surveillance IR, W. Swindeman, "The Mechanical Properties of INOR-8," USAEC Re- port ORNL-2780, Oak Ridge National Laboratory, January 1961. 10 Table 1.3. Physical Properties of INOR-8 Density, 1b/in.> 0.320 Melting point, °F 2470-2555 Thermal conductivity, Btu/hr-ft? (°F/ft) 12.7 at 1300°F Modulus of elasticity at ~1300°F, psi 24.8 x 10° Specific heat, Btu/lb-°F at 1300°F 0.138 Mean coefficient of thermal expansion in 8.0 x 107° 70-1300°F range, in./in.-°F Table 1.4. Mechanical Properties of INOR-8 Ultimate Yield Maximum Temperature Tensile Strength at Allowable (°F) Strength 0.2% Offset Stress? (psi) (psi) (psi) ok 82,000 26,200 17,000 1100 77,000 25,500 13,000 1280 62,500 24,200 6,000 1300 58,000 24,200 3,500 aASME Boiler and Pressure Vessel Code Case 131¢%. specimens, at 3- to 6-rmonth intervals, from the sample assembly located in the spare control rod position of the reactor vessel (see sec. 1.2.1). 1.1.3 Modergtor Material — Graphite Although the salt has moderating properties, the use of a separate moderator has the advantage of reducing the reactor fuel inventory. Un- clad graphite (see Table 1.5 for properties) was chosen as the MSRE moder- ator to avoid the problems of cladding, such as neutron losses and de- velopment of cladding technigues, and because no serious problems were foreseen with the bare graphite. At The time the decision was made, graphite that could meet the requirements cf high density and low per- meability to salt had been manufactured on an experimental basis by the 11 Table 1.5. Properties of MSRE Core Graphite, Grade CCB Physical prcperties Bulk density,® g/cm? Range Average Porosity,? % Accessible Inaccessible Total Thermal conductivity,® Btu/ft.hr.°F With grain At 90°F At 1200°F Nermal to grain At 90°F At 1200°F Temperature coefficient of expansion, With grain, at G8°F Normal to grain, at 68°F gpecific heat,d Btu/lb. °F At O°F At 200°F At 600°F At 1000°F At 1200°F Matrix ceoefficient of permeability to helium at 70°F,€ cm? /sec Salt absorption at 150 peig,? vol % Mechsnical strength at 68 °FD C OF—l Tensile strength, psi With grain Range Average Wormal to grain Range Average Flexural gstrength, psi With grain Range Average Normal to grain Range Average Modulus of elasticity, psi With grain Normal to grain Compressive strength,d psi 1.82-1,¢87 1.86 .49 x 10-° .8 x 10-° N O .14 .22 .33 .39 A2 % 104 WOOOOO0o O .20 1500=6200 1200 1100-4500 1400 3000-5000 4300 2200~3650 3400 3.2 x 106 1.0 x 108 8600 "Messurements made by the Oak Ridge National Laboratory. b and the Oak Ridge National Laboratory. Based on measurements made by the Carbon Products Division cMéasurements made by the Carbon Products Division. d Representative data from the Carbon Products Division. e Based on measurements made by the Carbon Products Division on pilot-production MSRE graphite, 12 National Carbon Company. Since it appeared that bars 2 in. by 2 in. could be processed without difficulty, this size was selected as the basic ele- ment for the MBRE core assembly. During the manufacture of the MSRE graphite, it was learned that scme cracking resulted from the impregnation and baking operations and that the density was not quite as high as had been expected. The cracks raised guestions about the strength and the permeability. The cracks in some of the poorest material are shown in Fig. 1.2. After extensive examing- tion and testing2 at the Oak Ridge National Laboratory, it was determined that the strength was as great or greater than specified, that the re- duction in density was inconsequential, that the cracks could not be propagated by mechanical forces or thermal stresses far in excess of con- ditions which will exist in the reactor, and that the total penetration of salt in the cracked graphite would be considerably below the specified penetration (0.5% at 150 psig). Furthermore, two to three months of in- pile testing of AGOT-grade graphite (a more permeable material with 12% salt penetration) with fuel salt at power densities more than six times the MSRE power density produced no observable damage to the graphite. Therefore, after consultation with representatives of the AEC Division of Reactor Development and the manufacturer, the graphite was accepted for use in the MSRE. 1.1.4 Compatibility of Salt, Graphite, and INOR-& Out-of-pile tests of combinations of fluoride salt, graphite, and INOR-8 over a periocd of several years have convincingly demonstrated the compatibility of these materials.? However, in-pile testing of fuel and graphite in INOR-8 capsules under conditions similar to those anticipated in the MSRE has been accomplished only since 1961. Although these first 20ak Ridge National Laboratory, "MSRP Semiann. Prog. Rep. Jan. 31, 1963," USARC Report ORNL-3419, pp. 70-76, and "MSRP Semiann. Prog. Rep. July 31, 1963," USAEC Report ORNL-3529, Chapter 4. W. D. Manly et al., "Metallurgical Problems in Molten Fluoride Sys- tems," Progress in Nuclear Energy, Series IV, Vol. 2, 164~179, Pergamon Press, London, 1960. 13 ‘Unclassified . Photo ¥-49020 . Unclassified __Photo ¥-49021 «} | Fig. 1.2. MSRE Gréfihite Shcnnng Cracks Resulting from Impregnation and Baking Operationms. o - , , W'} 14 irradiation experiments4 showed absolutely no evidence of attack on the INOR-8 or wetting of the graphite by the salt, appreciable duantities of Fo, and CF, were observed. Furthermore, in some of the capsules, Xenon was not found in the quantities expected. These anomalies were studied in two later series of in-pile capsule experiments during a total of seven months of exposure in the MIR. The following conclusions have been reached after a thorough analysis of in- formation accumulated during and after these experiments. Neither Fp nor significant quantities of CF, is generated when fission occurs in the molten salt at power densities as high as 65 w/cm®. However, Fp is re- leased from the irradiated salt after 36 or more hours of exposure to high-level beta and gamma fluxes at temperatures below 200°F. It also appears that CF, can be formed if Fp is present and avallable to the graphite when radiation is present at low temperatures or when the system temperature is raised to 1200 to 1400°F. Undoubtedly, the fluorine would also react with the INOR-& at these high temperatures. Finally, it must be concluded thas the "disappearance” of xenon in the tests resulted from a reaction of the fluorine to form one or more of the solid xenon fiuo- rides. These conclusions suppor: the important additional judgment that Fa and CF, formation should not affect the success of the MSRE. First, the salt in the MSRE should never be in the frozen state, except for small guantities at isolated locatiocns (e.g., freeze flanges and freeze valves). Even these isolated deposits would not be at temperatures as low as 200°F. Second, if Fp were evolved from the frozen seals, the rate of reaction with the INOR-8 metal at 200°F would be insignificant; at higher tempera- tures the back reaction would reccmbine the Fp and the reduced salt with- out an observable effect. Samples of INOR and graphite are located at a position of maximum flux within the reactor vessel to allow examination and testing after “0ak Ridge National Laboratory, "MSRP Semiann. Prog. Rep. Jan. 31, 1963," USAEC Report ORNL-3419, pp. 8§0-107, and "MSRP Semiann. Prog. Rep. July 31, 1963," USAEC Report ORNL-3529, Chapter 4. 15 various periods of exposure to salt. Salt samples for detaliled chemical gnalysis will be removed and will be analyzed daily. 1.2 System Components The components of the reactor are arranged in the fuel and coolant salt systems as depicted in Fig. 1.3. The individual pieces of equipment are described below. All the components are designed to meet Section VIII of the ASME Boiler and Pressure Vessel Code except that no pressure-relief devices are provided on the primary system. Instead, supply pressures (for example, cover gas) are limlted by control and protective devices. UNCLASSIFIED ORNL-DWG 63-1209R N i "% REMCTE MAINTENANCE f | - o .+ CONTROL ROOM : : ‘[ REAGTOR CONTROL [k““Am ROOM., o \ I, REACTOR VESSEL 7. RADIATOR S 2. HEAT EXCHANGER 8. GOOLANT DRAIN TANK U 3. FUEL PUMP 9. FANS 4 FREEZE FLANGE 10. FUEL DRAIN TANKS Y T 5. THERMAL SHIELD 11. FLUSH TANK AN T 6 COOLANT PUMP 12. CONTAINMENT VESSEL 13. FREEZE VALVE Fig. 1.3. MSRE Layout. 16 1.2.1 Reactor Vessel The reactor consists of a cylindrical vessel approximately 5 ft in diameter and 7 1/2 £t high, fitted with an inner cylinder that forms the inner wall of the shell-cooling annulus and serves to support the 55-in.- diam by 64-in.-high graphite matrix with its positioning and supporting members. Figure 1.4 is an assembly drawing of the reactor vessel and its graphite core, Fluid enters the vessel at the top of the cylindrical sec- tion and flows downward in a spiral path along the wall. With the design flow of 1200 gpm in the 1l-in. annulus, the Reynolds modulus is about 22,000. At the estimated heat generation rate of 0.24 w/cm3 in the wall, 28 kw of heat 1s removed by the incoming salt, while maintaining the wall tempera- ture at less than 5°F above the bulk fluid temperature. Design data for the reactor vessel are listed in Table 1,6, The fuel loses its rotational motion in the lower plenum and then flows upward through the graphite core matrix, which constitutes about 77.5% of the core volume. The moderator matrix is constructed of 2- by 2- by 63-in, stringers of graphite, which are loosely pinned to restrain- ing beams at the bottom of the ccre. Flow passages in the graphite matrix are 0.4- by 1.2-in. rectangular channels, with rounded corners, machined into the faces of the stringers. A typical arrangement of graphite stringers is shown in Fig. 1.5. Flow through the core is laminar, but because of the good thermal properties of the graphite and the fuel, the graphite temperature at the midpoint is only 60°F above the fuel mixed-mean temperature at the center of the core. The average power density in the fuel is 14 kw/liter and the maxi- mum is 31 kw/liter. The salt leaves the vessel at the top through & 10-in. pipe with a 5-in.-diam side outlet connecting to the circulating pump. A course screen is provided here to catch pieces of graphite larger than 1/8 in, in diame- ter should they be present in the salt stream. The 10-in. pipe accommo- dates an assembly of three 2-in.-diam control rod thimbles and permits access to a group of graphite and INOR-8 surveillance specimens arranged in the center of the core as indicated in Fig. 1l.6. The assembly 1is flanged for removal, if necessary, and is provided with cooling air to 17 UNCLASSIFIED ORNL-LR- DWG 61097 /1 FLEXIBLE CONDUIT TO fq,,. /CONTROL ROD DRIVES GRAPHITE SAMPLE ACCESS PORT f}p{; COOLING AIR LINES CORE CENTERING GRID e /é//rFLOW DISTRIBUTOR FUEL INLET REACTOR CORE CAN REACTOR VESSEL e RS T e ANTI-SWIRL VANES A X< MODERATOR VESSEL DRAIN LINE A SUPPORT GRID Fig. 1.4. Reactor Vessel and Access Nozzle. 18 SSSSSSSSSSSS PLAN VIEW TYPICAL MODERATOR STRINGERS REMOVABLE STRINGER NOTE: NOT TO SCALE Fig. 1.5. Typical Graphite Stringer Arrangement. SSSSSSSSSSSS RRRRRRRRRRRRRRRRRR ..-CONTROL ROD @by S\ & N \\/ 0 § GUIDE TUBE \\\\Q KJ/\\ GUIDE BAR TYPICAL X 7 NV AN N © \| NN ... \ 0.200-in. R/ 0.400 S R = REMOVABLE R STRINGER 4 GRAPHITE IRRADIATION 274 R N SAMPLES (7g-in.DIA)} 7 ‘ - 2 in. TYPICAL Fig. 1.6. Lattice Arrangement at Control Rods. 19 Table 1.6. Reactor Vessel and Core Design Data and Dimensions Construction material INOR-8 Inlet nozzle 5 in., sched-40, IPS Outlet nozzle 5 in., sched-40, IPS Core vessel Outside diameter 59 1/8 in. (60 in. max) Inside diameter 58 in. Wall thickness 9/16 1in. Overall height (to centerline of 100 3/4 in. 5-in. nozzle) Head thickness L in. Design pressure 50 psi Design temperature 1300°F Fuel inlet temperature 1175°F Fuel outlet temperature 1225°F Cocling annulus Inside diameter 56 in. Qutside diameter 58 in, Graphite core matrix Diameter 55 1/4 in, Number of fuel channels 1140 Fuel channel size 1.2 x 0.4 in. with rounded cormers Core inner container Inside diameter 55 1/2 in. Outside diameter 56 in. Wall thickness 1/4 in. Height 68 in. permit freezing the salt in the gap below the flange. The control rods and their drive mechanisms are described in Section 1.3. 1.2.2 Fuel and Coolant Pumps The fuel-circulation pump is a sump-type centrifugal pump with a vertical shaft. It has a 75-hp motor and is capable of circulating 1200 gpm of salt against a head of 48.5 ft. TFigure 1.7 is a drawing of the pump, and design data are presented in Table 1.7. The pump assenbly consists of motor and housing; bearing, shaft, and impeller assembly; and a 36-in.-diam sump tank. The sump tank, which is _ SHAFT SEAL— " (See Inset) LEAK DETECTOR LUBE OIL IN- BALL BEARINGS {Face to Face) BALL BEARINGS ~,__ { Back to Back) ‘ LUBE OfL OUT._ SEAL OIL LEAKAGE &> DRAIN — - LEAK DETECTOR-" SAMPLER ENRICHER - {Qut of Section) . {(See Inset) BUBBLE TYPE»//;«" LEVEL INDICATOR # OPERATING - I LEVEL " | To Overflow Tank .f"v P P | — ,J] l___4 - G BN SHAFT Cheom == = 3 NWATER COUPLING . Mo T T T e 2D COOLED T vy 1 : MOTOR A M~ 1 4 ‘ '4' s 7 - o 4 T i . UNCLASSIFIED ORNL- LR -DWG-56043-B _»LUBE OIL BREATHER 1{.@.__” ~BEARING HOUSING 7 __GAS PURGE IN A - " ' ’T" .~ SHAFT SEAL (See Inset ) — | ~~—SHIELD COOLANT PASSAGES {in Parallel With Lube Oil )} SHIELD PLUG GAS FILLED EXPANSION SPACE XENON STRIPPER {Spray Ring) >+ SPRAY Fig. 1.7. Fuel-Circulation Pump. /GAS PURGE OQUT ( See Inset)— oc¢ 21 Table 1.7. Design Data for Fuel and Coolant Pumps Fuel Coolant Design flow, gpm Circulation 1200 850 Bypass (50-gpm spray, fuel pump only) 65 15 Head at 1200 gpm, £t (fuel pump) 48.5 Head at 850 gpm, £t (coolant pump) 78 Motor, hp 75 75 Speed, rpm 1160 1750 Intake, INOR-8, sched-40 IPS, in. 8 6 Outside diameter, in. 8.625 6.625 Wall thickness, in. 0.322 0.280 Discharge nozzle, INOR-8, sched-40 IPS, in. 5 5 Outside diameter, in. 5.563 5.563 Wall thickness, in. 0.258 0.258 Pump bowl, INOR-8 Diameter, in. 36 36 Height, in. 15 15 Volumes, £t Minimum starting and normal operating volume 4.l 4.1 (including volute) Maximum operating volume 5.2 5.2 Maximum emergency volume (includes space 6.1 6.1 above vent) Normal gas volume 2.0 2.0 Overall height of pump and motor assembly, It 8.6 8.6 Design pressure, psi 50 50 Design temperature, °F 1300 1300 welded into the reactor system piping, serves as the expansion tank for the fuel and as a place for the separation of gaseous fission products. Separation is accomplished by spraying a 50-gpm stream of salt through the atmosphere of helium In the sump tank. The bearing houging is flanged to the sump tank so that the rotating parts can be removed and replaced. The motor is loosely coupled to the pump shaft, and the motor housing is 22 flanged to the upper end of the bearing housing to permit separate removal cf the motor. The pump is equipped with ball bearings, which are lubricated and cooled with oil circulated by an external pumping system. The oil is confined to the bearing housing by mechanical shaft seals. Helium is circulated into a labyrinth between The lower bearing and the sump tank. Part of the gas passes through the lower seal chamber to remove oil vapors which might lesk through the seal. The remainder of the helium flows downward along the shaft to prevent radioactive gas from reaching the oll chamber. Bubbler tubes are provided in the pump bowl to measure the liquid level so that the pump will not be overfilled and to permit a determina- tion of the salt inventory. The salt-sampling device ig attached to the bowl at another opening to allow the removal of approximately 10~-g por- tiong of salt for chemical analysis or to add enriched uranium salt in quantities of 150 g (20 g of U). Massive metal sections are incorporated in the pump assembly as shielding for the lubricant and the motor. The motor is enclosed and sealed to prevent the escape of radiocactive gas or fluids that might leak through the pump assembly under unusual conditions. Water cooling coils are attached to the housing to remove heat generated by the motor. Tmmediately underncash the purp is a torus-shaped tank (5.5 £t2 ) which serves as an emergency overflow tank for collecting the fuel in the event of coverfilling of the purp or expansion of the fuel in the pump as a result of an excessive salt temperature. The overflow tank is vented to the offgas system and is egquipped with level indicators. ©5Should salt flow into the tank, 1t can be pressurized back into the pump bowl. The pump bowl and the overflow tank are enclosed in an electrically heated furnace. The same type of pump, without the overflow provision, is used in the coolant system. This pump, however, dces not require asg much pro- tection against radiaticn. It is driven by a 75-hp motor and is designed to circulate 850 gpm of salt against a head of 78 ft of fluid. The com- plete design data for the coolant pump are included in Table 1.7. 23 1.2.3 Primary Heat Exchanger The primary heat exchanger (Fig. 1.8) contains 159 tubes (1/2 in. 0D, 0.042-in. wall) and is designed to transfer 10 Mw of heat from the fuel salt (in the shell) to the coolant salt {in the tubes). The ex- changer has a conventional, cross-baffled, shell-and-tube configuration. The tube bundle is laced with metal strips (not shown) to prevent vibra- tion of the tubes. Design data are listed in Table 1.8&. Space limitations in the reactor cell require a short unit. The U-tube configuration makes possible a length of & ft without greatly re- ducing the efficiency of heat transfer, as compared with a straight UNCLASSIFIED ORNL-LR-DWG 52036R2 FUEL INLET i/2-in.-0D TUBES THERMAL-BARRIER PLATE CROSS BAFFLES TUBE SHEET COCLANT INLET 16.4-in. OD x 0.2-in. WALL x 8-ft LONG COOLANT-STREAM T T COOLANT OUTLET SEPARATING BAFFLE fl FUEL OUTLET Fig. 1.8. Primary Heat Exchanger. 24 Table 1.8. Design Data for Primary Heat Exchanger Construction material INCR-8 Heat load 10 Mw Shell-side fluid Fuel salt Tube-side fluid Coolant salt Layout 25%-cut cross-baffled shell with U tubes Baffle pitch 12 in. Tube pitch G.775 in., triangular Active shell length ~6 1t Overall shell length ~8 ft Shell diameter 16 in. Shell thickness 1/2 in. Average tube length 14 £t Number of U tubes 159 Tube size 1/2 in. OD; 0.042-in. wall Effective heat transfer surface 259 1% Tube sheet thickness 1 1/2 in. Fuel salt holdup 6.1 ft° Design temperature Shell side 1300°F Tube side 1300°F Design pressure Shell side 55 psig Tube side 90 psig Allowable working pressure® Shell side 75 psig Tube side 125 psig Hydrostatic test pressure Shell side 800 psig Tube side 1335 psig aBased on actual thicknesses of materials and stresses al- lowed by ASME Code. 25 Table 1.8 (continued) Terminal temperatures Fuel salt 1225°F, inlet; 1175°F, outlet Coolant 1025°F, inlet; 1100°F, outlet Effective log mean temperature 133°F difference Pressure drop Shell side 28 psi Tube side 27 psi Nozzles Shell 5 in., sched-40 Tube 5 in., sched-40 Fuel-salt flow rate 1200 gpm Coolant-salt flow rate 850 gpm counter-flow unit, and eliminates a thermal expansion problem. The tubes are welded and back brazed to the tube sheet in order to reduce the prob- abllity of leakage between the fuel and coolant, The coolant pressure is kept higher than the fuel pressure to reduce the likelihood of fuel out- leakage in case of a tube failure. 1.2.4 Salt-to-Air Radiastor The thermal energy of the reactor 1s transferred to the atmosphere at a salt-to-air radiator, which is cooled by two 100,000-cfm blowers. The radiator contains 120 tubes (3/4 in. OD, 0.072-in. wall, 30 ft long) and is assembled as shown in Fig. 1.9. Design data for the radiator are listed in Table 1.9. Several features were incorporated in the design as protection against freezing of coolant salt in the radiator: 1. The tubes are of large diameter. 2. The heat-removal rate per unit area is kept low by using tubes without fins so that most of the temperature drop is in the air fiim. 3. The nminimum salt temperature is kept 175°F above the freezing point (850°F). A thermocouple is attached to each tube for temperature monitoring. /'/ DRI 7 S DOOR DRIVE MOTO GEAR REDUCE :‘5 ‘|‘ ;'1 1t SUPPORTING STEEL-TT DOOR CAM GUIDE ——J17 . BLDG. 7503, FIRST FLOOR (ELEV. 852 ft-0in.} . / ¢ i t ; 7 \ / L AIR INLET DUCT —~.__ | / ~ /" AIR DUCT FLANGE -~~~ 1 | Fig. _SHOCK SPRING 7| INLET DOOR --} S AIR BAFFLES ————F ,-"'}’- \ MAIN AIR BYPASS DUCT 7 26 v/ ;’:’/’ | P VE CHAIN ~_ ( @ R AND - R T ! e i - AN AIR FLOW / / / 1.9. Salt-to-Alr Radiator. S =~ . COOLANT PUMP UNCLASSIFIED ORNL-LR-DWG 55841R2 _~PENTHOUSE [—AIR CUTLET DUCT A 5 RADIATOR ENCLOSURE RADIATOR TUBES 27 Table 1.9. Design Data for Salt-to-Air Radiator Structural material INOR-8 Heat load 10 Mw Terminal temperatures Coolant salt 1100°F, inlet; 1025°F, outlet Air 100°F, inlet; 300°F, outlet Air flow 164,000 cfm at 15 in. HpO Salt flow 850 gpm at avg temperature Effective mean temperature difference 920°F Overall coefficient of heat transfer 53 Btu/hr-ft?.°F Heat transfer surface area 685 £t? Design temperature 1300°F Design pressure 75 psi Tube diameter 0.750 in. Wall thickness 0.072 in. Tube matrix 12 tubes per row; 10 rows deep Tube spacing i 1/2 in., triangular Subheaders 2 1/2 in., IPS Main headers g in., IPS Alr-side pressure drop 11.6 in. Hy0 Salt-side pressure drop 6.5 psi 4. The headers are designed to assure even flow distribution be- tween the tubes. 5. In the event of flow stoppage, doors on the radiator housing can close within 30 sec. 6. The electric heaters mounted inside the radiator enclosure are never turned off. 7. The salt can be drained in approximately 10 min. The layout of the tube matrix will allow movement of the tubes with minimum restraint during thermal expansion. The tubes are pitched to promote drainage. 28 The radiator is supported and retained in a structural steel frame that is completely enclosed and insulated. Reflective shields protect structural members from excessive temperatures. The frame also provides guides for the two vertical sliding doors, which can be closed to thermally isolate the radiator. The doors are installed on the radiator enclosure, one upstream and one downstream, and they can be raised and lowered at a speed of 10 ft/min during normal operation by a gear-reduced motor driving an overhead line shaft. The doors are suspended from wire rope, which runs over sheaves mounted on the line shaft. The enclosure is capable of sustaining full blower pressure with the doors in any position. The doors may be used to regulate the air flow across the radiator as a means of controlling the reactor load. However, the load is normally controlled by positioning a damper in a bypass duct and by switching fans on and off. More detail of the load-control plan may ke found in Section 2. Emergency closure is effected by de-energizing a magnetic brake on the line shaft; this permits the doors to fall freely. ©Shock absorbers are provided. Accidental closure has no scrious effects except for loss of time; it reduces the reactor vnower to <100 kw. 1.2.5 Drain and Storage Tanks Five tanks are provided for safe storage of salt mixtures when they are not in use in the reactor and coolant systems. They comprise two fuel drain tanks, a fiush salt tank, a fuel-and-flush-salt storage tank, and a coolant drain tank. Fuel Drain Tanks. The fuel drain tanks serve the important function of subecritical storage of the fuel. They are water cooled for removing fission-product decay heat, and an electric furnace is installed for maintaining the salte molten when the internal heat generaticn rate is low. Two tanks of the design shown in Fig. 1.10 are provided; each has a volume of 80 ft?. Each tank can hold an entire fuel charge, so one is for normal use and the other is a spare. The low moderating power of the salt makes criticality very unlikely, even with nearly double the planned U235 loading (see sec. 7.1.9). After long-term operation at 10 Mw, sudden draining of the fuel will regquire that it be cooled at a rate of 100 kw to prevent excessive fuel 29 UNCLASSIFIED ORNL-LR-DWG 61719 INSPECTION, SAMPLER, AND LEVEL PROBE ACCESS STEAM OUTLET STEAM DOME CONDENSATE RETURN WATER DOWNCOMER INLETS CORRUGATED FLEXIBLE HOSE STEAM RISER BAYONET SUPPORT PLATE STRIP WOUND FLEX!BLE HOSE WATER DOWNGCOMER BAYONET SUPPORT PLATE HANGER CABLE GAS PRESSURIZATION S AND VENT LINES INSTRUMENT THIMBLE FUEL SALT SYSTEM FILL AND DRAIN LINE SUPPORT RING FUEL SALT DRAIN TANK BAYONET HEAT EXCHANGER THIMBLES (32) TANK FILL LINE 3 FUEL SALT SYSTEM FILL AND DRAIN LINE TANK FILL LINE Fig. 1.10. Fuel Salt Drain Tank. 30 ° Evaporative cooling was chosen over gas or other means on temperatures. the basis of simplicity and independence from utilities. Heat is removed by 32 bayonet cocoling tubes (Fig. 1.11) inserted in thimbles in the tank. Water is fed through the center tube of the bayonet assembly, and steam is generated in the surrounding snnuvlus. Heat is transferred from the thimble to the cooling tube by radiation and conduction. Normally the steam is condensed in a water-cooled condenser, but it can be exhausted o the vapor-condensing system in the event of failure of the coolant supply. A 300-gal feed-vwater reserve can provide cooling for 6 hr. The drain tanks have dip-tube fill and drain lines and gas connections for maintaining a helium blanket for ventilating the space over the salt and for pressurizing to transfer the salt. Desigh data for the drain tanks gre presented in Table 1.10. | Flugh Salt Tank. The LiF-Bel's salt mixture with which the fuel sys- tem is flushed before maintenance will not accumulate sufficient fission products to require cooling during storage. For this reason the flush salt tank does not have a cooling system. Otherwise its design 1s similar to that of the fuel drain tanks. Coolant Drain Tank. A tank (see Table 1.1C) of 50-ft3 capacity that ig similar tc the fuel drain tanks but without cooling tubes 1s alsc pro- vided for the coolant salt. Tuel Storage Tank and Chemical Processing System. Batches of fuel or flush salt removed from the reactor circulating system can be processed in the fuel storage tank and its associated equipment to permit their re- use or to recover uranium. Salts that have been contaminated with oxygen conctituents as oxides can be treated with a hydrogen-hydrogen fluoride gas mixture to remove the oxygen as water vapor. A salt batch unacceptably contaminated with fission products, or one in which it is desirable to drastically change the uranium content, can be treated with fluorine gas to separate the uranium from the carrier salt by volatilization of UFg. In some instances the carrier salt will be °L. F. Parsly, "MSRE Drain Tank Heat Removal Studies,"” USAEC Report ORNL CF-60-9-59, 0Ozk Ridge National Laboratory, September 1960. 31 UNCLASSIFIED ORNL-LR-DWG 60838A1 - L i v e e 3 u s o < > w L ~ ) wy L | \ \ 1 AN A A X T L — z o i <[ = STEAM DOME LOWER HEAD BAYONET SUPPORT PLATE DRAIN TANK HEAD _?;_ \ __ _ _: Bayonet Cooling Thimble for Fuel Drain Tanks. Fig. 1.11. Table 1.10. 32 Design Data for Fuel Drain Tank, Coolant Drain Tank, and Flush Salt Tank Tyel drain tank Material Height Diametcr, outside Volume Total Fuel (normal) Gas blanket (normal) Wall thickness Vegeel Dished head Design temperature Design pressure Cocling methed Coocliling rate Coolant thimbles Number Diareter, IPS Coolant drzirn tank Mztcrial Height Diameter, outside Volume Total Coolant salt Gas tlanket Wall thickness Vesgel Dished head Degign temverature Desipgn pressure Cooling methed Flush salt tank Material Height Diameter, ocutside Velume Total Flush salt Gas blanket Well thickness Vessel Dished head Design temperature Design vressure Cooling method INOR-8 86 in. (without coolant headers ) 62 psi Boiling water in double- walled thimbles o0 kw INOR-& 78 in. 40 in. 50 T3 ~l T3 ~6 117 2/8 in. 5/8 in. 13C0°F 65 psi Nene INOR-8 84 in., 50 in. g2.2 £t 73.2 3 o f3 1/2 in. 3/4 in. 1300°F 65 psi None 33 discarded; in others, uranium of a different enrichment, thorium, or other constituents will be added to give the desired composition. The processing system consists of a 117-ft° salt storage and process- ing tank, supply tanks for the Hp, HF, and F, treating gases, a high- temperature (750°F) sodium fluoride adsorber for decontaminating the UFg, several low-temperature portable NaF adsorbers for UFg, a caustic scrubber, and associated piping and instrumentation. All except the UFg adsorbers (which do not require shielding) are located in the fuel processing cell below the operating floor of Building 7503. The process is described in Part VII of the MSRE Design and Operations Report. After the uranium has been transferred to the UFg adsorbers, they will be transported to the ORNL Volatility Pilot Plant, where the UFg will be removed and prepared for reuse. The entire processing system is separated from the reactor system with two freeze valves in series to ensure complete isclation. 1.2.6 Piping and Flanges The reactor vessel, pumps, and heat exchanger are interconnected by 5-in.-IPS, sched-40, INOR-8 piping. The 0.258-in. wall is several times the thickness necessary for operation at 1200°F and 75 psig. Piping of smaller diameter is used for the drain lines (1 1/2 in.) and other aux- iliaries, but no salt lines are smaller than 1/2 in., The components of the fuel-circulating system are equipped with a special type of flange, called a freeze flange {(shown in Fig. 1.12), to permit replacement. These flanges utilize a seal made by freezing salt between the flange faces, in addition to a conventional ring gasket seal which is helium buffered for leak detection. 1.2.7 TFreeze Valves The molten salt in both the fuel and cooclant circuits will be sealed off from the respective drain tanks by means of freeze valves in the drain lines. These valves (Fig. 1.13) are simply short, flattened sec- tions of pipe which are cocled tc freeze the salt in that section. Heaters surround each valve so that the salt can be thawed quickly when necessary to drain the system. The salt can also be thawed slowly without the 34 UNCLASSIFIED ORNL-LR-DWG 63248R2 CLAMP— — = NS o FLANGE \ f'n\"f"‘}é}n \\ ;GAPmeH BUFFER | CONNECTION (SHOWN ROTATED) MODIFIED R-68 | i< RING GASKET -~ FROZEN o 4% —in. SALT SEAL | v in. R 1 Y%-in-R (TYP) | SLOPE 1:4 //// % (TYP) !1/4 in.——[ 5-in. SCHED-40 PIPE f — . L I I ) - - ] . (3 Fig. 1.12. Freeze Flange. heaters by stopping the cooling. In addition to the freeze valves in each drain line, freeze valves are also provided in the transfer lines between the drain tanks, flush tank, and the storage tank. 1.2.8 Cover-Gas Supply and Disposal Because the fuel salt is sensitive to oxygen-containing compounds, it must be protected by an oxygen- and moisture-free cover gas at all times. The principal functions of the cover-gas systems are to supply an inert gas for blanketing the salt and for the pressure transfer of : ! UNCLASSIFIED PHOTO 70158 x THERMOCOUPLES RN el SAEME ‘ e - e Wi g - - 36 salt between components, to provide a means for disposing of radiocactive gas, and to maintain a higher gas pressure in the coolant system than in the fuel system. A simplified flow diagram of the system is presented in Fig. 1.1l4. The cover gas is helium supplied in cylinders at 2400 psig. It is purified by passage through filters, dryers, and oxygen traps (titanium at high temperature). Purified gas is then sent to two distribution systems, one for the fuel and one for the coolant salt. The total flow is about 10 liters/min (STP) at about 40 psig. A rupture disk protects the reactor against supply pressures greater than 50 psig. The largest flow of gas is directed to the fuel-circulation pump, the freeze-flange buffer zones, and the fuel drain tanks, where it is in di- rect contact with the fuel salt. Gas that passes through the fuel pump is circulated through a series of pipes where it is held and cooled for at least 2 hr to dissipate heat from the decay of short-lived fission products. Then it passes through & charcoal bed, where krypton and Xenon are retained for at least & and 72 days, respectively, and through a fil- ter and blower To the offgas stack. There 1t 1s mixed with a flow of 18,000 cfm of air, which provides dilution by a factor of approximately 10° and reduces the concentration of Kr85, the only significant isotope remaining, to 10-7 uc/cc at the stack exit. The charcoal bed is a series of pipes packed with activated carbon. It and a spare bed are mounted vertically in a secaled, water-filled secondary container; either or both beds may be used. Fuel salt transfers require more rapid venting of the cover gas, but the heat locad is low. A third charcoal bed is provided for venting those gases before they are sent to the stack. Figure 1.15 is the offgas dis- posal flowsheet. The cover-gas distribution for the coolant system (also shown on Fig. 1.14) supplies a small flow of helium to the coolant system, the sampler-enricher system, and to the coolant pump. Gas from the coolant system is vented directly through filters to the offgas stack, as indi- cated in Fig. 1.15. Monitors will stop the flow to the stack on indica- tion of high activity. 250 psig DRYER N TO ATM. OXYGEN REMOVAL UNIT 1200°F \ F_{ 7 HELIUM SUPPLY TRAILER 2400 psig Co SUPPLY . HEADER T ARRANGEMENT AT CONTAINMENT CELL WALL TYPICAL FOR SUPPLY FUEL PUMP SWEEP GAS NTAINMENT RUPTURE DISC 350 psig f— UNCLASSIFIED ORNL DWG. 64-600 Py LEAK DETECTOR SYSTEM TREATED HEL IUM SURGE TANK | REACTOR CELL TO: FUEL PUMP LEVEL BUBBLERS FUEL DRAIN TANKS Fig. 1.14. ZZ( ENRICHER-SAMPLER 250 psig HEADER GRAPHITE SAMPLER RADIATION MONITOR f-—\TO STACK RUPTURE DISC 50 psig ey SPENT FUEL PROCESSING 40 psig LEVEL BUBBLERS HEADER I FUEL SYSTEM DRAIN TANKS g COOLANT PUMP AND DRAIN TANK P FUEL PUMP SWEEP GAS e e PUMP O SYSTEMS Cover~-Gas System Flow Diagram. LE UNCLASSIFIED ORNL DWG. 64-594R RESTRICTOR\\ FUEL PSMP | [ CHARCOAL BED = HELIUM VOLUME s @r . VOLUME SUPPLY HOLDUP LOOP | HOLDUP LOOP SPARE L 1] I XF_cHarcoal Beo X l AUXILIARY N OVERFLOW 1 ™ CHARCOAL BED TANK - I | . ACTIVITY | - 1 I MONITOR¥ | I | TYPICAL DRAIN I g& 0.4 ¢ S / . »— i g / | x / 0.2 /,I // 2 z 00——"”‘ 0 10 20 30 40 50 60 70 DISTANCE INSERTED (in.) Fig. 2.4. Reactivity Worth of Control Rod as a Function of Depth of Insertion in Core. 52 UNCLASSIFIED ORNL—DWG 64— {1109R DISTANCE (in.) s b— — e e e XX' . 60 Q 0.25 0.50 0.75 1.00 .25 1.50 ELAPSED TIME (sec) CURVE A-REFERENCE CURVE QF SATISFACTORY SCRAM PERFORMANCE; BASED ON ACCELERATION OF 5 ft/sec? AND RELEASE TIME OF 0400 sec CURVE B-SCRAM PERFORMANCE FROM TESTS OF JAN.27-28,1964 Fig. 2.5. Control Rod Height Versus Time During a Scram. 4. The overrunning clutch transmits drive motor torque in the rod insert direction, and the control system contains an interlock with the safety system that turns the drive motor "ecn'" in the rod-insert direction following a rod scram. This will provide a force in the order of 350 1b to push the rod into the core if it is stuck. 5. A low-~angle, low-efficiency worm and gear between the servo motor and the sprocket locks the drive train so that if an external torgue in the direction tending toward rod withdrawal is applied to the drive sprocket, 1t will not rotate. A conservative stress ectimate of elements in the drive train from the worm gear to the sprocket chain gives a value of 463 1b as the gafe static torque. This torque is produced by a dif- ferential pressure of 450 psi acting on the projected cross-sectional area of the control rod. 53 6. The electromechanical clutch is a single-plate, flat-disk unit with the driven plate spring lcoaded: it has a stationary field winding. The clutch is disengaged with no field current. 7. The coarse-position synchro rotates 300° for full stroke and, hence, provides unambiguous information. The synchros transmit the angular position of the drive-sprocket shaft and do not take into account small changes of rod position caused by stretching of the drive chain and the hollow, flexible, support hose. A pneumatic, single-point, fiducial, zero-position-indicating device 1s provided that is activated by the change in pressure drop of the cooling air leaving the bottom of the rod. When the rod is at the zero position, the flow of cooling air from the radial exhaust nozzle at the bottom of the rod is impeded by a constricting throat near the bottom of the thimble and the sharp change in differential pres- sure indicates the location of the rod with respect to the thimble to an accuracy of 0.030 in. Figure 2.2 shows the nozzle and throat. This device will be used for periodic determination of the relation between synchro outputs and the position of the poison elements with respect to the thimble. 2. The shock sbsorber, Fig. 2.6, is of the spring-loaded "hydraulic" type, but it is unconventional in that the working "fluid" in the cylinder is hardened steel balls. The stroke ig adjusted between 2.75 and 4 in. by varying the spring constants and the preloads on the buffer return spring and the ball reset spring. The resultant deceleration during shock absorption is seven times gravity. Tegts! of the prototype rod and rod drive have included more than 1100 scrams from full withdrawal without any malfunction of the shock absorber. This i1s greatly in excess of the duty required in service. 9. The weight of the control rod plus the shock absorber produces an unbalanced weight on the sprocket chain of 18 1b. Static tests? of the prototype show that the internal friction of the drive unit plus the thimble-to-rod friction is equivalent to an upward force of 10 1b on the sprocket chain. Thig leaves an unbalanced force of 8 1b in the scram direction. This is in good agreement with the acceleration given in para- graph 2 above. M. Richardson, private communication, May 12, 1964, UNCLASSIFIED ORNL-DWG 64-983 54 T T Y T T e e o et o ool ool Z T BBl d Bl 2T T T T T T T T T TTTT g LTI e SR T AaH T ., M | ;wfiwmhfl...fluunnm,pawqm,flmu ......... - @ == - | &\\ \\W ) Q8 = SR A 00000008 Attty Ry 2o eiad = e \ > L ”I AN S S Yy ....Isl.liI_ LIRS e it ~—rme L U U] 770 RN SRS Bl 2 el O el _c¢ccc¢c 2N \\/\ e e i A e e /\ \ \ \ w\\,\ h o . z & 2 Q- B » z % £ L B g & Q 3 o 2 z P 5 x 9 EZ y 2 k- E R @ & 23 w W o _m bu o o & w5 m.M 2 & m ¥ o~ 4 Zz G g o £ 4 < & B @ = E o : t 2 2 3 Wi & z 3 3 S g 2 i & 2 ooh 3 8 B E g Control Rod Shock Absorber. Fig. 2.6. 55 10. The upper and lower limlt switches at‘the ends of the total rocd stroke are in independent pairs and are wired independently. All these switches are located at the upper end of the drive unit to eliminate long electrical leads adjacent to moving parts. This 1s also the most favorable location from the standpoint of temperature and radiation intensity. 11. A photograph of a prototype thimble is shown in Fig. 2.7. It is substantially the same ag the thimble that will be installed in the core vessel. The bottom of the thimble is closed and (see Fig. 2.1) is only 0.% in. above the graphite stringers on the grid plate. Thus there is ample protection against a rocd falling out of the core and initiating = nuclear excursion. The prototype rod drive assembly, supplied by the Vard Division of Royal Industries, Inc., Pasadena, Californis, is shown 1In Mg, 2. 8. 2.2 Safety Instrumentation The instrumentation and control devices classified as safety equip- ment are those whose Tailure to perform their protective function when required would result in an unacceptable hazard to personnel or unaccept- able damage to the reactor or its major components. The following principles were used in the design of the MSRE safety syetem in order to obtain the maximum degree of reliability: 1. Redundant and independent channels composed of high-grade com- ponents were provided for each protective function. Two degrees of redun- dancy are employed: (a) two independent channels, either of which will produce the required safety action, and (b) three independent channels in two-out-of-three coincidence. The second system requires agreement of any two channels to produce safety action. Removal of any channel for maintenance or operational checks i1s equivalent to a safety system input from that channel. In these circumstances the three-channel system re- verts to a simple two-channel system, 2. Provisions were made for periodic on-liine testing of each channel. 3. Continuous monitors were provided to disclose certain component failures and malfunctions, power failures, and loss of channel continuity. SRR T T AR A R AR T e T R TR R T A A e * UNCLASSIFIED PHOTO 39863 56 Control Rod Thimble. 7 2 Fig. ' L\" 0. _“. CONTROL ROD DRIVE - UNIT HOUSING ¥ THIMBLE ADAPTOR FLANGE 57 “7 UNCLASSIFIED © PHOTO 62054 _ POWER- AND POSITION- i INDICATOR PACKAGE _CHAIN DRIVE -~ . TOWBLOCK AND =~ " ROD DISCONNECT - e e e e 58 4. The use of all components in the safety system is restricted to providing safe reactor Qpefatibn; e.g., modifications to the safety in- strumentation for the confenience of data gathering by experimenters is prohibited. Similarly, the addition to the safety system of extra or auxiliary fUncfions that do not contribute to reactor system safety is considered a dilution of safety system efféctiveness and is prohibited. Typical alterations in this category are nonsafety interloéks and alarms. 5. Physical protection and separation of all components in each channel (conduiting, closed cabinets, etc.) are provided. 6. All safety-equipment components in each channel were identified as such. ’ ) | | Protection; separation, and identification (items 5 and 6)'are in- corporated throughout the system. Items 1 through 4 are discussed in detail inithe following paragraphs. A simplified block diagram of the MSRE nuclear and process safety system is shown in Fig. 2.9, which dis- pléys both the various inputs or conditions used to detect or to indicate the existence of unsafe conditions and the results obtained when these in- puts indicate that a hazard, real or potential, exists. The safety sys- tem input and output elements, their corrective actions, and the means em- ployed to attain the required reliability are listed in Tables 2.2 and 2.3, 2.2.1 Nuclear Safety System The nuclear safety system is required to decrease the reactivity re- liably upon the occurrence either of a high neutron flux or a high reactor outlet température. The primary safety elements for accomplishing this are the electromechanical rod-release clutches (for a description of the rod-drive mechanisms, see sec. 2.1). Deenergizing any clutch releases the drive sprocket (Fig. 2.3) from the motor and brake, and the absorber rod fglls into its thimble in the core. The instrumentation associated with these clutches and the initiation of a rod drop (scram) is described here. ’ A block diagram of the nuclear safety system instrumentation is shown in Fig. 2.10. The ranges of the flux safety channels, along with those of the flux channels used for control, are listed in Table 2.4. Details of chamber installations are given in Section 2.3. 4 59 UNCLASSIFIED ORNL-DWG 64-637 CONDITION CORREGTIVE ACTION FLUX SCRAM >15 Mw ALL RODS REACTOR OUTLET _ TEMPERATURE >1300°F — v OPEN - - BYPASS VALVES \ ¢ THE BYPASS VALVES AND ~ THAW DRAIN THE VENT VALVES ARE e FREEZE VALVES REACTOR REDUNDANT SETS, EITHER WiLL PROVIDE PRESSURE OPEN RELIEF IN DRAIN TANK * DURING A FUEL DRAIN VENT VALVES - ANY ROD NOT ABOVE B CLOSE MAIN He SUPPLY FILL POSITION 1 VALVE TO DRAIN TANKS 'FUEL PUMP BOWL CLOSE He SUPPLY VALVES PRESSURE > 2 psig USED TO FILL REACTOR FUEL PUMP BOWL He PRESSURE > 50 psig FUEL PUMP OVERFLOW TANK LEVEL >20% FS AGTIVITY IN REACTOR GELL AIR COOLANT PUMP OFF-GAS ACTIVITY > 20 mr/hr OPEN FUEL PUMP BOWL VENT VALVE CLOSE He SUPPLY VALVE TO FUEL STORAGE TANK i CLOSE BLOCK VALVES He LINES He SUPPLY PRESSURE CLOSE BLOCK VALVE IN CELL EVACUATION LINE < 30 psig CLOSE He SUPPLY LINES TO IN-CELL LEVEL INSTRUMENT BUBBLERS ACTIVITY IN QFF-GAS JTEMPERATURE AT RADIATOR OUTLET <1300 °F LOSS OF -COOLANT CLOSE BLOCK VALVE IN MAIN OFF-GAS LINE TO STACK CLOSE LUBE OIL SYSTEMS VENT TO OFF-GAS VALVES SALT FLOW ACTIVITY IN He LINE TO CLOSE RADIATOR DOORS IN-CELL- BUBBLERS REACTOR CELL GCLOSE LIQUID WASTE SYSTEM BLOCK VALVES _ PRESSURE >2 psig ACTIVITY N IN-CELL - _CLOSE INSTRUMENT AIR LINE BLOCK VALVES COOLING WATER SYSTEM ~ Fig. 2.9. f ’” . CLOSE IN-GELL COOLING WATER BLOCK VALVES Functional Diagram of Safety System. ' : ‘ UNCLASSIFIED : . : ORNL~DWS 64-638 CHAMBER VOLTAGE ‘ | MONITOR AND TEST - Q-260% CHAMBER VOLTAGE MONITOR AND TEST|[ | CHAMBER VOLTAGE MONITOR AND TEST e o e e i e o | e b Q-2601 HIGH TEMPERATURE TRIP e | Q-2601 oo [—— " - \ r---r r 7 '] CHAMBER FLUX ! | | cHAMBER | | cHamser | \ FLux ' I | Hv, supPLY AMP, | ‘ | | HwvsuPPLY i | WV suppRLY MP, | I Q-2602 | I | Q-2602 | - ——t Lo e e b o o e e e o e e =) DATA DATA DATA LOGGER LOGGER LOGGER Vo . ~ L L. | | | L I FAST TRIP FAST TRIP FAST TRIP FAST TRIP FAST TRIP FAST TRIP FAST TRIP FAST TRIP FAST TRIP COMPARATOR COMPARATOR COMPARATOR COMPARATOR COMPARATOR COMPARATOR COMPARATOR COM PARATOR COMPARATOR Q-2609 Q-2609 Q-2609 Q-2609 Q-2609 0-2609 Q-2609 0-2609 Q-2609 ECC ECC EcC gcCc | gce gce ROD REVERSE SAG BYPASS ROD REVERSE SAG BYPASS SCRAM ROD REVERSE SAG BYPASS ‘ : SCRAM SCRAM v, o £z J - . Ty Ny ’ | —l ! RELAY SAFETY RELAY SAFETY ELEMENT | RELAY SAFETY ’ ELEMENT | ' ece Q-2623 : ELEMENT Q-2623 | A T | Q-2623 _ Ecc | e | L £ec - ALARM ] ] ALARM N S 5 J - J [ COINCIDENCE COINCIDENCE COINCIDENCE MATRIX MONITOR MATRIX MONITOR MATRIX MONITOR ~Q-2624 Q-2624 Q-2624 NO. 1 ROD - . |noz rop NO.3 ROD CLUTCH . CLUTCH CLUTCH Fig. 2.10. Block Diagram of Safety Instrumentation for Control Rod Scram. ~ 09 O .‘ -) -t Drain fuel 61 Nf v C | : o Table 2.2. MSRE Safety System Inputs Condition or Situation Which Indicates - Causes of the Hazard, the Consequences, and v ) ‘ a Real or Potential Hazard the Corrective Action 1' - Supplementary Information : ; ‘ - ' : - : I. Excess reactor power; 9 > 15 Mw 1. Causes _ 1. Redundancy: Three independent channels provide flux \ : ' a. Uncontrolled rod withdrawal - information b. Premature criticality during filling (excess 2. Testig Response of each channel is tested, with the U23% due to partial freezing) exception of the ion chamber response to £lux changes ¢. Cold slug 3. Monj._toring. System design provides continucus check on ) d. Unimowvn or unldentified mechanisms circuit continuity and amplifier operation ‘ 2. Consequences . L 4. ‘- Safétx only?: Input information used solely for safety o , Excessively high temperatures; first in the fuel salt and to restrict reactor operation within safe limits loop, and ultimastely in the coolant salt loop, with L vt e ' resultant damage to equipment and, if unchecked, loss 5. Thig system employs two-out-of-three coincidence . of primary containment - Refer to Section 2.2 and Fig. 2.10 v 3. Corrective action | - a. Scram all rods _ : ! b. Open vent valves; relieves filling pressure in ! o : fuel dra.in ta.nk (see .1b above) - 9 S i -IT. Fuel salt outlet temperature 1. Causes [ Redfinda.ncz: - Three indepehdent channels greater than 1300°F Bee I above; also situations in which power genera-. 2. Testing: Response of each individual input channel may tion exceeds rate of heat rejection be, tested ‘ 2. Consequences - 3. Monfi.toring' A thermocouple break or detaclment from 'See T sbove pipe wall produces upscale reading and sa.fety a.ction _ . in that channel . 3. Correct;ve actions 4 ‘SaiJet onlv?: Yes a. Scram rods : ‘ o b. Drain fuel 5. This system employs two-out-of-thrge coincidence 6. Refer to Section 2.2 and Fig. 2.11 " III. Pump bowl helium pressure greater 1. Causes 1.” System design provides 5.0 £t of overflow capacity, and than 35 psig power level scram and fuel overflow tank level systems &. lzzpidruifil;:ckzgee;cpmgi;; Oioizein ga':;'z b{e:;{g::i (see I and IV, this table) protect against fuel salt inggerm:t ion’ 8 & PP exba.nsion caused by power excursion; safety system holds bypass valves (Fig. 2.14) open during operation, in b. TFallure of hellum off-gas letdown valve, ROV- which case the pressure rise will be moderate until all 522A)., combined with faillure of 40-psig inlet the overflow volume is filled Y - helium reguletion system and rupture disk - B c. Unknown or unidentified mechanisms 2. Redundancy: Two pressure channels, one in the pump : bowl and one in the overflow tank 2. Consequences 3. Testigg: Teet procedure’ checks entire channel, with ex- “ a. If unchecked, dama.ge to pump seals, which are : rated at 100 psig, and possible overstressing of . .celption of transmitter bellows (see Fig. 2. 15) primary loop ' 4. Monltoring: ILoss of helium flow to bubblers caused by b. Exceed capacity of off-gas system, with resulting v either low inlet pressure or line blockage is alarmed threat to contalnment . ¢c. Possible loss of primary containment Sefety only?: Yes i 6. Draining (see Figs. 2.9 and 2.14 and Table 2. 3) ensures 3. Corrective actions impediate opening of the bypass valves to back up ad- &. Drain fuel ministrative control. Venting to the auxiliary char- b. Vent pump bowl to the auxiliary charcoal bed cofenl bed is less effective but useful backup : IV. TFuel pump 6verfl%w tenk level 1. Cause 1. This backs up (anticipates) cause la in IIT (this table) greater than 20% full scale on . : . - Jevel instruments a. Overfill (malfunction in fill system, misopers- 2. Redundancy: Two independent systems provide level in- tion, ete.) formation - . Xpan 1 b; t - b Efi:s?s:e 1 :Iilgnlclafagg‘ere)caused Yy high tempera 3. T__e_SM: Test procedure checks entire channel with ex- ception of transmitter bellows (see Fig. _2.15) 2. Comsequemces 4. Monitoring: Loss of helium flow to bubblers caused by - Loss of capacity to handle fuel expansion or overfill either low inlet pressure or line blockage is alarmed (. 3. Corrective action 5. Saefety only?: Yes Table 2.2 (con%inued) Condition or Situation Which Indicates a Real or Potential Hazard '~ Cguses of the Hazard, the Consequences, and . 1 the Corrective Action Supplementary Information V. Helium pressure in fuel pump bowl greater than 2 psig during fill _ VI. Reactor cell air activity greater than 20 mr/hr VI1I. Coolant pump off-gas actlivity . greater than 20 mr/hr VIII. Control rods not ahove "f£ill" position during reactor filling Caus¢s a. chess £1lling rate or overfill emperature excursion during £111 (see I and II, ’ this table) c. -%aqunction in helium letdown system (lines 522, 324, and PCV-52241, etc., see Fig. 2.14) d. (losing of HCV-533, normally maintained open by gdministrative control, during the filling opera- fion Conseguences of itself, a hazard; & release of positive pres- may produce a sudden rise in fuel level during L and, possibly, & nuclear excursion Corre¢ctive action Draiq fuel Caus¢ 8 'Rupture or leak in the primary containment Consé uences nation of secondary contalnment and increased posglibility of contamination of area in the event tha} secondary containment fails Corrective actions a&. Drain fuel .b. Close in-cell cooling aeir syatem vent (to stack) valve Leak}in primary heat exchanger Consegquences a. This is.a loss in primary containment b. (ontamlnation of coolant salt, coolant salt loop, € components therein, with possible contamina- tion of area in the event radiator or other cool- ant system componehts fail Corréctive action Draiu‘fuel Caumgg Administrative Conse¢quences . Failpre to have rods cocked during f£ill is a loss of protection against the fill accident which can leed to a damaging temperature excursion Corrective action Prohjbit £illing [ This uses same instrumentation as III (this tdble) 1. 2. 3. 4‘ 'Testigg: Redundancy: Two channels are used; one in the pump bowl and the other in the overflow tank Testigg See III, this table ~ Monitoring: See III, this table Safety only?: Yes Redundancy: Two independent channels, either will pro- duce safety actlon Testing: Complete testing of each channel is possible Monitoring: Safetx onlz?: Certain system failures produhe alarms Yes During normal operation the coolant salt pressure in . : , the heat exchanger is greater than that of the fuel * salt; in these circumstances the contamination of the secondary ls that resulting from back diffusion through the leak Redundancy: , Two independent channels, elther will ini- tlate safety acticn Complete testing of each channel is possible Certain system failures produce alarms Monitorigg: Sefety only?: Yes i Redundancy: Action by apy two of the three rods affords protection Testing: This may be tested prior to f£illing Monitoring: A system test immediately before filling meets requirements Safety only?: Seme as I (this table) A Condition or Situation Which Indicates ~ & Real or Potential Hazard 63 Table 2.2 (continued) Causes of the Hazard, the Consequences, and the Corrective Action coedn Supplementéry Information IX. Supply pressure less than 30 psig in helium line 500 which supplies helium to all reactor cell com- ponents, drain tank cell, and fuel processing system X. Activity greater thsn 20 mr/hr in off-gas line XI. Activity greater than 20 mr/hr in helium line %o reactor cell level sensore (bubblers in pump bowl end overflow tank) Reactor cell pressure greater than 2 psig XII. = 1. Causes Loss of supply helium caused by: a. Empty tank b. Previous overpressure which operates relief valve and breaks rupture disk ¢, Malfunction of pressure-regulating valve PCV- 5OOG which mainteins supply at design-point value of 40 psig Consequences Possible loss of secondary containment (see 1b above) Corrective actions T ———— Block all hellum lines to reactor end drain tank cells Causes a. Charcoal beds not operating cbrrectlyf(overloaded, see III sbove) or pump seal rupture allowing dis- charge of activity to lube-oil system b. Activity in the coolant salt loop Consequences Radioactive gases discharged up stack - Corrective actions a. Close off-gas block valves b. Close lube=-0il systems vent valves Causes Reversal of flow in these lines (or back diffusion) from pump bowl, drain tanks, overflow tanks Consequences Radioactivity inside the helium piping in normally safe areas. This activity will be contained as long as piping is not breached Corrective action Close in-cell block valves Causes v a. Maximum credible accident b. Malfunction of cell pressure control system Consequences ! Not, of itself, & direct hazard; during normsl opera- tion the reactor cell is maintained at a negative pressure of —2 peig (13 psia) to ensure inflow in the event of a leak; the existence of positive pressure is eyidence that a malfunction or misoperation exists; a loss of secondary containment is a possible result Corrective actions a., Close instrument air block valves b. Close liquid waste system block valve (from réac- tor cell sump to waste tank) 4. . Testigg 'Redundancy: Safety only?: Redundancy: Three independent channels, any two will initiate safety action Testigg: Complete testing possible Monitoring: Testing meets requirements ‘Safety only?: Yes a . ) i i s' | Redundency: Two independent channels, either will ini- tiate safety action , 'Complete test;ng possible Monitoring: Saféty only?: Certain system failures are alarmed Yes x} Floy rate, in either direction, is limited by capil- laries; check valves, two in series, are used to back upiblock valves Redfindancz Three independent channels; any two will , initiate safety action Testigg: Complete testing of each channel is possible Mofiitorigg: Not applicable Yes Sefety only?: ‘. ~l Three independent chennels; any two will initiate safety action ‘Testigg: Complete testing is possible Monitor;gg: Not appliceble, testing meets requirements Yes il Table 2.2 (continued) Condition or Situation Which Indicates a Real or Potential Hazard auses of the Hazard, the Consequences, and i ' the Corrective Action * Supplementary Information XIII. XIV. XV, Rediocactivity greater than 100 mr/hr in in-cell cooling water system Loss of coolant salt flow Low coolant salt tempersture, measured at radiator outlet Cauj Max; teT es imum credible accident, which ruptures water sys- Consequences a. |Local contamination by radicactlve water lesk b. |Loss of secondary containment, with accompanying ‘leontamination if radiocactive material is dis- charged from the cooling-water system Cau_i;gg a. [Coolant pump stoppage b. |[Line break C ¢. {Unscheduled coolant salt drain d. [Plug in line or rasdiator Congequences Exc#pt for case VI1II ébove; no hazard exists; a flow logs at full power will freeze the radiator in 2 min if no corrective action is teken Corrective action Drop radiastor doors Causes a. {Malfunction of load control system (complex of doors, blowers, and bypass damper) b, |Cessation of power generation in core from any cause (scram, drain, rupture in primary contain- ment, ete.) c. |Loss of coolant flow (see XIV above) Congequences No flazard; warns that potential'radiator freezeup may be developing (see XIV above) - Cortective action Drop radiator doors .Redundancx: Three independent channels; any two will initiate pafety action Testing: Complete testing of each channel is possible- Monitoring: Certain system failures initiate alarms or produce safety action Safetx only?: Yes . ‘ . . , Redundancy: Two direct flow channels which receive information from a common primary element, a venturi, -plus two independent pump speed channels (Fig. 2.13) Testing: Partial testing possible; a test which in- cludes dropping radiator docors will perturdb operation; input elements not tested ' Monitoring: ' By surveillance and comparison Safetx only?: Yes | Redundancy: .Threé independent channels; any two will initiate safety action Testing: A complete test of each input channel is pos- sible; complete testing requires dropping doors, which disturbs operstion Monitoring: A thermocouple break or detachment from pipe will produce safety action in that channel Safety only?: Input channels used solely for safety | 3 - 65 Table 2.3. ct S " i + MSRE Safety System Output Actions e st b A L 14 im0 g . 11 Safety Action Results Produced Supplefientary Notes - Tnitiating Condition I. II' IIT. VII. ( VIII. Rod scranm b ‘Close freeze valve cooling air Resactor efiergency drain 1. Close freeze valve cooling air " valves HCV-919A and HCV-919B 2. (a) Open bypass valves HCV-544, 545, 546 (b) Open vent valves HCV-573, 575, 577 3. ‘Close helium supply valve PCV-517A1 valves HCV-909, HCV-910 5. Close helium supply valves HCV" 572A-1-, HCV" 574Al » HCV-576A1 Radiator door scram Close helium valvyes HCV-572, HCV-574, HCV-576 Close heiium supply. valve PCV-517A in main header serving all drain and flush tanks Close helium supply valve HCV-SBO to fuel storage tank Close valve HCV-516 in helium purge line to fuel pump Close valve HCV-512 in helium purge line to coolant pump .Reduces temperature at which reactor is critical; system is shutdown until new (lower) critical temperature 1is reached .Provides shutdown margin and transfers fuel salt - to a safe locatlion Thaws freeze valve FV-103, main drain valve at bottom of reactor vessel .Bqualizes fill pressure in drain tank and pump bowl pressure Reduces pressure in drain and flush salt tanks by venting to charcoal beds Shuts off pressurizing helium (40 psig) in main supply header to drain and flush salt tanks; this valve is closed during reactor operetion Thaws freeze valves FV-105 and FV-106, which admit fuel salt to drain tanks 1 and 2 Shuts off pressurizing helium (40 psig) to indi- vidval drain and flush salt tanks; these valves ~are closed during reactor operation Shuts off air flow acress radiator to prevent freezeup Preserves containment; shuts off supply of pres- surizing helium to individual drain and flush tanks and blocks these lines to prevent escape of radiocactive gases Preserves containment; shuts off 40-psig helium to drain and flush tanks and blocks escape of radicactivity or reverse flow from the drain tanks Preserves containment Preserves contalinment; blocks line 516 and pre- vents escape of radicactivity from containment Preserves containment; blocks line 512 and pre- vents escape of radiocactivity from containment Redundancy obta Action by any two of the rods is sufficient; rods tested at %ach shutdown N Complete emergency drain system may be eration 1f test period tested during does not exceed 5 min These valves ar must cperate Bypass valves ax by administrati closed to test The vent and b: drain tank fo I a redundant palr and with checkzvalves- Administrative dpntrol is used to keep one open during operation and there are two empty drain of these valve% to ensure that tanks to receiwe fuel See II.3 above;: closed and p 2 psig - g Complete testin% possible but will perturb pover generation i Redundahcy obta: ed in combination with V (below) and with check valves | | Redundancy obtajned in combination with IV (above) and with check valves 1Hfiw_wvwe;efi§§ i { Q . Bagic contalnment requirements met by this valve and check valves in series; ESV-516A1 and 516A2 provide redundancy Redundancy; 2.bérriers, provided by check valves plus hesit exchanger . { a redundant palr; only one kept open during operation e control and hence may be ‘4' Radicactivity 1 8s valves serving each ed in combination with individuel supply valves (see II.5 below) %ay be tested during opera- tion if PCV—511A is closed and if pump bowl pressure reduced to less than 2 psig; PCV- 517A may be tested if these valves are bowi pressure is less than 1. Flux signal, po Mw 2. Fuel salt outlet greater than 13 1. High reactor cu 2. Fuel pump bowl than 50 psig 3. Fuel salt overfl greater than 2 tainment cell 5. Redloactivity i coolant pump 6. During filling, r greater than 15 temperature O°F et temperature essure greater w tank level secondary con- T off.gas from el pumfi bowl pressure in excess of 2 psig 7. ‘During filling, drawn - Low coolant sald 2. Low coolant salt rods. not with- temperature flow l. Low helium suppl {less than 30 2., Fuel pump bowl excess of 2 psi 3. All shim rods ng drawn 1. Low helium suppl] Yy, pressure ig) essure in t with~ y pressure (less then 30 peig) 2. Fuel pump bowl p excess of 2 psi 3. A1l shim rods no drawn Low helium supply pr than 30 psig) Low helium supply pr than 30 psig) Low helium supply pr than 30 psig) regsure in £ t with- pssure (lees pssure (less basure (less 66 < Table 2.3 (continued) XII. XIII. XIV. XVII. Safety Action Block helium supply lines to bub- bler (level instruments in reac- tor cell) 1. Close valves HCV-593BL, HCV-593B2, HCV-593B3 2. Close valves HCV-599El, HCV-59932, HCV-599B3 3. Close valves HCV-595Bl, HCV-3595B2, HCV-595B3 Close walve HCV-511A1 in helium supply line (40 psig) to coolant drain tank Close HCV-557C; off-gas line from charcoal beds, coolant salt pump, coolant salt drain tank, and fuel end coolant pump lube oil systems- Close lube-oil systems vent velves PCV-510A2 (cool pump system) and PCV-513A2 (fuel pump system) Close valve PCV=-565A1 in yent line from reactor cell evacuation line N Close in-cell cooling water block valves: 1. FSV-837-A1 2. FSv-846-A1 3. FSV-846-A2 4. FEV-847-AL i 5 . FSV— 844"Al 6. ESV-STmA Close liquid waste system'block valves 1. HOV-343A1 and HCV-343A2 2. HCV-333A1 and HOV-333A2 Close instrument air line block valves Opefi fuel pump bowl vent valve to “auxiliary charcoal bed ‘Results Produced Preserves containment; prevents reverse flow of helium in lines to bubblers in tqe pump bowls and in the overflow tank Preserves containment i i Preserves containment; blocks helium flow which normally passes out off-gas stack 1 Preserves containment; blocks helium flow which normally passes out off-gas stack | | \ Preserves containment; blocks component cooling alr flow which would normally pass out through off-gas stack S-1 Preserves contalnment Blocks outlet line from drain tank cell space cooler Blocks outlet line from reactor cell space cooler No. 1 Blocks outlet line from reactor ce}l space cooler No. 2 Blocks outlet line from thermal shdeld -and fuel pump motor Blocks inlet line to thermal shiel? and fuel pump motor ‘ Blocks vent line from surge tank | Preserves containment; blocks reaétor cell and drein tank cell sump ejector disc rge lines to liquid waste stor&ge tank Preserves secondary containment Preserves containment Supplementary Notes Redundancy provided’by two check valves in series with each block valve Redundancy firovided by charcoal bed plué valves PCV-510A2 and PCV-513A2 (see XIT . below) Redundancy provided by valve HCV-557C (see XI above) and by pump seals; these valves - used for both control and safety Redundancy provided by reactor vessel; this valve forms part of secondary containment barrier ' Redundancy of valves 1 to 6, inclusive, provided by ESV-ST-A (6, below) This valve provides redundancy per above These valves normally closed during opera- tion; 1 and 2 are both redundant pairs Redundancy provided by pfimary containment vessel walls Initiating Condition 1. Low heliunm sfipply pressure (less than 30 psig) 2. Radiocactivity in lines to fuel salt level instruments (primary sensors) Low helium supply pressure (leés-fhan 30 psig) Radloactivity in off-gas line Radiocactivity in off-gas line Radioactlvity in reactor cell air Radiocactivity in cooling water system Reactor cell pressure greater than 2 peig Reactor cell pressure greater than 2 psig Fuel pump bowl pressure exceeds 35 psig 67 Teble 2.4, Operating Ranges of MSRE Nuclear Instrument Channels Fumber of Normal Operating Channel Type Detector Type Channels Range Wide-range Fission chamber 2 Source to five counting times full power Linear power Compensated ioniza- 2 15 w to 15 Mw tion chamber Flux safety Noncompensated 3 200 kw to 20 Mw ionization chamber The sensors used in the flux safety channels are uncompensated neu- tron-~sensitive ionization chambers. The linear Tflux amplifier in each channel produces an output signal of O to 10 v that is proportional to the ionization chamber current. The gain of the flux amplifier and the cham- ber locatlon is Tixed so that at design-point power the lux-amplifier output is 5 v; the setting of the scram point of the fast trip comparator is fixed at 7.5 v. During the early phases of MSRE operation, the scram point will be reduced by changes in either gain of the amplifier or chamber position or both. These changes are not routine and are not at the dis- cretion of the operator; they will be accomplished under administrative control and with suitable testing after they are made. The temperature channels, described in the next section, monitor the reactor outlet salt temperature over a range extending tc 1500°F. The scram point is adjustable about the nominal 1300°F presently intended but only under administrative control. A scram channel combines one temperature and one flux channel to pro- vide a trip 1if edither temperature or flux exceeds its scram point. The three scram channels thus Tormed are then arranged so that signals from two of the three scram channels are required to release the safety rods. Relay contacts in the relay safety element are used as the logic elements, with a separate two-of-three matrix being used for each rod. The coinci- dence matrix monitors are used to display the operation of the relay con- tacts in the matrices during operation, as well as during tests. 68 The flux and temperature safety channels are tested and monitored in service in the following ways: 1. A voltage ramp may be applied manually to the system that causes a current to flow through the leads from the ion chamber to the input of the flux amplifier. Channel response is checked by observation of panel meters. 2. A steady-state current may be applied manually to the input of the flux amplifier. Channel response is observed on panel meters or, if the applied current is large enough, it will trip the channel undergoing test and the trip so produced may be verified by observing the operation of relevant relays in all three relay matrices. The operation must reset the scram circuit after a test or, in fact, after any scram signal. Reset cannot be accomplished so long as the scram signal exists. On-line test- ing of each temperature channel is effected in a similar manner, except for the method of introducing the test signal (described in detail in a following section). Since the flux and temperature safety channels are in two-out-of-three coincidence, testing will not produce a scram. 3. A voltage loss at the ion chamber terminals will produce =z safety trip in the channel. This monitoring action is automatic and continuous, and a trip so produced is indicated by a lamp on the front panel. Throughout the nuclear safety system the three instrumentation chan- nels have been isolated from each other and from circuits used for control. Theilr components are ldentified as éafety devices and, as safety devices, are subject to the strictest administrative control in their maintenance. 2.2.2 Temperature Instrumentation for Safety System Inputs Both rod scram and radiator door scrams require reliable temperature input information. Figure 2.11 gives the details of one of the three identical channels used to provide the required three independent tempera- ture signals to the nuclear safety system shown in Fig. 2.10. The same type of system is used for the temperature input signals to the safety equipment that scrams the radiator doors. The interconnecting wiring of each channel is in its own conduit, and conduit runs are identified as part of the safety system. ZXach 69 UNCLASSIFIED ORNL-DWG 64-626 CURRENT ACTUATED SWITCHES TEST ASSEMBLY: T SaF roay CONTROL ROD a. ELECTRICAL OSYSSTEENTY L REVERSE r HEATER SEE FIG. 2.40 T0 | | b. TEST THERMQ- DATA | |~ COUPLE LOGGER | | (SEE NOTE 1.) URRENT c ) | /ACTUATED [ T SWITCH Q EmF_ TO 7 ISOLATION 7 CURRENT | AMPLIFIER 100 Q MEASURING CONVERTER — THERMOCOUPLE Lo ot t ° 400 0 L0 * o & I—— 10-50 ma 10 70 50 me METER CHANNEL NO.Z2 N0 SAME AS ABOVE CHANNEL NO.3 NOTE: {. POLARITY SHOWN FOR TEST THERMOCOUPLE 1S THAT USED TO SPARE TEST SYSTEM RESPONSE TO A : TEMPERATURE INCREASE e REACTOR OUTLET PIPE LINE NOQ. 100 AND RADIATOR OUTLET PIPE, LINE 202 Fig. 2.11. Typical Temperature-Measuring Channel Used in Safety bysten. channel is monitored, logged, alarmed, and pericdically tested during op- eration. Monitoring of the safety thermocouples is provided by the burnout feature of the emf-to-current converters and comparative survelllance of the three channels. If a thermocouple breaks or becomeg detached from the pipe wall, the burnout protection will cause the channel to fail toward safety, i.e., upscale. Wire-to-wire shorts or wire-to-ground shorts at points away from the thermocouple junction are detected by comparing the readings of the three channels. Operational testing of the individual channels is accomplished by use of a thermocouple test assembly (see Fig. 2.11), which consists of a small heated thermocouple assembly connected in series alding with the reactor thermocouple. When the thermocouple heater is off, the test thermocouple is at the same temperature throughout its length, and its net generated emf will be zero; however, when the heater is on, an emf will be generated that will add to the emf of the reactor 70 thermocouple and cause the alarms and control circuit interlocks associ- ated with that channel to operate. The use of two-out-of-three coinci- dence circuitry permits individual channels to be tested without initiat- ing corrective safety action. The thermocouple test unit described above satisfies the requirement that operational testing shall not disable a safety channel during the test and has the advantage of simplicity, re- liability, and low cost. The radiator thermocouple test is similar, except that the test thermocouple is connected in series opposition so that heating the thermo- couple reduces the net emf and simulates a temperature reduction in the radiator outlet pipe. 2.2.3 Radiator Door Emergency Closure System The radiator doors are dropped automatically if conditions indicate that there is danger of freezing the salt in the radiator. An analog computer calculation has shown that a loss of coolant salt flow, with the reactor at full power, would freeze the salt in a radiator tube in 3 The calculation assumed completely stagnant salt not less than 44 sec. in the radiator and made no allowance for time required for flow decelera- tion or for natural circulation if the particular situation permitted it. Two types of input information (see Table 2.2) are used to initiate closure. The low-temperature signal, measured at the radiator outlet, is the primary indication that remedial action is required, regardless of the cause. On loss of flow, where temperature measurements become meaningless, the doors are dropped on low-flow signals from (1) the venturi meter in the coolant salt loop and (2) the pump speed monitors. (The temperature input information safety channels are described in the preceding section. ) Also, individual radiator tubes are equipped with thermocouples which are con- nected to the temperature scan and alarm system. Coolant salt flow is measured using a single venturi in the coolant salt pipe loop. With the exception of the venturi, two input channels are provided. 1In a typical channel, differential pressure is converted into 33. J. Ball, "Freezing Times for Stagnant Salt in MSRE Radiator Tubes, " internal ORNL document MSR-63-13, April 19, 1963. 71 an electrical signal by a NaK-filled transmitter followed by an emf-to- current converter whose output is used to actuate alarm switches that serve as trips to actuate the clutch and brake in the drive mechanism. Coolant pump speed is measured by an electromagnetic pickup mounted very close to the pump shaft coupling. Teeth, similar to gear teeth, are cut on the periphery of the coupling, and the speed pickup generates a voltage pulse as each tooth moves past it. The count rate, as determined by the speed monitor, is thus an indication of pump speed and is used as an input for alarm and safety. The two flow signals and the speed signal are fed to a two-out-of- three coincidence matrix such that 1f any two reach unsafe values they will cause the doors to drop. The two flow channels are monitored by comparison, one with the other. A loss of flow signal is simulated by shunting a resistor across one leg of the strain-gage bridge in the pressure transmitter and observing the action of the relays in the coincidence circuitry. The pump-speed channel is fested by using the calibration switch on the speed monitor and observing relay response in the coincidence cir- cultry. Oscilloscope display of the input pulses affords a check on the integrity of the sensors and the wiring to the speed monitors. Low speed is annunciated. The combined system is diagrammed in Figs. 2.12 and 2.13. Interlocks are used to turn off the blowers when a safety signal to close the radiator doors is received. The radiator heaters are maintained "on" at all times when there is salt in the coolant loop. Heater current is monitored by operating personnel as a routine administrative procedure,. 2.2.4 Reactor Fill and Drain System A simplified diagram i1s shown in Fig. 2.14 of the reactor vessel, the drain tanks, the interconnecting piping, and the control elements re- quired to fill and drain the reactor system with fuel salt in a safe, orderly way. The control rods are ecssential to a safe filling procedure but are omitted from this diagram in the interest of simplification. The reactor is filled by applying helium pressure to the gas space in the selected drain tank and forcing the molten fuel salt up and into the 72 UNCLASSIFIED ORNL-DOWG 63-8390 ELECTRO- ELECTRO- /MECHANICAL BRAKE MECHANICAL / OVER-RUNNING CLUTCH Iuj IUJ// SHEAVES —— CLUTCH: - d’//FLYWHEEL A\ie | K K | B | £ \ = X0 ORIVE MOTOR WITH GEAR REDUCER. SINGLE MOTOR USED TO DRIVE BOTH DOORS~— SPRING SHOCK ABSORBERS A8 NOTE: THIS ARRANGEMENT 15 TYPICAL AND IS USED FOR BOTH DOORS RADIATOR DOOR 1770 b Fig. 2.12. Radiator Door Emergency Closure System. core vessel. As a typical example, if the reactor is to be filled from fuel drain tank No. 1 (FDT-1), pressurizing helium is admitted via lines 517 and 572 (see Fig. 2.14). The filling pressure is controlled by pres- sure-regulating valve PCV-517A. The pressure setting of this valve is controlled by the operator. The upstream capillary in line 517 limits the maximum flow rate in this line., Valves HCV-544 and HCV-573 in pipes which connect the gas space in FDT-1 to the pump bowl and to the charcoal beds are closed., Since the net pressure head available to produce flow de- creases as the fuel level rises in the core vessel, the fill rate becomes progressively slower as the fill proceeds if constant inlet helium pres- sure is maintained and if the pump bowl pressure remains constant. Helium pressure in the pump bowl is the second component of the dif- ferential pressure that drives fuel salt into the core vessel. A sudden reduction in pump bowl pressure during the reactor filling operation would cause an unscheduled rise in salt level in the core. If, at this time, the reactor were on the verge of becoming critical, an unexpected nuclear COOLANT PUMP - MOTOR : o MAGNETIC PICKUPS PUMP, SPEED]| S Z# INSTALLED ;// . - r—’ MONITOR SPARE, [veer} — ] o | PUMP SPEED MOTOR TO PUMP _ MONITOR SHAFT COUPLING VENTURI LINE 201 pume sowL (___ 7 LINE 202 HEAT - EXCHANGER THERMOCOUPLES —____ | DIFFERENTIAL | PRESSURE TRANSMITTERS . SAFETY _ | CIRCUITRY . $10R Sp Sy OR. Sy * INQEZE?QENT COINCIDENCE , A : Fy RELAY - et {0)2 FLOW — ul MATRIX F (b} SPEED £, il —e] __—RADIATOR . CIRCUITRY SEE FiG. 2.1 U THERMOCOUPLE oo CURRENT ACTUATED SWITCHES ] 3 EMF TO CURRENT CONVERTERS | Fig. 2.13. | iy H | ‘--—*—}3 A" w UNCLASSIFIED ORNL- DWG 63-8389 \'. SIGNALS COtNCIDENCE] RELAY MATRIX T INDEPENDENT [ ~ 15 ] TEMPERATURE {_Tg_ (2 OuT OF 3) (— NOTE: EMERGENGY GLOSURE PRODUGCED BY INP SIGNAL COMBINATIONS AS FOLLOWS: (a) S, OR S, + F, (b) S, OR S+ F, (e) Fy+ Fy (d} ANY TWO OF THE THREE TEMPE,RATUR'E INPUT SIGNALS . 4 ALARM - . Pump-Speed Monitoring System. L= 70 READOUTS —= LOGGER ETC, UT . CLUTCH SEE FIG. 2.12; ) CLUTCH BRAKE UPSTREAM DOOR =2 3 E BRAKE DOWNSTREAM DOCR 74 System Conditien PresFill Gpocete Yeive No. o Norm Draln Inte with Clre, Nermst Fill feam Fill from Fill from Ewerp. Helium Ha Opsrstion FOT#1 FOT#2 FFT FOTAl FDT#2 FFT Dewin PCY S, Fill prons o x x 0 0 o x X X HCV 873, Fill, FOT »1 x X x o x x x x X x P HCV 574, Fill, FOT #2 x x % x o x X x X x ‘ FUEL PUMP HCY 378, Fill, FFT X x X X X 0 X X X X f—\ ‘ BOWL OFF-GAS HCY 344, By pass o o o X 0 ° o X E o - .5 (3%2) 320+ HOLDUP AND HCV 345, By po 0 x E E o ' P f— H i HCV 546, : :.:: g g :. 0 0 : E g o o INLET}HELIUM ) @ COO_LER NCY 573, V. X x x x x x o. o LINES - - l i J_'_i 140 liters , Yeat ‘ SEE|FIG. 217 =TT 1 , Vo X x X PT HCY 577 Ve X ox o x xax . % : > 539 600 : — HCV 523 HCV 533 Al o x 0 o o x X X x : $ (589) ° A S T A | ——e— || | ] _ HCY 910 Al o o o X o X X 0 X X | ; ‘AR HC 519 Asn' o X x X x x x x x x 1 : | OVERFLOW ( Hev 528t 0 0 o o 0 o . o o o o ‘ | ! ! Loy TANK pevsna gl 5 5, 0 o 0 o o o o o INPUT TO — —= EMERGENCY DRAIN e S xke % o cx X 8 1% . | SAFETY SYSTEM | __ OPENWCV___ . _ . - FV 106 X X OwX* 0O XX ° X x o ! SEE FIG. 2.15 533 " iy 07 x x X x X X X X ; ‘ :: 108 : : : x X X X X x X X BYPASS LINE =20 fs—s? FY 109 : X x X X X X X X X - N X -:lou‘:o-o;-n;‘! = sither span or cloasd, . . '_ *During nermel oparation wither, (but net beth) ¥V 105 or FV 106 is apen. . . ) ;HCV 518 AL B eponed after comploting fill and betore sporation to ensure thet drein line 103 is clesred, ) . THCY 523 clesed enly when H-‘rln‘ solt gut of averilow tenk, BYPASS . BYPASS INITIAL ! LOADING V REACTOR CONNECTION M) VENT =5 VESSEL ' LINE HCV ( HGV HoV ST . 546 3 HCV ° &19 x HCV HCV } Py FV103 919A 9198 {103) 7 ; FV103 ; io TO FUEL . FV106 FVIOS 1 FLUSH TANK @ v I : } Fvi08 TO FUEL _ FV107 e D . (108)- TRANS. AND Jla | No2 |FLUSH, CAPILLARY ‘ Tfig:‘ Sl So | ‘ TANKl 1 FLOW RESTRICTOR o\, | Nt o T z ! o | | HELIUM WE st | i 3 > 576 S x HIx & X SUPPLY, w g 40 psig : {(575) FREEZE VALVE 5198 5194 Fig. 21 COOLING AIR 14. Reactor Fill and Drain System Valving UNCLASSIFIED ORNL-DWG 63-8395 PCV 522 A4 TO CHARCOAL BEDS TO AUX. (G61)~= CHARCOAL BEDS f FREEZE VALVE Lcoon.ms AlR 75 excursion could not be ruled out. The safeguard provided is tc allow filling only when the positive pressure in the pump bowl is equal to or less than 42 psig and thus keep the maximum possible increase in net Till- ing pressure within safe limits. Two causes Tor a sudden decrease 1n pump bowl pressure are considered. First, opening valve HCV-533 after reactor filling has started would vent the pump bowl to the auxiliary charcoal beds, which operate at close to atmospheric pressure. The resultant pres- gure change in the pump bowl would be -2 psi. Administrative contrecl isg used to maintain HCV-533 open just before and during filling. Second, a rupture or leak would allow the escape of pump bowl helium to the reactor cell stmosphere, which is held at =2 psig (12.6 psia). This condition would cause a maximum pressure decrease in the pump bowl of 4 psi. If the fuel pump bowl pressure exceeded 2 psig, the safety equipment (1) would initiate a drain by opening the bypass valves, HCV-544, -545, and -546, and equalizing helium pressure in the drain tanks and pump bowl and (2) would shut off the supply of pressurizing helium. This is the helium valve condition shown on Fig. 2.14 and is required by normal operation with the pump bowl at 5 psig; therefore, the 2-psig channel need not be disabled during normal operation. Additiocnal safety considerations in- volving pressure in the fuel salt system and the instrumentation used for measuring helium pressures is discussed in the following section. During reactor operation helium pressure in the pump bowl is main- tained at 5 psig by means of throttling valve PCV-52241 (Fig. 2.14), which controls the flow rate of off-gas from the pump bowl. This valve is actuated by a signal from a pressure transmitter, PT-5224, on line 592, which admits helium to the pump bowl. Excessive pressure (50 psig) in the pump bowl is relieved by opening HCV-533. During reactor operation, with the bypass lines open, one or more of the drain tanks would be subjected to this pressure. Considering the gas volumes involved, a 50-pslg overpressure in these circumstances is of questionable credibility. The input signal used to open HCV-533 originates in either of a redundant pair of pressure transmitters, PT-592B or PT-589B, on inlet lines to the pump bowl and overflow tank, respec- tively. Thege same input channels provide the 42 psig safety signal dis- cussed in the preceding paragraph. 76 As a Turther safeguard during filling, the system requires that all three control rods be partially withdrawn in order to pressurize the drain tanks. The weigh cells on the drain tanks provide information used to monitor the filling rate and the total amount of fuel salt moved into the core vessel. The possible filling accidents are discussed in Section 7.1.4. Briefly, these accidents are (1) premature criticality during filling caused by an overly high concentration of uranium brought about by selec- tive freezing in the drain tank; (2) premature criticality during filling of low-temperature (200°F) fuel salt of normal concentration, whose normal critical temperature is 1200°F; and (3) premature criticality during fill- ing of normal fuel salt at normal temperature with all control rods fully withdrawn. Protective action is the same for all three cases: (1) the control rods are scrammed, and (2) the reactor vessel is drained to ensure permanent shutdown. Rod scram is produced and the vent valves HCV-573, -575, and =577 are opened by the excess flux signal. ©Since the criti- cglity would be occurring before loop circulstion was attained, the cut- let temperature sensors would be ineffective. The safety system also invokeg a reactor emergency drain to enhsznce ccntainment if there is evi- dence that radiocactivity is escaping from the primary fuel loop. These subsystems are covercd in g subsequent section. Emergency drainage is effected by the following actions (see Fig. 2.14): 1. The freeze valve in line 103 which connects the reactor to the drain tanks is thawed. Thawing is accomplished by closing valves HCV-919A and -919B, a redundant pair, to stop the flow of cooling air. The systen is designed with a heat capacity sufficient to thaw the plug ixn 1> min. 2. The helium pressures in the fuel drain tank and the uniilled portion of the fuel salt loop are equalized by opening bypass valves HCV- S4de, =545, and -546. 3. Vent valves HCV-573, -575, and -577, which release pressurizing helium in the drain tank to the off-gas system, are opened. 7' 4. Pressure regulating valve PCV-517A1 in line 517/, the inlet header that supplies pressurizing helium tc all the drain tanks, is closed and shuts off the supply of pressurizing helium. 5. The drain tank pressurizing valves HCV-572, -574, and -576 in the helium supply lines are closed to halt further addition of filling pressure. Acticons 2 or 3 are immediately and independently effective in revers- ing a fi1l1l, and hence the valves are redundant. Similarly, PCV-517A2 and the individual pressurizing valves, 4 and 5 above, form series pairs and provide redundancy. For example, it can be seen from Fig. 2.14 that, taking fuel drain tank No. 1 as typical, pressure equalization and vent- ing are accomplished by opening HCV-544 and HCV-573, respectively. Ad- ministrative control is used to ensure that, when the reactor is filled with fuel salt, both drain tanks, FDT-1 and FDT-2, are empty and that one of the freeze valves, FV-106 or FV-105, is thawed. The tank condition is monitored by reference to the weigh cell and level instrumentation on each tank. 2.2.5 Helium Pressure Measurements in the Fuel Salt Loop One channel of the instrumentation which measures helium pressure in the fuel pump bowl and the overflow tank is shown in Fig. 2.15. The com- ponents and their installation are designed to meet contaimment require- ments. Since the secondary containment is held below atmospheric pres- sure and since the temperature therein is not closely controlled, it is necessary to use a variable-volume reference chamber to maintain atmo- spheric pressure on one side of the measuring bellows in the transmitter. Two channels are used: one in the pump bowl and the other in the overflow tank. These normally operate at the same pressure. Any safety signal from either channel initiates the appropriate safety action. This is discussed in the preceding section and outlined in Tables 2.2 and 2.3. The system may be tested periodically during reactor operation by (1) Observing system response to small operator-induced pressure changes and (2) by shunting the torque motor in the pressure transmitter. These tests will establish that, in the channel undergoing test, the lines from NOTES: i, OPERATION OF THE TEST SWITCH SIMULATES THE APPLICATION OF PRESSURE TO THE TRANSMITTER AND GCHECKS ALL COMPONENTS IN THE SAFETY CHANNEL EXCEPT THE CONVERSION OF PRESSURE TO FORGE BY THE TRANSMITTER BELLOWS. 2. THE 2 psig SAFETY CHANNEL IS IN THE ACTUATED MODE DURING NORMAL REACTOR OPERATION UNCLASS!FIED ORNL-DWG 64-627 TO SAFETY SYSTEM SEE F!G. 214 AND TABLE 2.2 ,SINGLE ALARM SWITCH, SET AT 40 psig ~DUAL ALARM SWITCH 7 SWITCH SET TO PRODUCE 2 SAFETY INPUT SIGNALS \ aLarM | — VENT TO STACK CONTAINED. ZERG 1 IF PRESS IN LINE EXCEEDS o e i | . . . - , AN psig REFERENGE J%N‘ | oW 2psig AND 50 psig J \ . CHAMBER. RATED | g2 N TO DATA N AND TESTED AT 100 psi | S0 psig; 2ps LOGGER FROM FUEL SALT SLACK M ! ps19 T OVERFLOW TANK PRESSURE DIAPHRAGM ~ l . o TRANSMITTER L TEST SWITCH , o ~ rrfTer,, 1 | SEE NOTE f o] S 5~ METER E A [=10TO 50ma = 4000 ‘~/)1 ' 1 E R ~ =10 TO 50 ma 2eh T v T WELD SEALED PRESSURE " ISOLATION AMPLIFIER INLET HELIUM * 592> TO PRESS TRANSMITTER USED FOR CONTROL TO OVERFLOW TANK, INLET TRANSMITTER WiITH PRESSURE- TO-CURRENT TRANSDUCER-AMPLIFIER, RATED AND TESTED AT 250 psig REFERENCE LINE TO FUEL PUMP BOWL TO FUEL PUMP INSTRUMENTATION SAME AS ABOVE HELIUM Fig. 2.15. the Primary Loop. BOWL GAS SPACE Typical Instrumentation for Measuring Helium Pressures 1in 84 79 the pump bowl (or overflow tank) are clear and that the electronic equip- ment and associated wiring is capable of operating the output relays. Since it takes 5 min to thaw drain valve FV-103, this time can be used to observe the response of the thermocouples on the freeze valve as 1t heats up but before actual salt flow begins. The valve can then be refrozen before an actual drain is initiated. In such a test, valve HCV-533 will be opened, and the regulated pressure (5 psig) in the pump bowl will be lost for the duration of the test. The response of HCV-533 can be noted by observing the actuation of the position switch on the valve stem. This test procedure checks the entire safety channel, except the ability of the transmitter measuring bellows and the associated linkage to transmit the pressure. Monitoring of these channels is accomplished by indicating and alarm- ing the pressure downstream from the hand throttling valves in each helium line serving the primary elements (the bubbler tubes) in the pump bowl and in the overflow tank (see Fig. 2.17 in sec. 2.2.7). A downstream flow stoppage by a blocked line is indicated by a high-pressure alarm; low flow caused by a loss of upstream (supply) pressure, an upstream blockage by foreign matter, or a hand valve closure is indicated by a low-pressure alarm. The fuel level in the overflow tank is measured by the differential pressure across the helium bubbler probe which dips into the fuel salt in the tank. The degign criteria for testing and monitoring these differen- tial-pressure-measuring channels are the same as those which guided the design of the pressure channels described in the preceding varagraphs. The reference chamber is not required. Two channels are employed for safety and either will initiate a reactor drain if the fuel salt level in the overflow tank exceeds 20% of full-scale level indication. 2.2.6 Afterheat Removal System The drain tank afterheat removal system, typically the same for both drain tanks, is shown in Fig. 2.16. Once placed in operation the evapora- tive cooling system is designed to be self-regulating and to operate with- out external control., Reliable operation of the afterheat removal system 80 requires (1) that the feedwater tanks contain a supply of cooling water, (2) that an ample supply of cooling water is available to the condensers, and (3) that the system includes reliable valves to admit feedwater to the FROM COOLING TOWER COOLING WATER TO OTHER COMPONENTS TO CONDENSATE STORAGE TANK HCv-882 c < FROM PROCESS WATER MAIN TOWER 70 COOLING =—@56T—— DRAIN TANK CONDENSER NO.1 Fig UNCULASSLEFIED ORNL—-DWG 63—8388 TO OFF-GAS STACK RUPTURE g/ DISKS MOTOR OPERATED VALVES HAND VALVE 560 TO VAPOR CONDENSING | SYSTEM o a 812 = TO DRAIN TANK ~ CONDENSER NO.2 FEEDWATER TANK NOA < ® 878 < — CONDENSATE 3 T ¥ ESV-806 0 PN T 7 et A L LT iy ] | | | A STEAM | x' ORUM | i 1N f = E -—g‘j - ——-—-‘-HA \ ::v' S DRAIN TANK S CELL= — ~wy k i’/‘/ ; FUEL é - DRAIN - TANK | 7 ., . 2.16. Afterheat Removal Systemn. 81 steam drums. Administrative control is relied on to ensure that the feed- water tanks contain water. Valve ESV-806 opens to admit water automati- cally to the steam drum when the salt temperature in the drain tank exceeds 1300°F and recloses at some lower temperature to be determined experimen- tally. This valve is in parallel with manual valve LCV-806A, and the pair are a redundant means of valving water to the steam drum., Normally the condensers are cooled by tower cooling water, but diversion valve HCV-882Cl provides an alternate supply of condensing water. Loss of tower water is detected by pressure switch PS-851B, which operates diversion valve HCV- 882C1 and supplies condenser cooling water from the process water main. Since it takes over 12 hr for excessively high afterheat temperatures to develop after a drain, there is sufficient time to effect a transfer to the other drain tank in the event of a failure or malfunction. This is also ample time to make connections from the ccoling water system to a tank truck in the event that the normal water supply is i1noperative,. 2.2.7 Containment System Instrumentation Contaimment requirements are met by providing at least two independent reliable barriers, in series, between the interior of the primary system and the atmosphere. For example, the two-barrier concept is fulfilled by: 1. Two independent, reliable, controlled block valves with independent instrumentation. 2. One controlled block valve plus a restriction such as a charcoal bed which will limit the escape of activity to the stack to less than the maximum permissible concentration. One controlled block valve plus two check valves. solid barriers (vessel or pipe walls ), One solid barrier and one controlled block valve. One solid barrier and one check valve. o W oW = O The general considerations outlined in this section apply to the instrumentation and control equipment used to operate the block valves. Rlock valves are not located at such a distance from the containment pene- tration that the lines become tenuous extensions of the containment ves- sel., Valves and other devices used in lieu of solid barriers will be 82 routinely tested and demonstrated to be capable of maintaining leakage below the specified tolerance when closed. Helium Supply Block Valves. A diagram is presented in Fig. 2.17 that shows the helium supply lines tc the primary containment vessel and the associated valves used for contrcl and blocking these supply lines against the escape of radicactivity from the primary system as a result of reverse flcw or back diffusion. Two types of input signals are used tc initiate block-valve closure (see Tig. 2.9)., The first, a reduction in the supply pressure from its normal value of 40 psig to 30 psig actu- ates pressure switches and closes all the inlet helium bleck valves. The second, excess radiatior in any of the helium lines supplying the level probes (bubblers) and pressure-measuring instruments in the pump bowl and overflow tank, closes the block valves in these lines. A reduction in helium supply pressure in line 500 from its regulated value of 40 psig indicates a leaky rupture disk or leaky piping and a loss of primary con- tainment. In-service testing cf the loss of pressure channels is accomplished by cpering the hand valves on the lines to the pressure switches on the main supply pipe (line 500), one at & time, and observing the action of the relays irn the twc-oui-of-three coincidence matrix in the control room. Actuzl block-valve closure is not tested with this procedure. Low- and high-pressure alarms are prcvided on line 500; these will actuate before the safety system pressure switches close the blcck vaives. Additional testing of the solenoid block valves in the helium bubbler lines is provided by clesing each velve individually and cbserving the pressure change downstream of the hard throttling valve, The three radiation-monitorirg safety input channels, RM-596A, B, ard C, are tested by exposing ezch irdividual radiation element to a source and noting the response ¢of the cutput relays which produce valve closure. Since two-cut-of-three coincidence 1s used, this test willl not perturb the system; nelther dces it provide a valve-clcsure test. Off-Gas System Monitoring and Block Vslves. The off-gas systemn, Figs. 2.18 and 2.19, is monitored for excess radiation in four places: 83 TO FUEL . TO psig ALARMS - ¥ UNCLASSIFIED ORNL-DWG 63-8398 TO COOLANT SALT - PUMP BOWL BUBBLERS PRESSURE GAGE - AND :IPRESSURE SWITCHES WITH HIGH AND TO LEVEL PROBES, (BUBBLERS) IN FUEL SALT OVERFLOW | TANK.SEE FIG.2.14 - TO LEVEL PROBES { BUBBLERS) IN FUEL SALT PUMP BOWL. - SEE FIG™2.14 . PUMP— > ° @ LOW PRESSURE ALAI‘\;M 1S TYPICAL OF ALL BUBBLER LINES CESV % ESV : _ ' 516 A1 § § si6a2 — & MoV ] * ¥ 59983 BI&—D} - == @ i K ’ I £t SHAl ggA?%O%:NNJ ' - f * q: so3e1 | 511 I~ e - . D<= (598 Y s 1 T | e o | ' * 1 59382 312 T~ - TO COOLANT PUMP >pg——G2 : pk L= > _ * el W50 | : : ETN ] (50D -GoD /I X - CLOSES : 7 o P BLOCK VALVES -SEE NOTE 3 | (deos ) (522 )(522) SEE NOTE | f : A B Y 2/ ] " TO FUEL STORAGE L 2 OUT OF 3 \ @ A TANK AND CHEMICAL COINCIDENCE @ ‘ @ PROCESSING SYSTEM SNUBBER SNUBBER r'_‘ 3 INDEPENDENT ~ b SNUBBER VENTED TO OFF-GAS SYSTEM DUCT (832 (TYPICAL) EXCESS FLOW RUPTURE DISK @ [ s0esig SIGNALS ' THREE CHANNELS FOR INPUT SIGNAL. PRESSURE SWITCHES ACTUATE AT 35 psig — — ——y — — HELIUM SUPPLY REGULATED AT 40 psig - > PCV | . ; i 6054 MATRIX PCV *i S17TA X REFERENCE CHAMBERS SHOWN ON FIG. 2.5 % 4 1 1 1 INDIVIDUAL TESTING | CAPILLARY , OF EACH CHANNEL | RESTRICTOR v / AND ARE SECONDARY CONTAINMENT BARRIERS BY OPENING HAND | * | f ‘ = - g VALVES T mov DTN (572 TO FOT- . - I 574 A1 | \ ‘ , A 2 OUT OF 3 E]; q & 2 COINCIDENCE Aoy RN &3 = TO FOT-2 |, MATRIX @ 5T6A1 : * |F ——i—x =N G790 T0 FFT ESV ‘ 5194 : CLOSES * 4 53¢ ' , 5198 TO REACTOR DRAIN . BLSSJEcg VALVES < X pr——T"—T\———= LINE [NO.103) PURGE i / | STACK ™ SECONDARY ' CONTAINMENT " ’ | RELIEF © NOTES: _ L VALVE 1, AN EXCESS RADIATION SIGNAL FROM 2. A LOW HELIUM SUPPLY PRESSU)?E PCV - J,500¢ FROM HELIUM STORAGE Fig. 2.17. Helium Supply Block Valving. RE 596, (ANY 2 OF 3) CLOSES HELIUM SUPPLY VALVES , HCV 593 B1, B2, B3, HCV 599 Bi, B2, B3 TO BUBBLERS IN FUEL PUMP BOWL. 4, THIS LINE, 588 GOES TO AUXILIARY SIGNAL LINE 500, (ANY 2 OF 3] PRECS. SWITCHES) CLOSES HELIUM BLOCK VALVES MARKED WITH ). ASTERISK, THUS % |} i , CHARCOAL BEDS, LINE 56%, VIA DRAIN TANK VENT VALVES. SEE FIGS. 2.14 AND 2.15 3. LINE AND ASSOCIATED PXM'S IS A ZERO psig REFERENCE LINE WHICH MAINTAINS CONTAINMENT AND TRANSMITTERS. g T ) A ‘ , ( ' UNCLASSIFIED L . - . ) ORNL-DWG 638397 - (560) © O IN-GELL COMPONENT T —mwesToy , COOLING AIR HEADER - / ' CC — . . . \ . =51 _ % oy REACTOR CELL , - AIR COOLER : EVACUATION LINE , NO.1 . I 565 Al y / ; ' A""v‘v"""““‘v - x cor | Cee—— v | : ‘ NO.2 p ' ’ > T PRESSURE BLOWERS e ‘ HELIUM FROM | ! : FUEL DRAIN TANKS "= D -— i GHARGOAL BEps M DG5 PUMP, : L NOS. 1 AND2 J'F , A | s @ : 13) HELIUM FROM COOLANT SALT ~— G- N> .GIRC. PUMP. ‘ ABSOLUTE: FILTER @7 -~ - : _ Sem—e—|——n HELIUM FROM | HEY J COOLANT DRAIN TANK | 557Ct AND LUBE OIL SYSTEMS - 55D NS SAMPLER- ENRICHER I @D - v 0 &0 3 T - ) \ ‘ :\ | | , AE 0 2 : 1 ) 3 £3 :: 1 J _ VENTILATION AIR ] / _ . S : FROM: / ' B | < : . : 1. LIQUID WASTE CELL "-:!a ;\J_h v N ,93'\\-/ I o THIS LINE USED ° . ' 2. DECONTAMINATION CELL I TT ¢! =~ . i ONLY DURING MAINTENANCE / 3. REMOTE MAINTENANCE [ — ' PUMP CELL 4, HOT STORAGE CELL | 5. FUEL STORAGE TANK : : VENTILATION CELL J AIR' FROM REFESER F!Ss.o%qs o : : SE VICE TUNNEL - FOR ILS OF LU 6. SPARE CELL OIL SYSTEMS IN-CELL COMPONENTS o rrrrre @39 Y / a | oo c o | H ANK 1 CELL 3 4B 1 > >~ v 000 coxoarrr REACTOR [ELL / VAPOR ; _ PROCESS WATER —— CONDENSING ] ; TANK NO.2 ° VAPQ - FILTER 250 CONDEN ' LTE ctm .o TAN : vT FROM RUPTURE DISK ON THERMAL SHIELD COOLING WATER LINE, SEE FIG. 2.22 F\ig. 2.18. Off-Gas System Instrumentation and Valving. LOG AND +——m ALARM ~ COVER GAS PRESS CONTROL SYSTEM 85 UNCLASSIFIED ORNL-DWG 63-83293 SIGNAL TO CLOSE FROM RADIATION MONITORS RM-557 A48 | —— N3 - | QUIGK ™~ | oisconnecT 99 - THIS VALVE BLOCKED \\\T——————-— | A WHEN HELIUM SUPPLY POy | ooy | HELIUM | SUPPLY (513) <} I 40 psig I ALARM .i.. LOG OIL SUPPLY | D 5 TANK FOR MON FUEL SALT RE Pugf1 G ~——708)- QT8 = —(715)——<4——=—— EXCESS FLOW RETURN FROM PUMP OUTLET ALARM TC CIL G1ar D PUMPS 713 e FOP-1 FOP-2 NOTE : BOTH LUBE OiL SYSTEMS ARE SIMILAR. NUMBERS ENCLOSED THUS (O ARE EQUIVALENTS IN v COOLANT SALT PUMP COOLING QUICK LUBE OIL SYSTEM WATER DISCONNECT Fig. 2.19. Lube 0il System Off-Gas Monitors. 1. Line 557, which carries off-gas from (a) all charcoal beds, lines 562, and 557, (b) helium from the coolant salt pump, lines 526, and 528, (c¢) helium from fuel and coolant salt pump lube-oil systems, lines 534, and 535, and (d) helium from coolant salt drain tank and line 560. cates 1 (b) abovel. Line 528, which carries helium from the coolant salt pump bowl [dupli- 3. Line 565, the reactor cell evacuation line. 5 4. Line 927, the main inlet line to the stack fans. As shown on Fig. 2.9 and Tables 2.2 and 2.3, these different radiation monitors do not have a common output. A high radiation signal from RM- 557A or RM-257B in line 557 blocks helium flow from the lube oil systems, Fig. 2.19, by closing valves PCV-210A2 and PCV-513A2. It also closes HCV-557C1l and blocks line 557, which is, in effect, a header carrying the flows listed above. Blocking of the lube-oil-system off-gas is, there- fore, redundant. A high radiation signal from the radiation monitors in line 565 closes HCV-565A1 and blocks this line only and thus prevents cell evacuation. A high radiation signal from the radiation monitors in line 528 carrying helium from the coolant pump initiates a drain. This particu- lar monitor is deemed necessary becauge it provides an indication that there exists a leak in the salt-to-salt heat exchanger (from the fuel salt loop to the coolant salt loop). Such a lezak could put fission producte in the coolant salt; however, when the coolant salt circulating pump 1s running, the coolant salt pressure in the heat exchanger tubes is greater than the fuel salt pressure in the shell., The fission-product gases would be carried from the free surface in the ccolant salt pump bowl into line 528. Thig activity would be read by RM-557A or B or both shortly after being flret read by RM-528A or B cor both. Blocking as described above would then fellow snd the protecticon afforded against this heat exchanger leak is thus redundant. The systom ig tested by inserting a radiation source in the radiation monitor shields and observing (1) the radiation-monitor indicator, (2) the control circuit relays, and (3) the action of the control valves which provide blocking. The three pairs of radiation monitors, RM-5Z&A and B, RM-55%A and B, and RM-565A and B, all operate such that an excess radia- tion signal from either channel in a pair will produce safety action. Insertion of a radiation source in any one monitor will cause all block valves asscclated with that particular monitor to close. Testing of RM-565 or RM-52¢ will initiate a reactor drain, and unless the test — is completed within 5 min, the time required to thaw freeze valve 103, 87 an actual drain will be established. The other channels may be tested for longer periods of time without affecting operations. The stack gas monitor (item 4 above) on line 927 monitors off-gas activity from all the other sources Just before the filtered off-gas enters the stack blowers. Ixcess activity is alarmed in the control room and another alarm signal is transmitted to the ORNL Central Monitoring Facility. In-Cell Liquid Waste and Instrument Air Block Valves. As indicated by Fig. 2.9, the in-cell (secondary contaimment) liguid waste lines are blocked if the reactor cell pressure exceeds 2 psig (17 psia). The normal operating pressure is 13 psia. Excess cell pressure is a symptom of sys- tem malfunction or, at worst, the maximum credible accident (see sec. 8.5). The system and its operation are shown on Figs. 2,20 and 2.21. Three independent pressure-measuring channels with trips (pressure switches) are employed as inputs. The pressure-switch outputs go to a two-out-of- three coincidence matrix. If any two input channels indicate a positive cell pressure equal to or more than 2 psig, the block valves in the lines from the reactor and drain tank cell sumps to the liquid waste tank are closed., The instrument air lines into the containment vessel are blocked at the same time. Since each pressure-switch contact is connected to a different power source (see Fig. 2.21), any single power failure will not produce blocking, and no single channel or‘component failure will disable the systemn. The pressure switches are mounted close to the cell wall and are accessible during reactor shutdown. The pressure gages and hand valves are mounted on a test panel located in an area which is accessible during power operation. Fach pressure switch is connected to a separate tap on the reactor containment vessel. Iengths of lines between the snubber, hand valve, gages, and the switch are as short as is consistent with the location requirements. Wiring in the three channels is run in conduits separate from each other and from control-grade wiring. The switches are connected toc a two-out-of-three solencid matrix in the manner shown in Fig. 2.21. The circuit shown is a simplified schematic. A relay will be required to operate other block valves, SERVICE PRESSURE INDICATOR SINTERED UNCLASSIFIED ORNL~DWG €3-8396 NITROGEN ———— SUPPLY 50 psig MAX HAND VALVES R DISK SNUBBER ' SECONDARY PRESSURE SWITCHES / CONTAINMENT, (- @ (ACTUATE AT +2 psig) s 0@ " o3 SEE FIG, 2.21 \; REACTOR [ \\ / (669) | CELL @ \ @_ ) 1 NOMINAL | F \ 2 OUT OF """ T OPERATING B ~{ge-g/— — | 3 COINCIDENCE [~ — ' ceveL ! 1 ; - i PRESS.= | o 7’| RELAY MATRIX » J [ 1 -2 psig (3 psia) / | o TRANSM. (| 1 DRAIN | / A1 et / ' I TANK CELL A Qx 7| Pume / . —bfl-—m {4 : RCF } : s — Gt ‘ JET TO CLOSE 100 psig | & PUMP o \SUMP | AIR = o . : ‘ Teve, | : = c osss‘ss_ocx ‘ 1 . PROBE (BUBBLER) m“ Sump oy LT T T VALVES IN INST, ’ - : FCV ’ PCV 3431 i 24322 ~ AIR LINES 33128 »< bt <] e:l:’ - 332 = LIouiD. ! WASTE - TANK 11,000 gal / , Pig. 2.20. In-Cell Liquid Waste and Instrument Air Block Valving. 88 89 UNCLASSIFIED ORNL-DWG 63~-8392 25 kva AG EMERGENGCY M-G SET DC POWER + PRESSURE LINE FROM REACTOR _—.| SWITCH —_— CELL / (TYF’ICAL) / \ ¢ m PSS PSS ; m PSS . RC-B RC G \ . RG-F w T/ T RC- — - RC" ESV ESV ESV ESV ESV ESV 1A1 142 181 182 1G4 1C2 -AC DC NEUTRAL NEGATIVE 80-psig EMERGENCY ey X AlIR HEADER - ENERGIZED ' L Pl : DE-ENERGIZED 25 kva 1A4 TVA 5 o ESV cov {0 e A VALUE FLOW, i TYPICAL OF ALL B ESV'S 1A2 VENT v — TVA +DC 7. ESV Esv (1] e F—y - Pl 183 COINCIDENGE MATRIX OF +DC Pl 25kva --—— VALVES OPERATED fAS .y BY PRESSURE Ecsv [tq; }@ ‘L'bJ‘ sAo SWITCHES 1C1 BLOCK VALVE i N HEADERS TO REACTOR AND DRAIN TANK i CELLS i — VALVES INST AIR LINES Fig. 2.21. Pressure Switch Matrix Used with Instrument Air Line Block Valves. 90 The pressure channels will be tested by slowly opening the hand valves in the nitrogen supply lines, one at a time, and observing pressure changes in the pneumatic circuits that operate the block valves (for example, see Fig. 2.20). If the nitrogen valve in the left hand line is opened, pres- sure switch PSS-RC-B will be actuated and the actuating pressure may be checked with the gage, PI-RC-B. As shown in Fig. 2.21, actuating this pressure switch deenergizes two solenoid valves, ESV-1A1 and ESV-1A2, in the six-valve matrix. The line carrying PI-1A5 will be vented tc the atmo- sphere, and PI-1A5 will read zero. The other two channels are tested in a similar way by observing pres- sure gages PI-1A1 and PI-1A4, respectively. The tests do not actuate the block valves, since the coincidence matrix requires agreement of any two of the three input channels to vent the block valve headers to atmospheric pressure. In-Cell Cooling-Water Block Valves. A simplified flow diagram of the in-cell cooling-water system, with instrumented block valves, is shown in Fig. 2.22. The signal tc close the block valves 1s provided by an excess radiaticn level in the pump return header, line 827. Three independent sensocrs, 3=-827A, B, and C, are used that operate in two-out- of-three coincidence; i.e., block valve operation 1s initiated when any two of the input channels indicate excess radiation in line 827. This system permits the loss of one sensing channel by malfunction or for maintenance. In these circumstances, a signal from elther of the re- maining channels will actuate the block valves. Operation of the system is apparent from Fig. 2.22. Testing is accomplished during operation by manually inserting a radiation source in the monitor shield and observing that the indicated radiation level increases and that the proper relays operate. The construction ¢f the sensor shield is such that each radia- tion monitoring channel can be tested individually. Since the monitor contacts are arranged in two-out-of-three ccincidence, testing of indi- vidual channels does not close the valves. The valves will be tested individually by operating a hand switch and observing that flow stops in the line under test. The complete system can be tested during shutdown. NOTES: . ON EXCESS RADIATION SIGNAL FROM ANY UNCLASSIFIED ORNL- DWG 64638 2 OF 3 RADIATION MONITORS ALL VALVES BIES ;E;QL_%E__ . AN SOPPY MARKED WITH ASTERISK, THUS (D% ARE CLOSED M 43 psig . 0 GAS | T 828 COOLER 42 psig IN SPECIAL | I \| © EQUIPMENT 2. RUPTURE DISK AND PRESSURE SWITCH PROTECT | ROOM GAS PUMP THERMAL SHIELD FROM RUPTURE IN THE EVENT | T'TJEA‘?FTE?D \\ | oL COOLERS OF EXCESS WATER PRESSURE | cooLer /| A b | - | - THROTTLING "/ ;W' R —_—— e T/ vawes 51 psig J \ FROM COOLING / 3 TOWER / PUMP Y ‘= o NO. @ / z g ®2D o wn —@23 o : o % o \ 8 £ \3 F’@’;’ (835 58 psig PRESSURE gy : SWITCHES o Z SEE NOTENO.2 T - 25 3 o RUPTURE _Lay © COOLANT CHEMICAL 5K L SALT PUMP ADDITION FROM SPECIAL FUEL PUMP . i ; ; MOTOR o EQUIPMENT M \ i L £ T ROOM COOLANT SALT SPRING SO LOADED ‘ - HAND MOTOR l—m VALVE —» S NUCLEAR ] ks, @ @ INSTRUMENT Ty PENETRATION % N C . C [ L= DRAIN — | ‘ r' TANK 3 INDEPENDENT RADIATION L L Lo CELL MONITORS RE 827 A, B, E,\R’E‘I\RE) y ro- "AND G <] T AR | - 2 OUT OF 3 St X COOLERS < =7 r COINCIDENCE {————=——~~ v X R | MATRIX | 3 RELIEF . ; VALVES ] L~ 10 BLOGK vaLvEs ~ 100Psa ] 1 SEE NOTE NO. 1 } THERMAL / v SHIELD ~23 psig {84&r ; P :], SECONDARY - | (84— v CONTAINMENT : @D ’, § 0 . ! WASTE TANK he e e . A Fig. 2.22. In-Cell Cooling Water System Block Valving. T6 92 Valve ESV-ST-A serves as a backup to FSV-837, FSV-846, FSV-841, and FSV-847 and provides redundant blocking in the system. The thermal shield is protected from excess hydraulic pressure by interlocked blocking of the inlet, line 844. The pressure switch on this line closes valve FSV-844A1. The rupture disk protects against overpressure if a temperature rise takes place when the thermal shield is isolated by the block valves or other flow stoppage. Discharge through the rupture disk passes into the vapor- condensing system and is contained. 2.2.8 Health-Physics Radiation Monitoring The purpose of the radiation monitoring system is to provide personnel protection throughout Building 7503 from radiation hazards due to airborne and fixed-gource activity. The monitors composing this system are placed at strategically located points throughout the building (sce Fig. 2.23). A1l the monitors that measure area activity and airborne activity produce signals which are transmitted to a central annunciator and control panel locgted in the reactor auxiliary control room. These monitors also pro- duce a building-evacuation signal based upon a coincidence of combined monitor operation. The criteria fcr the number, location, and arrangement of the Health Phyeics menitors were established by the ORNL Radiation Safety and Con- trol Office assisted by the reactor operations group. The locations of the monitors in Building 7503 are shown in Figs, 2.23 and 2.24. The contiruous air monitors (seven required) are located within the building so that the air-flow patterns will sweep air past these units tc menitor for airborne zctivity. Twe monitcrs are in the high-bay area, one in the ccntrol rcom office area, and three In the basement working area. A seventh instrument is a mobile unit for special operations. Eight fixed mcnitrcns for lcow-level gamma detecticn are located in the same general arrangement as the constant air monitors. Very sensitive beta-gamma mornitcrs (GM-tube count-rate meters) ére used in contamination free areas for clothing and body surveys. Two of these devices are used for process monitoring, one in the cooling water room and one in the vent house. , ‘ . . i \ ‘ N m‘ 4 ' | N \ 93 . i . : . ‘ UNCLASSIFIED . . . f . ORNL-DWG 64-6209 T ! . - o . T . . BLDG. 7509 | ‘ ; ‘ N — pY i fis} fi) \ | . ¥ T I \8 T ] [ | : ) _ : & ‘ : ‘ DOWN et ;—1. L || =N L o P P A 5 | L \2 u - — GUARD PORTAL g MAIN CHANGE ! | . . FIRE HOSE o PANEL"B" - ROOM o o—— 'A\__&—&& /7' °\ - . HOT CHANGE 4 4 ‘ : : FIRE THROWOVER SWITCH FIRE ‘ __Roow : : ; , EXTINGUISHER bl EXTINGUISHER ik : : o _ : - “FIRE T N _ = e | FILTER P e . : EXTINGUISHER : ‘ e T : - . g o LIGHTING A : - eI : '. - PANEL"G" = ' . i - o ‘ HOT CHANGE T~LIGHTING PANEL"J" ' . O— . — : oow 3 [T T | o , C ' - 10-ton CRANE v : *: | ' FTT T[T TTT FIREHOSE= i ~ 0o ‘ N lFUIELI | o . HOT STORAGE | 1ennSFER ! . : 0 _ : Lo | ! R R R C o ‘ \ | I CELL, J _ _ _ o S LL L e ojojooofg]ofo]olo a0 o WFTTTW ol J ' o 1 v b 4 | _~RADIATOR PIT - : NATION. Srogy | | DRAIN TANK CELL & PENTHOUSE TOP : Lt |y srorace | - : ELEVATION. < . {11 1eew | ' 86111 Oin. ADSORBER - ALl | 30=ton 0R , = CRANE «__ i : - MAINTENANCE ot ELEZ CONTROL ROOM Ly FLOOR ELEVATION . . ! : 8621t Oin. -”- WEST SPACE COOQLER ' ~ EAST SPACE COOLER ' A ) _ , : SWITCH HOUSE : 1 ‘ | ' o ’ ‘ DOWN—T ' F . ‘ : L o : EAN HOUSE i M AR MONITER |Q-2240 . * _ _ | - - i mmme MONITRON Q-H54 ] . @ CRM (BETA AND GAMMA) Q-2091 _ W EVACUATION HORN w | 1 ! @ ALARM BEACAN LIGHT ‘ o i 48 HAND AND FOPT MONITOR Q-t939 '. o , T . o . : i A , ' - : | D : i DIESEL HOUSE | Y : ; ! ) PLAN-ELEVATION 852t Oin. - ‘ | ' PUMP NO. 2 : . gwmp NO. ! ' \ : e—COOLING TOWER g C | , L Fig. 2.23. Reactor Building (7503) at 852-ft Elevation Showing Locations of Monitors. - F ’ b o UNCLASSIFIED . S " ORNL-DWG 64—-6210 N -— - ® @ ® ® ® uP - . O—F ] : ' » - BATTERY DOWN DOWN ROOM FLooz’gNEL 533 # ’}E lfih ’ IN BREAKER e &ng:grioc NO.2 . FIRE EXTINGUISHER =S/'a7 l=0IST. NO. 2 - : ®&——uh o o D40-ton CRANE o b DIST. NO. 1 @V e . MG NO.2 AND NO.3 R MOTOR AND GENERATOR ————— ——— MAIN BREAKER NO.{ \ 48 volt DC BATTERY CHARGING SYSTEM DIST. PANEL 1A ©—) & [a SPECIAL MAIN TRANSFER EQUIPMENT O SWITCH TO it x EMERGENCY ROO 1 | GENERATOR @— : 1 - " | | o I e e M Il ’:‘ S 1Y 7 Loy stacken BLOCK COOLING WATER o . L EQUIPMENT ROOM—T[ ‘ » Bl AR MONITOR. G-2240 - - amme MONITRON Q-54 , — . . , @) CRM (BETA AND GAMMA)} Q-20Q1 . FAN HOUSE _ W EVACUATION HORN . ‘ RAMP DOWN M ALARM BEACON LIGHT ' 'PLAN-ELEVATION 840 ft Oin. Fig. 2.24. Reactor Building (7503) at 840-ft Elevation Shdwing Locations of Monitors. 95 A selected group of constant air monitors and monitrons form the two input channels, contamination and radiaticn, respectively, which actuate the building-evacuation system. The output signals from the monitors in each channel are transmitted to a central annunciator panel located in the auxiliary control roomnm. Excessive radiation signals from two monitors in either the radia- tion or contamination channel will cause the evacuation siren to sound. At the same time, four beacon lights at the corners of the building will begin flashing and a signal will be sent to Guard Headquarters. Also, a manual evacuation switch is available on the reactor console. On the constant air mcnitors and the monitrons, there is an inter- mediate alarm. When a monitor reaches this set point, the indicator lamp on that monitor input module is lighted and a lccal buzzer sounded. This buzzer is located in the HP annunciator panel. At the same time this buzzer sounds, an annunciator in the control rocm indicates that a health physics instrument 1s in the alarmed state. This alarm signal is also produced by an instrument mslfunction, since both the constant air moni- tors and monitrons produce a signal upon instrument failure. An administrative prccedure involving both reactor operations and health physics is used to ensure adequate protection during reactor main- tenance periods without cauvsing false evacuation signals. This is imple- mented by two key-locked switches, one in the main control room and the second in the maintenance control room (see Fig. 2.39 in sec. 2.6). Oper- ating either of these switches prevents the monitors from actuating the alarm sirens and beacon lights and switches any evacuation signal to Guard Headguarters. The established procedure requires the Chief of Operations to call Guard Headquarters and to evacuate the area of all personnel except those essential to the particular maintenance operation involved. The Chief of Operations then throws the key-locked switches and disables the alarm sirens. The manual evacuation switches can be used to override the alarm cutout should evacuation be required. The individuval monitor alarms and the main annunciator still continue to function. 96 2.3 Controcl Instrumentation 2.3.1 Nuclear Instrumentation The normal operating ranges of the nuclear instrumentation are listed above in Table 2.4, and the safety channels were described in Section 2.2. The nuclear instrumentation used for reactor control consists of two wide- range counting channels and two linear power channels, as described below, In addition, provision is made for two temporary BF; counter channels (Fig. 2.25), which will be used for the initial critical experiments and other low-level nuclear tests. These counters will be located in vertical holes near the inner periphery cof the annular thermal shield, as shown in Fig. 2.26. Alsc, extra tubes are available in the neutron instrument pene- tration, as described below. All permanent neutron sensors and a portion ¢f the associlated electri- cal and electronic ccmponents shown on Figs. 2.10, 2.27, and 2.28 are lo- cated in the nuclear instrument penetration, a 36-in.-diam, water-filled, steel tube. The upver end of the tube terminates on the main floor (852-ft level), end the lower end penetrates the thermal shield at the horizontal midplane cf the reactor core. The penetration is inclined at an angle of 42° from the horizontal; its design and location are depicted in Figs. 2.26 and 2.29. Fach sensing chamber is located and supported within the penetraticn by an individusl aluminum tube. The chamber housings are lcose fits in the ¥ UNCLASSIFIED ORNL -DWG &4-624 —— PREAMPLIFIER LINEAR PULSE SCALER E—— ORNL MODEL AMPLIFIER ORNL MODEL Q-1326 ORNL MODEL Q-1326 Q-1743 BF3 COUNTER HIGH VOLTAGE SUPPLY ORNL MODEL Q-2057 Pig. 2.25. Typical Low-Level BF3 Counting Chamnel for Initial Criti- cal Tests. UNCLASSIFIED ORNL—DOWG 64-984 / INSTRUMENT PENETRATION SEE FIG. 2.29 l - BFy HAMBER TUBE 97 UPPER END OF NEUTRON INSTRUMENT PENETRATION REACTCR CELL~. -2 -NEUTRON BF3 /!.//, “~ ’// Y /, A . 1 ., ™, § ! A i o - . . ; N, \.~ hY i ;. A . ; x 5 o ,_ s 5 Wi L Lo o - W.L = oot Ll Ll Wl — M m ! , T L SP : | =@ sl ! . L) W —J D @ L 3 N I, = n N ~ w > Y o @ " o S , 2 < @ ] *, 1 N, < < \ \ ,,. } *, < k! S N \.\ \ P /. \\\ _,Xx.\‘ — - = - o S e © Y e N N > - Locations of BF3 Chambers. Fig. 2.26. . r——o—zsw-q ———-I ' (— FISSION I CHAMBER -PREAMP ‘ PREAMP. PWR. SUPPLY \ PULSE AMP PULSE DISC, SCALER PULSE AMF. AND CRM Q-2614 LINEAR |- CRM LOG CRM — FUNCTION GENERATOR I Q-2616 [ i I I I L INTERPOLATING VERNISTAT POT. / | | | | | 1 | | | | i I I | | l | e e e e e e o e i Q-2605 o 28 \§ 3 REACTOR PWR 1 Q-2609 Q-2609 >2cps 1 <10%cps FAST TRIP FAST TRIP COMPARATOR COMPARATOR e e COUNTING CHANN CONFIDENCE Q-2615 EL SERVO PWR AMPLIFIER TACH, MOTOR NOTE: Q-2605 IS ANALOGUE COMPUTER TYPE OPERATIONAL AMPLIFIER DIFFERENTIATING AMPLIFIER — —1— Q-2609 @-2609 Q-2609 < 200 kw »>500kw >1.5Mw FAST TRIP FASTTRIP FAST TRIP' COMPARATOR COMPARATOR COMPARATOR SAG RUN PERMIT ROD REVERSE IN START MODE LINEAR FISSION POT. CHAMBER POSITION REACTOR PERIOD Fig. 2.27. Wide-Range Counting Channel.. ! UNCLASSIFIED ORN(-DWG 64-633 Q-2609 Q-2609 Q-2609 T30s8ec FAST TRIP FAST TRIP FAST TRIP COMPARATOR COMPARATOR COMPARATOR ROD ROD INHIBIT RUN PERMIT REVERSE ‘ AND ALARM O h 86 99 UNCLASSIFIED ORNL-0OWS 64-635 — DATA LOGGE R - N | DATA l LLOGGFRI P -39k puh - £ e \ CONSOLF *L 7 COMPENSATED TON CHAMBER Q-1045 TO INLET DATA TEMPERATURE NOTE: Q-2605 IS OPERATIONA. |LOGGE) SENSOR AMPLIFIER & TWO PEN . N o ECORDFR FOXBORO i RANGE ] ]CHANGER|_;<_( RANGE LEVEL . E1 SOLE CHANNEL -~ i 1 1 | co s ! i '[ Q-2605 DATA : SELECTOR SWIiTCH 553_(1 RESET LOGGER AV DS i < - | F I“Z AMPLIFIER \\* ., o ¢ \ \_\_\ L,m\‘"; Y / S SN ,‘RQTJGE SEAL 1 f / - . | ~— ~.. [ T p I p Q-2605 Q-PE05>—— 7 o ! ke e ~ COMPENSATED o N T NG _RANGE “J‘ T 7 [ 10N CHAMBER - o) CONSOLE 1-Mw | SEAL ” G-1C45 I FLUX = = CLAMP ... FOXBORO e F/I TEMPERATURE / SERVO DEMAND| ] 1 1" DRIVE UNIT . , 333'9 0 5528 () consore Vv . TOOUTLET oM PUTVED TEMPERATURE /FLUX DEMAND SENSCR | - FCC FAST TRIP f—> COMPARATOR RUN PERMIT SERVO Q-2609 < iMw | on Y — R"'N # o SERVO FLUX DEMAND | — S| ROD MOT: —1 LIMITING CRT. * Q-2605 PiLOT ’ 0D MOTOR AL CONTACTOR : ! RELAYS - - M < b << 11 Mw NG RUN - | ACHOMETER : I ) L =l T NO RANGE SEAL ! ] L TR EMA\JD ~——\ ] . ‘ MANJAL - ; " ‘ CONTROL SYNCHRONOUS MOTUR RANGE SEAL i DEMODULATOR 1-Mw : o FLUX YT T T T T T T T e e " DEMAND oo Fig. 2.28. Linear Power Channels and Automatic Rod Controller. tubes so that chamber positioning can be accomplished by mcving the chamber axially within the tube. A float-type water-level indicator is used that will alarm on low water level. Water supply to the tube is controlled manually by a hand valve. An overflow pipe near the top of the penetration limlts the upper water level. The normal water level provides ample biological radiation shielding. In addition to the level alarm, monitron No. 2 (see Fig. 2.23) will alarm on an excessive radiaticn level in the region near the top end of the 100 241t 2in, NEUTRON INSTRUMENT JUNCTION BOX UNCLASSIFIED . ORNL-DWG 64-625 _i ' ¢ FLANGE FACE U | . HIGH BAY AREA /WATER-SAND ANNULUS 1 REMOVABLE CONCRETE SHIELOING - FINISHED FLOOR EL. 852 ft Oin \ SR REACTOR THERMAL SHIELD INSULATION 2.29. pehetration. ON CHAMBER {TYPICAL) LT ) CHAMBER GUIDE TUBE (TYPICAL) —t END OF 48-in. OD SLEEVE EL. 8411t |-§ in. 48-in.0D OUTER SLEEVE ASSEMBLY (WITH EXPANSION JOINT) —EXP JOINT t CONTAINMENT PENETRATION EL. 8341 52 In. - ¢ EL.830 #t 3in. SHELL 36 in. 3stinin/0 QUTER STEEL SHELL REACTOR CONTAINMENT FACE - 41* 50 FACE OF INST. SHELL FLANGE EL. 851t Bin. % WATER LEVEL IN: STRUMENT Nuclear Instfumentation_Penetration. geneous Reactor Experlment (HRE-2). Wide-Range Counting Channels. A 51m11ar installation was used successfully in the Homo- v The entire operating range of the re- actor is monitored by two counting channels, using automatic positioning of the fission chambers to extend the limited range of the more usual ar- :'rangement.4 The block diagram of one such channel is shown in Fig. 2.27. The technique mekes use of the variation of neutron flux with detector 101 position, which is nearly exponential in most shielding configurations, The function generator is used to make the detector-position signal more nearly proportional to the logarithm of the neutron attenuation. Adding this signal to a signal proportional to the logarithm of the counting rate gives a resultant signal proportional to the logarithm of the reactor power. Computing a suitable derivative then yields the reactor period. The chamber is moved by a drive mechanism under the control of a small servo system, whose function is to attempt to maintain a constant counting rate of 10% counts/sec. When the reactor is shut down, the counting rate is much less than 10% counts/sec, and the servo drives the detector to its innermost posi- tion. As the reactor is started up, the counting rate increases and causes the indicated reactor power to increase correspondingly. As the startup proceeds to higher power, the counting rate reaches 104 counts/sec, and the servo withdraws the detector, keeping the counting rate constant; the change in detector position then increases the indicated reactor powver. The servo need only be fast enough to follow normal maneuvers; any transients which are toc fast for the servo will change the counting rate, and the channel will read correctly in spite of the lagging servc. The advantages of this technique are the absence of any necessity for range gwitching or any other action by the operator, the wide range covered, and the elimination of the 'dead time' induced when the chanber is with- drawn quickly, as in other methods. Further, operation is always at an optimum counting rate when sufficient flux is available. Tinear Power Channels and Automatic Rod Controller. There are two linear power channels in the MSRE. Each channel consists of a compengated ionization chamber, power supply, and linear picoammeter, with remote range switching. Either of the two channels can be used to provide the flux input signal to the automatic rod controller, and both can be used for monitoring reactor flux level over a range of about six decades. A block diagram of the two flux channels and the automatic rod controller is shown in Fig. 2.28. “R. E. Wintenberg and J. L. Anderson, "A Ten Decade Reactor Instru- mentation Channel, Trans., Am. Nucl. Soc., 3(2): 454 (1960). 102 The automatic rod controller has two modes of operation. When the reactor is in the "start" mode (see discussion of modes in the next sec- tion), the automatic rod controller is a conventional on-off flux servo. The set point, in terms of neutron flux, depends upon the range of the picoammeter selected by the operator. Continuous adjustment is available. The maximum set point attainable in this mode corresponds to a power of approximately 1.5 Mw. At such low power levels the effect of power on temperature 1s slight; however, the operator must adjust the thermal load to keep the system temperatures within acceptable operating limits. When the reactor is in the "run" mode, the automatic rod controller becomes & reactor temperature controller, with the temperature set point adjustable by the operator, When the servo controller is operating as a temperature controller itg principal function is to augment the natural temperature coefficient of the reactor and to provide the reactivity changes required to compen- sate the small [-107% (Sk/k)/Mw] power coefficient of reactivity. The average reactor power will, over a period of time, tend to follow the load demands with only small changes in average core outlet temperature, The thermal capacities of the system are large and the power density is low; this, plus the long transport lags in the system, result in long time constants and delayed coupling between the power demand and the power being generated in the core. Analog simulation hags shown that without supple- mentary control, the reactor's response to changes in load are slow and characterized by flux and temperature overshoots and oscillations. The rod control servo, operating to hold core outlet temperature constant, makes up for these deficiencies in inherent control at a rate which causes a minimum of system perturbation. The servo controller,5:6 see Fig. 2.30, compares measured reactor flux, ¢, with the measured temperature rise, AT, in the core and computes 53. J. Ditto, "Preliminary Study of an Automatic Rod Controller Pro- posed for the MSRE," internal ORNL document MSR-62-96. ®E. N. Fray, S. J. Dittc, and D. N. Fry, Design of MSRE Automatic Rod Contreller, "Instrumentation and Controls Div. Ann. Progr. Rept. Sept. 1, 1963," USAEC Report ORNL-3573, Oak Ridge National Laboratory. 103 UNCLASSIFIED ORNL-DWG 64-6206 SUMMING AMPLIFIER (TYFICAL), OUTPUT IS NEGATIVE SUM OF INPUTS INTEGRATOR 1 %), € > +8, INSERT ROD €< =8, WITHDRAW ROD z € ], 0 INSERT € 0O WITHDRAW - INSERT € O WITHDRAW P 0 INSERT € 0 WITHDRAW T = FUEL SALT TEMPERATURE, CORE INLET T, =FUEL SA_T TEMPERATURE, CORE QUTLET NEUTRON FLUX [E; 0 SET POINT, FUEL SALT OUTLET TEMPERATURE | —= T{ME 0 et . Fig. 2.30. Computer Diagram for Servo-Controller Simulation. 104 a flux demand signal which, in turn, is used to produce the appropriate regulating rod motion so that reactor flux is adjusted to meet the load demand at rates equal to the rates of demand changes. The controller includes a slow reset element that corrects (slowly reduces) the error signal on a long time scale. The rate of reset is such that it has little effect on the transient error signal brought about by load changes. This reset element acts so that slow drifts and calibration errors of the tem- perature and flux sensors dc not affect the servo's ability to make the reactor follow changes in load. If, however, the outlet temperature thermocouple drifts, the reactor outlet temperature will follow the drift and system temperatures will be raised or lowered accordingly. Such a change will be apparent to the operator since system temperatures are monitored frequently and in many places. Figure 2.31 shows the result of an analog simulation of the response ol the reactor to step changes in load from 10 to O Mw and back to 10 Mw. The temperature of the salt leaving the reactor shows nc appreciable de-~ viation from set pcint, except after t = 250 sec, when the controller is forced by its limiting action toc hcld the power at ~0.5 Mw with no heat loss. The behavior of varicus system temperatures as the temperature set point is linearly decreased and increased at 5°/min, while the load is held ccnstant at 10 Mw, is shown in Fig. 2.32. 2.2.2 Plant Centrel Because of the conflicting control reguirements under different oper- ating conditions, the control system has been designed to operate in sev- eral modes selected by the cperator. These are shown in the tlock diagram in Fig. 2.33. The "prefill" mode is that mode of operation during which the reactor is empty and certain manipulations of the helium and salt sys- tems are required. This mode provides interlocks to prevent filling of the reactor while allowing transfer of fuel or Tlush salt between storage tanks., Circulation of helium for cleanup and for heating or cocling can also be maintained. The "operate” mode is used for f£illing the reactor and for all operations associated with a full reactor. In the "operate” mode the transfer of fuel or flush salt between tanks is prohibited. TEMPERATURE (°F) PCWER (Mw) 1250 1225 1200 1475 1150 1125 1400 1075 1050 1025 7 Fig. 2.31. UNCLASSIFLED ORNL—DWG €64—-620 ~REACTOR QUTLET TEMPERATURE DEMAND INCREASED TO 410 Mw RADIATOR INLET SALT TEMPERATURE RADIATOR OUTLET SALT TEMPERATURE 1 -t - | . o | DEMAND-(RADIATOR HEAT REJECTION) RESPONSE—REACTOR POWER {¢) | ! | | e 100 150 200 250 300 350 400 450 500 550 600 650 TIME (sec) Results of Analog Simulation of oystem Response to Step Changes in Power Demend with Reactor on Automatic Temperature Servo Con- trol. SALT TEMPERATURE (°F) UNCLASSIFIED CRNL-DWG 64-621 T 1 ~REACTOR QUTLET / TEMPERATURE | — | ‘ . I ] I MAXIMUM DEVIATION 6°-] I — —— T —— s T 1050 - _~REACTOR INLET 1 | - TEMPERATURE 1450 [~ - - - ” ; 1 o4 CONTROLLER SET POINT (REACTOR OUTLET TEMPERATURE) ~~ _RADIATOR QUTLET t——1- TEMPERATURE | _-RADIATOR INLET | TEMPERATURE | 1400 - ) : 7fififi.w.J_ | T I | e 1000 950+ - 900 —-—_T_/ f | P i | | CONTROLLER SET POINT (REACTOR . OUTLET TEMPERATURE) RAMPED AT 5 °F/min WITH LOAD AT 10 Mw HOLD ' __ NEGATIVE RAMP ! l | e | - | $w+_%;fi_ | PoaTwF RAMP HOLD | | o ig. 2.32. 100 200 300 400 500 600 800 300 TIME (sec) 700 Results of Analog Simulaticon of Reactor Response to Ramped Changes in Outlet Tem- perature Set Point with Reactor Under Automatic Control. GOT 106 UNCLASSIFIED ORN_—DWG 64-632 NOT i | PREFILL tMOBE HOT 1N GPERATE rACOE AT _EAST ONE F!LL AND DRAIN VALVE THAWED ore MCDE NG r RECUEST ‘ HFJUEb flPE«JE N'LDE i - - NG REGUEST [ oPeRaTe | L TmMoDE , RUN MODE _ ' l (—REQUEST "!NO REQUEST{ . FULL - —t = __L l SEAL | T 7 1 CUMP VALVE FRDZEN REACTCR ALTORATIC PERWMISES wE CONTTION RO N LOAD G 7 HUN MCDE 0K FOR RUN / —_— - ’ / f - L—-‘ 1 MOVERTARY SWITOH AST UATION WiTH | / L 3 So NG RETURN 1 O RECTION 5 o / o NO SAG / —— / " £ . / ."I . A _ - / = - p RESLLTING CONSITION 0% ACTION [W START z START l . \_/ fi reouesT (7777 7T ecovest | R 1 ,,,,,, : RELOTOR CONTRGZ MOLE Fig. 2.33. BSimplified Master Plant Control Block Diagram. Inasmuch as the reactor may be empty, yet conditions for the prefill mode are not satisfied, or full and the conditions for the operate mode are not satisfied, the third mode is defined as the mode when neither of the other two exists and is called the "off" mode. When the off mode exists, none of the usual operations involving pressurization, circulation of helium or fuel, or manipulation of the control rods is permitted. For nuclear operation of the reactor it is necessary that the reactor be full, and the fuel should be circulating. Under these conditions the reactor system may be operated in either the "start" mode or the "run" 107 mode. (These are, in a sense, submodes of the operate mode.) The start mode is used for all power levels up to gbout 1 Mw, at which point thermal effects become significant. The run mode is used for power levels between 1 Mw and the design point of 10 Mw. Control of the nuclear operation of the MSRE includes control of heat generation and of heat removal. To attain steady state, it is clear that the two quantities must be made equal. For low-level operation the reactor neutron flux is consldered to be the most significant reactor parameter, although system temperatures must be held within safe limits. ITnasmuch as at power levels lesgs than about 1 Mw the maximum rate of change of system temperature with no heat loss is of the order of a few degrees per minute, the use of flux control at these power levels does not pose much of a problem in adjusting the cooling capacity of the radia- tor to control temperature. Therefore, when the reactor system is in the start mode, where the maximum power allowed is 1.5 Mw, the flux and ther- mal load are independently controlled. The use of an automatic flux con- troller aids in such independent control by effectively decoupling the flux from any temperature feedback brought about by load changes or by changes in electrical heat input from loop heaters. In the power range above 1 Mw, the problems of independent ccontrol of power and of temperature become more difficult. The control system for high-power operation of the MSRE, with its capability demonstrated by analog analysis, permits the operator tc adjust the heat removal rate by manipulation of the radiator compconents while an automatic rod con- troller is used to maintain the desired reactor salt outlet temperature. The automatic rod controller has a temperature setpoint that is adjustable by the operator at a maximum rate of a few degrees lahrenheit per minute. In addition, the control of the load has been programmed so that the opera- tor need only actuate a single switch to increase or decrease the load over the entire range at a rate compatible with the temperature control- ler's capability. In practice, the automatic rod controller changes its mode of operation from flux to temperature when the reactor operating mode is changed from start to run. 108 The capsability of the automatic rod controller to make the reactor closely follow its load at all power levels from 1 to 10 Mw, with inde- pendent control of reactor outlet temperature and load, will be a great asset to the reactor operator. However, manual modes of operation have been provided for greater flexibility in the experimental progran. Rod Control. A simplified block diagram of the rod control system is shown in Fig. 2.34. In general, manual withdrawal of the three shim rods is permitted if the reactor period is not shorter than 10 sec and if no automatic rod insertion ('"reverse") signal exists. Reverse and the conditions producing it are discussed in a subsequent paragraph. In addition to manual operation of the rods, there is an automatic rod control, or "servo, ' mode of operation. In the servo mode, one of the shim rod drive motors is controlled by signals from the automatic rod con- troller described in the preceding section. In such operatiocn, the opera- tor exercises control cver the position of the servo-controlled rod (called the shim regulating rod) by adjustment of the other rods. The regulating rod limit assembly (Fig. 2.35) is an electromechanical device used to limit the stroke of the servo-controlled rod. Figure 2.36 is a diagram of the associated circulitry. This diagram is a simplified version of the actual control circuits in that it omits administrative and supervisory contacts, upper and lower rod limit contacts, etec. Referring to Fig. 2.35, 1t can be seen that, by means of the mechani- cal differential, the net rotation of the limit switch cam is the resultant of two components: (l) the rotation produced by the balance motor and (2) the rotation of the shim locating motor. When the reactor is in the servo control mode and if the operator is not relocating the position of the regulating rod span, the balance motor, driving through the clutch (the brake is disengaged) and the differential, rotates the limit switch cam through an angle that is directly proportional to the linear motion of the servo-controlled rod. It is assumed for the sake of discussion that the reactor 1s being operated in steady state and that no fuel is being added. Burnup and poisoning will cause slow withdrawal of the servo controlled rod unlessg the reactor is manually shimmed by the other two rods. If no o 1-»-1 . /s . l o 10° i f L } ‘ ' . hi‘ b UNCLASSIFIED R N ORNL-DWG 64-634 } ’t\ . \ | | | | | . l | | | | ONE WIDE . FUEL OUTLET - 1 SERVQ 5 . -1 IN RUN FLUX>12 Mw FUEL SALT P ROD § SWITCH GROUP SWITCH ROD 2 5WITCH GROUP SWITGH ROD 3 SWITCH GROUP SWITCH RANGE COUNTING TEMP.>1275 ' " REV INSERT .. " A b " . CHANNEL OPERABLE MOOE (2 OF 3) (2 OF 3) LEVEL>T5% F'ffif?xf?'sM" T 20 sec Lo ] REG.TROD ' " " " " INSERT LIMIT REVERSE = IN START ~IN RUN IN START SERVO "OFF SERVO "ON I MODE MODE | _ MODE I NO REVERSE b 1 ROD 1 NOT AT 1 SERVO "ON* \ — REG. ROD ; WITHDRAW LIMIT § : l________l : ! ROD % NOT AT NOT AT ROD 3 NOT AT . b i SHIM INSERT INSERT INSERT IN START I LIMIT LINIT LIMIT IN OPERATE MODE MODE i L ' INSERT INSERT INSERT INSERT — REG. ROD L ‘ L foD LIMIT RAD ROD 4 : NC.1 NO 2 NO.3 I ] 7 ] SWITCHES ROD 1 SWITCH GROUP SWITCH whow ROD 2 SWITCH GROUP SWITCH ROD 3 SWITCH GROUP SWITCH '"WITHDRAW "WITHDRAW" SIGNAL “WITHORAW" “WITHDRAW" " WITHDRAW" “WITHORAW" ROD 1 NOT AT f g ' F REG. ROD / - WITHDRAW LIMIT SERVO "OFF" SERVO “ON* * | ; . - I Lé : AUTOMATIC PERMISSIVE CONDITION _ Foo SERVO “ON" ' 5 i POSITION OF A MANUALLY OPERATED SWITGH 1 ‘ S \ | ROD 1 NOT AT - ROD § NOT AT ROD 2 NOT AT ROD 3 NOT AT RESULTING CONDITION OR ACTION SHIM WITHDRAW - REG. ROD WITHDRAW WITHDRAW LT INSERT LIMIT LIMIT -LIMIT ! WITHDRAW REG. ROD WITHDRAW ROD NO.2 Fig. 2.34. Simpli] WITHDRAW ROD NO.3 e e s e T ———— arme NOTE: * THESE CONDITIONS REFER TO POSITION OQF THE SERVO CONTROLLED ROD WITHIN TS AUTOMATIC SPAN AS DERIVED FROM-THE REGULATING ROD LIMIT SWITCH ASSEMBLY. fied Rod Control Block Diagram. ROD DRIVE MOTOR, SERVO OR MANUALLY CONTROLLED JA . y LOWER LIMIT SW ITCH—-”"’g; [ " CONTROL ROD | SHIM LOCATING MOTOR, OPERATOR CONTROLLED GEARBOX NOTE: AS REDUCED TO PRACTICE THE CONTROL ROD MOTION IS REPRODUCED IN THE CONTROL ROOM AND LIMIT SWITCHES ARE ACTUATED B8Y THE REGULATING ROD umMIT SWITCH ASSEMBLY. SEE DIAGRAM. —————— REGULATING ROD SPAN Ay, REGULATING ROD LIMIT SWITCHES 3% AVAILABLE TO SERVO CONTROLLER WITH REG. ROD SWITCHES L ATy, 1 SYNCHRO POSITION UNCLASSIFIED ORNL-DOWG 64-622 REGULATING ROD LIMIT SWITCH ASSEMBLY : "IN INSTRUMENT CABINET SHIM LOCATING MOTOR - ROD POSITION OPERATOR CONTROLLED TRANSMITTER BY CONTROL ROD ACTUATOR SWITCH. USED DURING = SERVO OPERATION ONLY COARSE r- | : csms.ox ! I I | . | Hl— - g SYNCHRO _ ROD DRIVE MOTOR: | | CONTROL 'BALANCE e L . OPERATED BY ROD CONTROLLER | | | TRANSFORMER MOTOR ONLY WHEN IN SERVO OPERATION AND BY ROD ACTUATOR SWITCH | | | cLutch-BRAKE: CALIBRATION IN MANUAL OPERATION - o. IN SERVO OPERATION THE DIAL . CLUTCH 15 ENGAGED AND ~_. . ; V THE CAM 1S COUPLED TO SLIP CLUTCH R I UNT I b} tie eaLance woror ] - 1 i b. IN MANUAL OPERATION GEAR BOX FOR - THE OUTPUT SHAFT IS ) | | LOCKED AND THE SPAN CHANGE GEARS | | BALANCE MOTOR | DECOUPLED (?D' I REGULATING ROD | LIMIT SWITCH | | | cam anp umir switcres LOWER | | WITH MECHANICAL STOP , | | REGULATING ROD UIMIT REGULATING: ROD WITCH UPP | | SWITCH UPPER POSITION TRANSMITTER I CLOCK CAM ROTATION (SYNCHRO) I | equiv. TO: | | 0. WITHDRAW ROD NO.{ | I b, INSERT LIMIT SWITCH / b SPAN { DECREASE y,) | - ' | | ] @ @ _ SEE FIG.2.36 FOR COARSE - FINE CIRCUIT DIAGRAM . REG ROD POSITION SHIM ROD INDICATOR {RELATIVE POSITION TO SPAN LIMIT INDICATORS - \ SWITCHES) . 34 . . —‘ CONTROL CONSOLE ' Fig. 2.35. Regulating Rod Limit Switch Assembly. OTT - - i ? ' , ‘ | !‘ . _?f 111 < : - . . . C ‘ ~ A ) ) , UNCLASSIFIED .« ) : . . ) ORNL~DWG 64—-6082 ’ HSv AC ' . OPENS WHEN SHIM REG. ROD NO. 1 ROD NO.{ . SWITCHES ; , o : ROD LIMIT SWITCH ) ;é’msac — ASSEMBLY CAM AT K183D ;é ' _LzKatToE : I KAI7OF LOWER LIMIT St 1 71~ CLOSED WITH OPEN WITH 1% Iv SERVO “ON" SERVO "ON" , - . | : ' _L_kazao0c . _ _lzkB170 | .- ‘ o T CLOSED WHEN SHIM-REG. | 71~ CLOSED WITH I' _ | sion | . - _ ROD ASSEMBLY CAM AT | “SERVO" ON ! 2%(240 | K244 o _ UPPER LIMIT o ) o ' : , . SHIMNO.{ ' ' 1 SERVO CONTROLLER | <% i "“ _l_ Ki38aA _L_KiTaA ACTUATE IR ~T~ WITHDRAW.ROD NO.1 . : | , f;?g‘igp,?.t‘“-' ‘T GROUP WITHDRAW . wITHDRAW : ' : - _ INSERT | WITHDRAW | WITHDRAW S20B] SHIMS NO. 2 _ [/ K240F _lzK240a S | : I ¥ _ - AND NO.3 7T~ OPENS WHEN SHIM 7~ OPENS WHEN SHIM o - _ DRIVE MOTOR _ _ | SHIM LOCATING MOTOR | : $218 | ACTUATE REG. ROD LIMIT SWITCH REG. ROD LIMIT ROD NO.4 | ~ | SEE FIG. 2.35 5224 ‘ ‘ CAM AT UPPER LIMIT SWITCH CAM AT i Lol Posimion || J _ fiz GROUP ROD ° | L _L -~ o | uppEr LMIT } TRANSMITTER [[ ™ 1 KANT? . ' WITHORAW |/ ky70H : L OPéuovcmH _Lz karoo | i S MECHANICAL DIFFERENTIAL (MANUAL} . >A— CLOSED WITH | servo “on" . 7T CLOSED wiTH é SERVO "ON" _ _ SERVO "ON" : : ‘ : 0 NOTES ! i - . ' : _ , : . SHIM LOCATING e ,‘; [t ome] ~ [ hop o s womon] | woron rorarion 71~ OPENS WHEN SHIM REG. - £ ki83c ' : ‘ {INSERT) o - AT LOWER LIMIT : ‘ ’ o AND VICE VERSA FOR COUNTERCLOCKWISE CAM ROTATION | o NO.1 INSERT O | KI7?5 ' ‘ KiT6 K182 ‘ K183 | § ' ' 2. REFER TO FIG. 2.35 FOR DETAILED DIAGRAM OF CAM DRIVE . THIS RELAY, WHEN ENERGIZED, THIS RELAY, WHEN o THIS RELAY, WHEN ENERGIZED, " THIS RELAY, WHEN ENERGIZED, , ‘ o ' _ K174 OPERATES SHIM LOCATING ENERGIZED, OPERATES OPERATES SHIM LOCATING OPERATES ROD ORIVE MOTOR : 3. SWITCH AND RELAY CONTACT ASPECT (GPEN OR CLOSED) ON - / - MOTOR N REG. ROD LIMIT ROD DRIVE MOTOR IN MOTOR IN LIMIT SWITCH INSERT IN “INSERT" DIRECTION CAM - THIS SKETCK IS FOR FOLLDWING CONDITIONS : SWITCH. WITHDRAW DIRECTION ROD WITHDRAW DIRECTION, DIRECTION, CAM ROTATES ROTATES WITH SERVO "ON m . 1. SERVO "ON", ZERO EXROR SIGNAL CAM ROTATES =~ CAM ROTATES WITH : ' fl 3 2. REACTOR SYSTEM IN|"OPERATE~RUN" MODES ) o SERVO “ON O : _ 3 ‘ 3. REGULATING ROD BETWEEN LIMITS GROUP ROD REG. ROD LIMIT. ~ ROD NO.1 REG. ROD ROD NO. 1 { ' , WITHDRAW SWITCH ASSEMBLY WITHDRAW LIMIT SWITCH INSERT : _ - WITHDRAW - . INSERT ‘ ' ‘ Fig. 2.36. Regulating Rod Control Circuit. e g i T L eatR, 112 shimming taekes place, the limit switch cam'wiil continue to rotate clock- '(]' - wise, and unless‘the shim locating motor is actuated to produce counter- | clockwise cam rotation the upper shim regulating/rod limit switch»wili be opened, its associated relay No. K240 will drop out, and contact No. K24O0F (see Fig. 2.36) will open. This wili prevent the servo system withdraw relay from transmitting power to the rod withdraw relay (No. K176). Note that the servo system is not turned off and that it continues to exert _ control in the "rod insert"‘direction; i.e., if for any reason the servo calls for the insertion of negative reactivity with the Servo rod at the upper lifiit of the regulating rod span, the rod insert requirement wiii be transmitted to relay K183, which causes the rod drive motor to ifisert the rod. This rod No. 1 insert relay, K183, cross interlocks the No, 1 rod withdraw relay, K176, such that an "insert" request from any source véverrides all "withdraw'" requests. The operator must now shim the reactor. If he chooses to shim by using rod No. 1, the servo controlled‘rod, he does so by means of the individual rod control switch (819) used for manual | bperation. This closes contact S19A and energizes K175,.the relay which causes the shim locating motor to turn the cam counterclockwisé. The limit switch closes, contact K240F closes, and the servo is in command of rod No. 1. The servo now withdraws rod No. 1 until the upper limit is again reached or until shimming requirements are met. OSince the éhim relocating motor moves the cam at a slightly lower speed than does the balance motor; the actual withdrawal of rod No. 1, in these circumstances, is a series of start and stop movements until the span of the regulating rod is re- located. The alternate method of shimming is to withdraw either or both | manually controlled rods Nos. 2 and 3 until the servo automatically returns rod No. 1 to a suitable place within the regulating rod span. ' Should the servo controller malfunction and ask for a continuous out- of-control withdrawal of rod No. 1, the withdrawal will only persist until the upper regulating rod limit is reached. The on-off power amplifier re- ~lays in the servo controller do not exert direct control of the rod drive motors (see Fig. 2.36); instead they control the "withdraw" and "insert" relays (Nos.,Kl76 and K183), and & servo malfunction has to Be accompanied by faiiure of other control-grade interlocks to become 6f consequence. ' (’ 113 Should the regulating rod limit switch mechanism misoperate in con- Junction with a servo controller malfunction such that the cam fails to rotate as the rod 1s being withdrawn by the servo, the control system interlocks call for a "reverse” (group insertion of all three rods) for any of the following conditions: 1. Reactor power above 1.5 Mw in start mode. 2. Reactor power above 12 Mw, 3. Reactor period less than 10 sec in start mode. 4. Reactor outlet temperature above 1275°F. 5. Excess fuel level in the overflow tank. "Reverse' by the control system is independent of the condition of the servo controller and is also invoked should the operator inadvertently withdraw either shim rod so as to produce any of the above listed condi- tions. '"Reverse," while not a safety system action, is used to keep re- actor system parameters within well-defined operating limits and away from other limits, such as scram level power, which are classified as safety limits. If one of the shim regulating rod limit switches remains actuated or if the cam fails to move off the mechanical stop, rod No. 1 is inhibited from further motion in one direction only. The operator can provide con- trol with rods 2 and 3 and can switch over to manual control until the trouble is cleared. The span, Ayy on Hig. 2.35, of the regulating rod limit switch assem- bly and, hence, the reactivity available to the servo controller 1s ad- Justed by means of change gears in the cam drive train. The mechanism 1s not directly avalilable to operating personnel, and all changes in span are subject to administrative control, Load Control. Control of the heat removal rate of the MSRE is achieved by controlling the effective area of an air-cooled radiator and the mass flow rate of alr through the radiator. Control of the effective area 1s accomplished by adjusting the positions of a pair of docors, one on either side of the radiator. These doors may be moved individually or together by a single motor through a clutch and brake arrangement. As indicated in Section 2.2, the doors are closed automatically to prevent 114 freezing of the coolant salt. Control of the mass flow rate is accom- plished by the use of one or two main blowers and by adjusting the posi- tion of a2 bypass damper. In principle it i1s possible to control the heat rejection rate manu- ally; however, the very long time constants in the system and the complex interactions between the varicus control devices have led to the develop- ment of a programmed sequence for increasing or decreasing load. This sequence consists of a series of steps involving movement of radiator doors, switching blowers on and off, and changing a differential pressure contrcller set point to regulate the position of a bypass damper. The operator need only operate a single load demand switch to increase or decrease load over the entire range and may interrupt the sequence at any load between 1 and 10 Mw. During these lcad changes the automatic rod controller holds the reactor outlet temperature at its set point. The reactor power rollows the load with a time lag resulting primarily from the thermal capacities and transport lags of the salt systems. TFor added flexibility in experiments, manual control of the various load control devices has also been included, with sultable provisions to limit tran- sients during transfer from manual to programmed operation. Interlocks and Circuit Jumpers. Throughout the plant there are many control devices designed to assist the operator to attain orderly opera- tion. These provide interlocks that prevent certain undesirable operations under specific conditions but are not necessary for safety. DBecause the experimental nature of the reactor demands considerable flexibility in the control system, an arrangement has been provided for bypassing these con- trol interlocks. A Jjumper board located on the main control board is used to facilitate alteration of interlock circuits and to enhance administra- tive control of such changes. It is also desirable, when the reactor is not operating, to run opera- tional check tests of the vital components that are subject to safety sys- tem control. A typical example is a prestartup test to verifly that the radiator doors operate correctly. The safety system maintains the doors closed for either the condition of low salt flow or low salt temperature (see Fig. 2.9) and would prevent such a test in an empty, nonoperating 115 gsystem. 1In order to permit such essential tests, the safety system design permits bypassing the necessary safety system contacts by using the jumper board. Jumpering of safety system circuitry is subject to stringent ad- ministrative control; i1.e., permission of the Chief of Reactor Operations must be obtained before using a jumper. The safety system Jjumpers and associated circuitry, Fig. 2.37, fulfill design criteria, as follows: 1. Safety circuit isolation and separation are maintained., This is accomplished by the "Bypassing Relay' on Fig. 2.37. 2. The Jjumpers (plugs) ére readily visible to supervisory and opersa- tions personnel in the main contrel area. 3. The circuit and its condition (whether or not energized) in each string of contacts is displayed to personnel in the main control area by means of Iindicating lamps. 4, The presence of a safety system jumper is annunciated. 5. If any safety system contact is bypassed by a jumper the control gystem cannot be put in the "operate” mode (see Fig. 2.33). 6. Failures of components in the Jumper board circuitry will not Jeopardize operation of the safety circuits. Control interlock (nonsafety system) Jumpering meets the criteria of items 2, 3, and 6 above. Figure 2.37 is a much simplified version of a typical safety circult in that 1t shows only one relay contact in the safety relay circuit. In an actual circuit there are several contacts, all of which may be wired for Jjumpers. In such a string of contacts with indicating lamps to show contact condition, the remote possibility exists that the lamps could bypass enough current around an open contact to keep the safety relsay operated. This situation requires that the lamp neutral be open so that the current path through the lamps passes to neutral via the safety relay. All safety con- tact-indicating lamp circuits contain a dropping resistor and a silicon diode in series with the lamp so that normal current through one lamp is much less than required to maintain the safety relay operated. 1In the event of an open lamp neutral, the diodes are "back-to-back"” in the sneak circuit through the safety relay. A typical silicon diode will pass only 116 UNCLASSIFIED ORNL-DWG €4-6080 JUMPER BOARD SAFETY SYSTEM ON MAIN CONTROL BOARD RELAY CABINET LS /99@754464//» ;7?25609?99006664447779066644/790- o1 / /1 / SAFETY % ? g / SYSTEM ; 7 ] BY-PASSING RELAY g “ RELAY [/ / 2] ! / / . : — U g 1 V] 7 ; g ’ T T 7777 7 I T T I NOTE: THE JUMPERS AND THE INDICATING LAMPS ARE ON THE MAIN CONTROL PANEL AND VISIBLE FROM THE OPERATOR'S CONSOLE. Fig, 2.37. Diagram of Safety System Bypassing with Jumper Board. a fraction of a milliampere of reverse current. Since relay holding cur- rents are over 100 mamp, the relay will not be prevented from dropping out. The Jjumper board presents a graphic display in full view of the opera- tor of all of the circuitry subject to bypassing. The indicating lamps are located on the board between each contact subject to Jumpering and provide immediate information as to circuit condition (whether energized or not). 117 A circuit jumper, requiring administrative formality for its insertion, can not become a forgotten clip lead. 2.4 Neutron Source Considerations MSRE fuel salts provide a substantial inherent (a,n) source.! When the core vessel is filled with fuel salt containing sufficient uranium for criticality, this intrinsic source produces more than 10° neutrons/sec. The resulting fission rate is such that statistical fluctuatlions are neg- ligible and the reactor's behavior is accurately described by the well- known kinetic equations. Kinetic calculations have established that this initial fission rate is high enough to make tolersble the worst credible startup accident. This accident is caused by the uncontrolled, simultane- ous withdrawal of all three shim-safety rods and is described in detail in Section 7.1.2. It can be concluded that the inherent (a,n) source 1s adequate for safe reactor operation. Since this source 1s intrinsic with the fuel salt, instrumentation is not required to verify its existence. Subecritical testing will recuire an external [other than (a,n) in the fuel salt] neutron source and highly sensitive detectors. This external neutron source is also used during normal, routine operation for monitor- ing the fill and startup procedures and for partial shutdowns and, thereby, provides information required for consistent, orderly operational routines. The control system is interlocked so that a count rate of at least 2 counts/sec is required during fill and startup. During partial shutdowns the fuel salt i1s maintained at the reactor operating temperature in the reactor primary loop, and the shim-safety rods are inserted. The external source is used to monitor the reactivity deficit. The source is not located inside the core vessel because (1) typical MSRE temperatures exceed the melting point of antimony and its oxides and the required cooling system would be unduly complex, (2) accessibility and, hence, source removal would be restricted, and (3) an additional in-core 7P, N. Haubenreich, "Inherent Sources in Clean MSRE Fuel Salt," USAEC Report ORNL-TM-611, O=2k Ridge National Leboratory, Aug. 27, 1963. 118 penetration is undesirable. The locations of the source and the sensitive BF3 counters are shown on Fig. 2.26, An antimony-beryllium source pro- viding 107 neutrons/sec is adequate for all purposes.8 The foregoing evaluation of source requirements takes no credit for the large photoneutron source that will be present after the reactor has been operating at power for a relatively short time. 2.5 FElectrical Power System A simplified version of the power distribution system for the MSRE is shown in Fig. 1.17 of Section 1. Normally all electrical power is sup- plied by TVA. Two main 13.8-kv supplies are available. These main supply lines originate at geographically separated locations. Switchover from the "normal" to the "alternate" 13.8-kv bus is done automatically. Standby power is provided by three diesel-engine-powered generators, and the switching operations required to put the diesels on the line are manual, as 1s diesel startup. Uninterrupted ac and dc power for the vital instruments, the control system, and emergency lighting and the power to operate the larger circuit breakers is supplied by a motor-generator- battery system that operates continuously. Normally the 250-v battery flcats on the output of the motor-generator and is kept charged. Direct- current requirements are supplied from the battery, and a 25-kw motor- generator set powered by the battery provides ac power. This 25-kw motor- generator set is normally connected to its load and is run continucusly sco that no switchover is required in the event of a TVA outage. This elimi- nates any momentary interruption that would be produced by switching time and the time required for the 25-kw motor-generator set to pick up its load. ©Such an interruption would, among other things, scram the control rods unnecessarily. The power distribution system was not designed for the purpose of insuring uninterrupted, full-power operation of the MSRE. In the event of a TVA outage, the reactor power level will be reduced until the standby diesels are started and are producing sufficient power to run all pumps 8J. R. Engel, P. N. Haubenreich, and B. E. Prince, "MSRE Neutron Source Requirements" (report to be igsued). 119 and blowers. The reliable storage battery supply described in the preced- ing paragraph has sufficient capacity to operate for spproximately 2 hr. This is ample time to start the diesels. If, after a reasonable pericd of time, it 1s not possible to restore sufficient electrical power, either from diesels or TVA, to operate the reactor system in an orderly fashion, the reactor may be drained. 2.6 Control Room and Plant Instrumentation Layout 2.6,1 Main Control Ares The main control area, shown in Figs. 2.38 and 2.39, contains 12 con- trol panels and the control console. The 12 panels comprise the main con- trol board, which is in the form of an arc centered on the console. The main control board, Fig. 2.40, provides a full graphic display of the reactor system and its associated instrumentation. These panels contain the instrumentation for the fuel salt system, the cocolant salt system, the fuel and coolant pump oil systems, the off-gas system, the Jumper board, and a pushbutton station for the various pumps and blowers. The primary recorders for the nuclear control system are alsc on the main board. The instrumentation on the console consists of position indica- tors, control switches and limit indicators for the control rods, radiator doors and Tission chambers, and other switches required for normal and emergency operation of the reactor. Routine control of the reactor is accomplished from the main control area. Control of some of the auxiliary systems is also accomplished here, with remaining portions controlled from the local panels shown in Figs. 2.38 and 2.41. 2.6.2 Auxiliary Control Area The auxiliary control area, Fig. 2.39, consists of eight auxiliary control boards, five nuclear control boards, one relay cabinet, one thermo- couple cabinet, four diesel and switching panels, and a safety relay cabi- net. The eight auxiliary panels contain the signal amplifiers, power com- puter, switches and thermocouple alarm switches, auxiliary indicators not reguired for immediate reactor operation, the fuel and cooclant pump 120 ORNL~DWG 64-808 W UNCLASSIFIED : ~ e __ BUILDING 7509 1 \ b (orrice) HEALTH PHYSICS - : - ROOM DOWN INS“{ggMGEENT STORA Tl L) Al 4 N | | / : OFFICE OFFICE OFFICE — . . ' DOWN ) INSTRUMENT SHOP - } MAIN : "F CHANGE. : . ROOM i _ i F i [l sfifxf'?z\%?" | wor I NUCLEAR Ll - : DATA ROOM CHANGE I | PENETRATION - SEE FIG.2.43 ay ROOM L __ROOM CONTROL ROOM %OT CHANGE ROOM REACTOR - CELL ENRICHER PANELS \ ‘ DRAIN TANK CELL i ) VENT fl HOUSE ROOF v REMOTE MAINTENANCE 1 e CONTROL ROOM _ ELEVATION 862ft Qin. N Fig. 2.38. Main Floor Layout of Building 7503 at 852-ft Eleva- tion. . " ALY - UNCLASSIFIED 7 ORNL—DWG 64-630 . N OBSERVATION @ -— fi‘) ' cnu.cmr\ . : r . Y ~ . . . ¥ - - F - - F - - © : 4 THERMOCOUPLE CABINET \_I : POWER (B IshlflelE/om = . PANELS . , AUXILIARY PANEI.S _ _ DATA AND DIESEL AUXILIARY CONTROL AREA NUCLEAR SUPERVISOR PANELS / ROOM 4"%‘1 ! '\ s‘/-'.. oaan |5I4I3I=I'I | RELAY cnemsr/ SAFETY CABINET/ - - I - T - - @ ~ 1 I Fig. 2.39. Main and Auxiliary Control Areas. NS » a A » Fig. 2.40. Main Control Board. UNCLASSIFIED PHOTO 37955 L] 122 UNCLASSIFIED ORNL-DWG 63-8399 up [ T - [ up : = L ATTERY ' - - ] % RooM o LUNCH - Mgl MAINTENANCE - R | SHOP . v : SERVICE "Jgg\ S ) ] g | ROM N = ,f)> LUBRICATION OIL Nt [[ _ TRANSMITTER CONTROL. PANELS . - ROOM pown TC SCANNER o I g . ' ‘_-:-iJ_.:-;.J LJ PANELS : T SPECIAL : R NGl i EQUIPMENT - X REMOTE {1 _ ROOM "] MAINTENANCE | REACTOR - CELL CELL LIQuUID DECONTAMINATION ” FUEL m CELL PROCESSING Cl WATER ' ' EQUIPMENT 1, SWITCH P ” HOUSE ' A —— o BLOWER BUILDING C%QTTERR/ HOUSE VENTILATION PANEL SYSTEM FILTER HOUSE COVER GAS DIESEL TROL PANE bo DIESEL Jj_con OL PANELS Wn BUILDING 7503~ELEVATION B840 ft Oin. LEVEL Fig. 2.41. Layout of Building 7503 at‘840¥ft Elevation. microphone amplifier and noise level indicator, and the substation alarm monitors. The relay cabinet contains oniy the relays for the main control circuitry. The thermocouple cabinet contains a patch panel with pyrometer jacks. All thermocouples in the system except those on the radiator tubes are terminated in this cabinet. | The five nuclear panels contain some of the instrumentation for the process radiation system, the oil system, and the sampling and enriching system. The health physics mbnitofing alarm system, the high level gammé- chamber electrometers and switching panel, and the reactor control nuclear amplifiers, servo amplifiers, and other equipment for the nuclear control system are also located here. Four panels contain the instrumentation for the operation of the'emergency diesel generators. The distribution panels w 123 for the control circuits are located on the south wall of the auxiliary contrcl room. 2.6.3 Transmitter Room The transmitter room, Fig. 2.42, consists of nine control panels, one solenoid power supply rack, and a transmitter rack. Panels 1 and 2 are for the leak detectors system, and panels 3 and 4 are for fuel drain tanks 1 and 2, coolant drain tank, fuel flush tank, and the fuel storage tank welgh instrumentation. Panels 5 and 6 are the fuel and coolant pump bub- bler-type level control panels, panel 7 is the freeze valve air control panel, panel 8 is an installed spare panel, and pancl 9 is for the sump bubbler system. The solenoid power supply rack contains most of the sole- noids for the fuel salt system and the power supplies and distribution panels for the electric transmitters and differential transmitters. The UNCLASSIFIED ORNL-DWG 64-629 N 1 ) - © X 1 | r ’ 1 LDK 4 ¢ 2 9 >~ ' 3 REMOVABLE 8 ACCESS PLUG WEIGHT < - || PUMP 4 p—— —— . 6 5 SOL. RACK TRAI;%I(\;III(TTER LADDER ~a DOWN {1 1l \I/F——| Fig. 2.42. Transmitter Room. 124 transmitter rack contains the remote amplifiers for the electric signal transmission system and also the current-to-air transducers. A1l electrical and pneumatic lines that connect transmitter room equipment to input signals from the reactor, drain tank, water room, and all other areas outside the building enter through sleeves in the floor. Output signals from the transmitter room equipment are conducted via feeder cable trays to the main cable tray system, which runs north-south in the building on the 840-ft level. 2.6.4 Field Panels Additional panels are required for the auxiliary systems. These (see Figs. 2.38 and 2.41) are located near the subsystems which they serve. Four panels for the cooling oil systems are located in the service room and service tunnel; one coocling and treated water system panel is located in the fan house; three panels are located on elevation 852 ft at column line C-7 for the sampling and enriching system. Two cover gas system pan- els are located in the diesel house. Two containment air system panels are provided. One is located at the filter pit, and the other is located in the high bay area. 2.6.5 Interconnections The control areas and the controlled systems are interconnected by cable trays and conduit. This system includes the main cable tray system running north-south in the building, with branch trays and conduits Jjoin- ing the main tray. These main trays carry all tubing for the pneumatic system, thermocouple leadwire, and the ac and dc control wiring. Each category of signal runs in a different tray to avoid mixing of the signals. Risers connect the panels and cabinets in the control area on elevation 852 ft with the main trays. No exposed trays are in the main or auxiliary control rooms. Safety system wiring is completely enclosed in conduit and junction boxes and is separated from control and instrumentation wiring. 125 2.6,6 Data Room The data room (Figs. 2.38 and 2.43) is designed primarily to house the data-handling system, but it also serves as the information process- ing center for the reactor. The room contains storage space for all types of instrument data and log sheets. It also containsg the data-plotting equipment used with the data-handling system and other special data dis- play and processing equipment. This room 1s located in the north end of the reactor building next to the main controcl room. Therefiare entries from the hallway and the main control room. A large glass window is located so that the main control panel can be seen from the room. The data-handling system is installed along the west and north walls of the room, Fig. 2.43. The data system console is located so that the reactor control panel can be seen by an operator at the consgole. At least two desks and several tables are located in the room for use of reactor operations and analysis perso?nel. Data-plotting equipment and record storage cabinets are placed at cconvenlent locations. A typieal process computer system is shown in Fig. 2,44 . The input signal cables for the data system are brought up to the bottom of the room in cable trays and conduits. Holes are cut in the floor beneath the signal input cabinet for routing the cables to the cabinet. Cable trays and conduit are also provided beneath the floor for signal lines to and from the main and auxiliary control areas. The data room is air conditioned by the system which is used for the control room and offices. © 126 @ ORNL-DWG 64-628 j 7 UNCLASSIFIED @ BT s 7 27 e i e e ALt o e 2 L ot bk el XL ; N B . N N N N 9 N N N [ N N g ' \ - \ - N T N © \ . N § ’ \ N R Q N N N \ N \ N N N N INPUT | anatos | AMOC TO INPUT N OUTPUT INPUT NVERSION OUTPUT R NO.2 NO. 3 NO. 4 S . . SYSTEM 4 . N , \ / N ANALOG NO. § g ! s ACCESSORIES ° N & § ™~ N \ N N BASIC Y INPUT CORE DRUM s N TRA N outpur | JNPUT | CENTRAL | yemory | meEmory X nNo.o | QUTPUT |PROCESSOR| iy UNIT N - SYSTEM . N \ N N \ - N £ N o ' S MAGNETIC < ;: TAPE " N N S C 3 — ] N MAGNETIC 30in. 30in. N TAPE LOGGER LOGGER N N ) N T N PUNCH AND 'Q READER N [ N \ D N N CONSOLE COMPUTER | —=7T—KEY BOARD 16in. LOGGER DESK 1 Fig. 2.43. JIS in. LOGGER Layout of Data Room. —— %flf/////////////////////// AL " " AL L) UNCLASSIFIED ORNL-DWG 64-631 PROGRAMMING AND MAINTENANCE STATION FLEXOWRITER HIGH SPEED PAPER TAPE PUNCH HIGH SPEED PAPER TAPE READER BASIC COMPUTER 5YSTEM CENTRAL PROCESSOR CORE MEMORY DRUM MEMORY INTERRUPT SUBSYSTEM MASTER INPUT-OUTPUT CONTROL SUBSYSTEM PROGRAMMING AND MAINTENANCE PANEL INPUT-OUTPUT SYSTEM ANALOG INPUT SUBSYSTEM ANALOG OUTPUT SUBSYSTEM CONTACT CLOSURE INPUT SUBSYSTEM CONTACT CLOSURE OUTPUT SUBSYSTEM COUNTER SUBSYSTEM OPERATOR-PROCESS COMMUNICATION STATION LOGGING TYPEWRITER OPERATOR’S CONSOLE CONTROL PANEL DISPLAY INPUT SWITCHES FUNCTION SELECTOR FUNCTION MATRIX ALARM PRINTER Fig. 2.44. Typical Process Computer System (TRW-340). LZT 128 3. PLANT LAYOUT 3.1 Equipment Arrangement The general arrangement of Building 7503 is shown in Figs. 3.1 and 3.2. The main entrance is at the north end. Reactor equipment and major auxiliary facilities occupy the west half of the building in the high-bay area. The east half of the building contains the control room, offices, change rooms, instrument and general maintenance shops, and storage areas. Additional offices are provided in a separate building that is east of the main building. Equipment for ventilating the operating and experimental areas is located south of the main building. A small cooling tower and buildings containing supplies and the diesel-electric emergency power equipment are near the west side of the maln buillding. The reactor primary system and the drain tank system are installed in shielded, pressure-tight reactor and drain tank cells, which occupy most of the south half of the high-bay area. The reactor cell is 24 ft in diameter and 33 ft in height. It is surrounded by a 30-ft-diam steel tank, and the 3-ft annular space is filled with a shielding mixture of sand and water. The top of the cell is covered with removable shielding blocks that extend 3 1/2 ft above the floor elevation of 852 ft and are sealed with a welded membrane. Ad joining the reactor cell on the north side is the drain tank cell, which extends 5 ft deeper than the reactor cell to provide gravity flow from the reactor to the drain tanks. This cell is rectangular (17 1/2 X 21 1/2 f£t), with 3-ft-thick heavily reinforced concrete walls lined with stainless steel. Access to the equipment is through the removable roof blocks, which are arranged in two layers with a sealing membrane between. Several other shielded cells are located in the north end of the high-bay area, as shown in Fig. 3.1. They provide shielding and venti- lated isolation for the fuel processing cell, the waste cell, and two cells for equipment storage and decontamination. Fig. 1.3 (see sec. 1.2) shows how the components of the fuel circu- lating system are arranged in the reactor cell. The reactor tank and the " ’ - W ¥ UNCLASSIFIED ORNL DWG. 63-4347 OFFICE FUEL PUMP -~ BATTERY LUBE ROOM MAINTENANCE SHOP : CHEMICAL SERVICE LABORATORY TUNNEL COQOLANT PUMP SPECIAL : EQUIP, ROOM "TpAfis 1l e CELL I m- SPARE CELL. Q. STORAGE: MAINTENANCE cel TO0 VAPOR PRATICE CELL - . SYSTEM PUMP WASTE - ELL - 1 . {REGULATOR . CELL FUEL SALT FUEL FLUSH L DRAIN TANK TANK NO. 2 ¥ | [ . ‘ FUEL DRAIN THERMAL T BLOWERS TANK NO. t- SHIELD - [ REAGTOR . b - CELL C.. . ANNULUS *ELEC. SERVICE AREA BELOW . RADIATOR . BLOWERS + i Fig. 3.1. First Floor Plan of Reactor Building. 4 OlL. SYSTEM 10 FILTERS ' - AND STACK CELL VENTILATION AND BLOCK VALVE COOLANT SALT DRAIN TANK . 62T UNCLASSIFIED , ORNL DWG. 64-397 : 3 AND 10-TON CRANES CI0LANT SALT PUMP SHIELD LIQUID WASTE CELL DRAIN TANK CELL RADIATOR BYPASS DUCT— h COOLANT SALT DRAIN TANK / FUEL ACTOR VESSEL USH RMAL SHIELD FUEL ANK NO. 1 ORAIN LINE FUEL DRAIN TAN" NO. 2 - Fig. 3.2. Elevation Drawing of Reactofr Building. octT 131 thermal neutron shield are located south of the center of the containment cell, and they rest on the floor. The primary heat exchanger is positioned north of the reactor vessel and above it in elevation. The freeze flanges on the fuel and coolant salt lines attached to the heat exchanger are placed with adequate working space for remote-maintenance operations. The fuel-circulating pump is mounted east of the reactor vessel and above it and the heat exchanger, sc that the free surface of salt within the pump bowl 1s the highest point in the system. Underneath the bowl is the overflow tank and surrocunding it is an electric furnace. All ser- vice piping flanges and electrical Jjoints are located close to the pump bowl so that remote replacement of the pump bowl will be easier. The arrangement of the drain tank cell is illustrated in Fig. 1.3. The flush salt tank is at the south end of the cell, and the two fuel salt tanks are north of it. Space is allowed to the east and west for remote cutting and brazing equipment in case either of the tanks must be removed. In the northwest corner of the cell 1s a 25-kva transformer, which supplies the power for resistance heating of the 1L l/2-in. drain line connecting the reactor vessel to the tanks. The coolant cell abuts the reactor cell on the south. It is a shielded area, with controlled ventilation, but it is not sealed. The coolant salt circulating pump is mounted high in the coolant cell, as shown in Fig. 1.3. The radiator is at a lower elevation, to the west, and confined in the air-cooling duct. The coolant drain tank is underneath the pump and radiator, at the bottom of the cell. The blowers that supply coocling air to the radiator are installed 1in an exist- ing blower house along the west wall of the coclant cell. Rooms containing auxiliary and service equipment, instrument trans- mitters, and electrical equipment are located along the east wall of the reactor, drain tank, and coolant cells. Ventilation of these rooms is controllied, and some rooms are provided with shielding. The high-bay area of the building over all the cells is lined with metal to minimize inleakage. Ventilation is controlled and the area is normally operated at slightly below atmospheric pressure. The effluent air from this area and from all other controlled-ventilation areas is 132 filtered and monitored before it is discharged to the atmosphere. The containment ventilation equipment consists of a filter pit, two fans, and a 100-ft-high steel stack. This equipment on the south side of the main building is connected to the building by one ventilation duct to the bot- tom of the reactor cell and another along the east side of the high bay. The vent house and charcoal beds for handling the gaseous fission products from the reactor systems are near the southwest corner of the main building. The carbon beds are installed in an existing pit that is filled with water and is covered with concrete slabs. The vent house and pit are also controlled-ventilation areas. Gases from the carbon beds are monitored continuously for radiocactivity and are discharged into the ventilation system upstream of the filters. 3.2 Biological Shielding The MSRE work areas are divided into five types based on expected radiation levels: (1) areas with high radiation levels that prohibit entry under any circumstances, such as the reactor and drain tank cells; (2) areas that can be entered a short time after reactor shutdown, such as the radiator area; (3) areas that can be entered at low reactor power levels, such as the special equipment room; (4) areas that are habitable at all times; and (5) the maintenance control room, which is the only habitable area on the site when certain large-scale maintenance operations are being performed. The MSRE shielding i1s designed to permit prolonged operation at 10 Mw without exposing personnel to more than a few mrem per week. The areas which will be entered routinely and will have unlimited access will be essentially at the normal background level for the Oak Ridge vicinity. However, there are several class 2 and class 3 areas which will have activity levels considerably above background and which will be entered only occasionally. These are located near the reactor or drain-tank cell penetrations. One such limited access area is the coolant cell, in which the background activity might be as high as 100 mr/hr. The blower house is also a limited access area, since the radiation level may be about 20 mr/hr near the blowers. Although the special equipment room 1s considered 133 a limited access area, the radiation field should not exceed about 10 mr/hr. The south electric service area, another limited access portion of the building, will have a generally higher radiation level of approxi- mately 200 mr/hr. All these estimates are based on operation of the re-~ actor at the 10-Mw power level. On the infrequent occasions when these areas must be entered, radiation surveys will be made so that the working time and exposures can be kept to safe levels, When the reactor is subcritical, all areas, except Tthose of type 1 (the reactor, drain tank, and fuel processing cells), may be entered a few minutes after the reactor is shut down and after a radiation survey is completed. 1In general, it is eXpected that access can be on an un- limited basis unless "hot spots' are found. In any direction from the reactor vessel, a sufficient thickness of shielding material has been provided to reduce the radiation level in habitable areas to background levels. In some directions the shielding exists as several widely separated barriers. The general arrangement is described below, beginning at the reactor vessel. Reactor Cell. The reactor vessel at the south side of the cell is surrounded by a stainless steel tank with a 16-in. thickness of iron and water. This thermal neutron shield is located within the secondary con- tainment vessel which, in turn, is within another tank to provide a 3-ft- wide annular sgpace, which is filled with megnetite sand and water., The outer tank is surrounded by & cylindrical, monolithic concrete wall 21 in. thick, except in the area facing the coolant cell, 1In this area special shielding materials are installed to give equivalent protection. There are no equipment or personnel access openings other than those through the ftop of the cell. The top of the reactor cell is flat and is covered with two 3 1/2-ft layers of large, removable, concrete blocks, as shown in Fig. 3.3, The bottom layer is high-density (sp gr = 3+) concrete, and the upper layer is ordinary concrete. The joints in the lower layer of blocks are filled with steel plate inserts. A stainless steel sheet (1/8 in. thick) is sandwiched between the two layers and is welded to the edge of the tank to provide a gastight seal, 134 UNCLASSIFIED ORNL DWG. 64-598 L /./ \R \ “ o 2 O__—’B_:O——f)-gfig oIy / %o/"/.?jrfi: *fir&\% TN Yy’ d ‘\o o/// ‘ \0\ \ M | \i \ o;/// l \\o f —= . ,q\L_l{ T ] \ eyt [{’ Y \5" //__\ \ i..xx.. —— o | "‘} o ‘ ) - d ! ” | \ O N / \\ 1o ,:—: ~—=r /// \ IRy N I L/ o§ I I —L ;/o o F\\ I / ® / oY o \ \o}#\\dx | &fg z/ N g fi/ ) PLAN - /_ggiDréDOWN > N i “ N A\ \lRON SUPPORT BEAM \ /Access SHIELDING PLUGS INSERT SHIELDING SO CELL WALL SECTION "xXx" Pig, 3.3. Arrangement of Shielding Blocks on Top of Reactor. 135 The service penetrations through the reactor cell walls pass through sleeves filled with magnetite concrete grout or magnetite sand and water. Where possible, these lines have an offset bend. The penetration of the 30-in.-diam air exhaust line through the bottom hemisphere of the con- tainment vessel required special treatment because of the size of the opening. A shadow shield of a 9-in. thickness of steel is provided in front of the opening inside the cell, and a 12~in.-thick wall of stacked blocks is erected outside the cell at the foot of the ramp to the coolant cell, Coolant Cell., The top and sides of the coolant and coolant drain tank cell provide at least 24 in. of concrete shielding as protection against activity induced in the ccolant salt while the reactor is pro- ducing power. The large openings provided between the coolant cell and the blower house for the cooling air supply to the radiator, however, make it difficult to shield the blower house from this induced activity. Space has been provided for additional shielding in the form of stacked blocks, should they be found necessary after full power is reached. Drain Tank Cell. The drain tank cell has a minimum thickness of 3 ft for the magnetite concrete walls facing accessible areas. The top of the cell comsists of a layer of 4-ft-thick ordinary-concrete blocks covered by a layer of 3 l/2-ft—thick ordinary-concrete blocks. The pipe lines penetrating the cell walls have off'sets, and the smaller pipes are cast into the walls. Shielded plugs are provided for the larger penetra- tions. Other Shielding. The 1/2-in. offgas line from the reactor cell to the charcoal beds is shielded by 4 in. of lead as it passes through the coolant drain tank cell. Barytes concrete blocks are stacked to a thick- nesg of 5 ft above the line in the vent house, and 17-in.~-thick steel plate is provided above the line between the vent house and the charcoal beds. The charcoal beds are covered with two 18-in.-thick by 10-ft-diam barytes concrete blocks and an additional 30-in. thickness of stacked barytes blocks. The walls of the filter pit for the containment ventilation system are 12 in. thick and the roof blocks over the filters are 18 in. thick. 136 The thicknesses of the walls and tops of the auxiliary cells are given in Table 3.1. Additional shielding is provided by blocks stacked on the west side of the fuel processing and the decontamination cells. Fauipment in the fuel-circulating and drain tank systems will be repaired or replaced with remote-handling and -viewing equipment. A heavily shielded maintenance control room with viewing windows is located above the operating floor. This room will be used as a protected place to operate remotely controlled equipment when several roof shielding plugs are removed and radiocactive equipment 1s to be transferred to a storage cell. Fguipment in the coolant cell cannot be approached when the reactor is operating, but the induced activity in the coclant salts is sufficiently short lived to permit The ccolant cell to be entered for direct mainte- nance shortly after reactor shutdown. Table 3.1. Descriptions of Auxiliary Cells Location Concrete Wall Thickness Thickness N £ Cell Inside Dimensions Fioor of Top ame oL e N-5 E-W N-8 X B-H Flevation N s B W Blocks Columns Columns (in.) (in.) (in.) (in.) (in.) Tuel processing cell =5 A-B 12 £+ 10 in. X 14 £t 3 in. 831 £t O in. 18 4y 182 128 48 Decontamination cell 3-4 A-B 15 £t Q0 in. x 14 Tt 3 in. 832 £t 6 in. 18 18 18 12 30 Liguid waste cell 2-3 A-B 13 £t 0 in. X 21 £t 0 in. 828 £t 0 in. 18 18 18 18 30 Remote maintenance cell 2-3 C-D 13 £t O in. X 2L £t O in. 831 £t O in. ig 18 18 18 30 Hot storage cellP 34 C-D 15 ft 0 in. X 14 ft 3 in. 832 £t 6 in. 18 18 122 18 30 Spare cell 45 C-D 13 f4 6 in. X 14 ft 3 in. 831 £t 0 in. 18 12 12 18% 30 aPlus additional stacked blocks as required. bThe hot storage cell is lined with ll-gage stainless steel to the clevation of 836 ft 6 in. LET 138 4. SITE FEATURES* 4,1 Location The Molten-Salt Reactor Experiment is located in the Roane County portion of the Oak Ridge area of Tennesgee. The site is owned and con- trolled by the Atomic Energy Commission. The MSRE is in the Melton Valley area of the Oak Ridge National Laboratory southeast of the Bethel Valley, X-10 area, in Building 7503, which formerly housed the Aircraft Reactor Experiment (ARE). Building 7500, which formerly housed the Homogeneous Reactor Experiment No. 2 (HRE) and which is 2000 ft west-northwest of the MSRE, now houses the Nuclear Safety Pilot Plant (NSPP). The High Flux Isotope Reactor is being constructed 1500 ft south-southeast of the MSRE site. The main laboratory area of X-10, which is approximately 1 mile to the northwest, 1s separated from the MSRE site by Haw Ridge. Melton Hill bounds the valley on the southern side. The MERE site is located within a well-established AEC-controlled area. The ARC Patrol covers the roads and adjacent area to restrict public access to certain designated routes through the controlled land. A perimeter fence encloses the entire area of the MSRE. Approximately 40 people will be present in the exclusion area during the day shift and 6 or 7 on the evening and night shifts. The location is shown in respectively increasing scale in Figs. 4.1, 4,2, and 4.3. 4,2 Population Density The total population of the four counties (Anderson, Knox, Loudon, and Roane) closest to the MSRE site is 370,145. Of this number, 177,255 people are located in cities with populations greater than 2500 persons. The rural population density in these four counties is about 135 persons per square mile. The average population density within a radius of 27.5 miles of the MSRE site, as determined from the data obtained in the 1960 census, is 147 persons per square mile. Table 4.1 lists the surrounding ¥This chapter was compiled by T. H. Row, Reactor Safety Group of the Reactor Division, Osk Ridge National Laboratory. UNCLASSIFIED ORNL-LR-DWG 35219 - T T 1 ” e A —em | “\\ \N \(\ ) ,‘/‘j—\‘\:&v/ // ' \ \ Midglesboro AL —.u——?S s R - ' --——---—7’:’-——--——--— 55 - R W P F"\A /fli?_’» - Jellico T = 25/ LnFoHefle NORRIS DAM 23 miles Lake City r“ CHEROKEE DAM| § , e[ | < Ciintordid ) ok AN IR L] "y 7 Ly, Wortburg Jefferson Crty %} ai-| DOUGLAS DAM |3 N . et - ; Spurh\’\\\\\\\\\\\\\\\\\\\\\\\\\\\\\ . e arville E&a miles / W z [ 1 /, Alcoa P .'“"“‘\/\o - Gotlinburg ’_/_/'/WQ,&\A Y FT. LCUDQUN DAM - < (FQ\‘ e, m—— - | McMinnville 12 miles A\~ - w- ” GREAT FALLS DAM cnear swocr W 3am ;o . ———— ¢ . o \ 75 miles 7 / . 2. o o, NAT. PAR -l L= o /D z S "‘:f'“‘f Cherokee /"/ - 1 v'?\J‘“\(‘\} % - n\\ ., 123 5592, Forgson gy s Dilsbore t ! ¢ y 4 N ' ‘ 3 S, \“‘-fl y ‘ 2 ~ ; ' . e ! . . Franklin fi\ N % : )rfi\\**s ; / TRy W o [ "-:;C?, \\\ p | : / ) b Murphy . { 1 3] » c = i /1_\7"&‘? = . \ .‘\\(/\ l ‘ 3 Ducktown ; e I/ —l S ; ____‘/ ! Fig. 4.1. Area Surrounding MSRE Site. 6T 140 UNCLASSIFIED ORNL-LR-DWG 4406R \ TO ORGDP ‘.\- \o 5 mile O _TOWER SHIELDING FACILITY WHITE WING MELTON GATE HILL DAM Fig. 4.2. Contour Map of Area Surrounding MSRE Site. g v R 141 - ¥ hm.,, Eategnee WATTS BAR LAKE - TECLS { cHE BASIN AREA MICAL WASTE LY . T [ . ‘»\ + = PATROL - ™ > 2R OA D gu. [ofe —— S > MIOWAY ENTRANCE I \ Ofo Olo ' N ‘ - = Y3 o= ' - _ w : . ) . L H - ; , . : o o UT-AEC. AGRICULTURAL // MAIN RESERVOIR oW : _ ! RESEARCH L ABORATORY 5 o ‘ i : . ' , ROAD QEFICIAL . . 2 WOAD VvALLEY i d () ’ ARC-TVA PROTECTIVR FENCE SOUTHERN REGID A ~ & x 1000 300 O 1000 2000 3000 4000 8000 ==ttty SCALE IN FEEY = 3 KEAR MOLLOW ENTRANCE~ = N KERR—s . HOLLOW—s—ROAY Route 62 ’ , . Lo 7 : X %{ : X ‘ UNCLASSIFIED ORNL DWG, 63-3741 2y ”"w, P e & - o Fig. 4.3. Map of Ok Ridge Area. bttt e A b R 142 Teble 4.1. Population of the Surrounding Towns ‘Based on 1960 Census City ~ Distance Time Downwind: or from Site® Direction Population (%) Town‘ (miles) - Night Day Osk Ridge 7 NNE 27,124 5.6 5.5 Lenoir City 9 SSE 4,979 4,3 6.0 Oliver Springs 9 ‘N by W 1,163 2.3 2.7 Martel 10 SE 500P 1.4 2.8 Coalfield 10 W 650P 0.5 1.1 Windrock 10 N by W 5500 - 2.3 2.7 ‘Kingston 12 WSW - 2,010 9.5 11.3 Harriman | 13 W o 5,931 2.2 3.7 South Harriman 13 W 2,884 2.2 3.7 Petros | 14 W by 790P 1.4 2.8 Fork Mountain 15 ' - NNW 4 700P 2.3 2.7 Emory Gap 15 W ' 500P° 2.2 3.7 Friendsville 15 SE - 606 1.4 ‘2.8 Clinton 16 NE 4,943 11.6 2.0 South Clinton ‘16 NE 1,356 11.6 9.0 Powell 17 ENE 500P 8.3 6.8 Briceville - 19 NNE 1,217 5.6 5.5 Wartburg 20 NW by W 800P 1.4 2.8 Alcoa 20 ESE 6,395 - 2.0 2.0 Maryville 21 ESE 10, 348 2.0 2.0 Knoxville 18 to 25 E 111,827 1.5 2.7 - Greenback B 20 S by E 960° 5.5 4,9 Rockwood 21 Wby S 5,343 2.2 3.7 Rockford 22 SE 5,345 1.4 2.8 Fountain City - 22 ENE 10,365 8.3 6.8 Lake City 23 NNE 1,914 5.6 5.5 Norris - 23 NNE 1,389 5.6 5.5 . Sweetwater 23 SSW 4,145 8.4 12.7 Neubert 27 ENE 60QP 8.3 6.8 John Sevier 27 E 7520 1.5 2.7 Madisonville 27 s 1,812 5.5 11.9 Caryville 27 N by E o 1,234P 9.5 6.1 Sunbright , 30 W 6002 0.5 1.1 Jacksboro 30 ‘ N by E 5770 9.5 - 6.1 Niota | 30 SSW | 679 8.4 12,7 gBased on data obtained for HFIR site. ?Taken'from 1950 census. “ () 143 communities with a population of over 500 and their approximate distance and direction from the site. The rural population density in the four surrounding counties is given in Table 4.2. A number of facilities are located within the AEC-controlled area, and the approximate number of employees at each plant is given in Table 4.3. The numbers of employees indicate the total employment at each facility and do not attempt to show the breakdown according to shifts., However, most of these employees work the normal 40-hr week on the day shift. Table 4.2. Rural Population in Surrounding Counties . Estimated Population Total Rurald ~ Topulation - County Area® Popula- (NDGHSItyl Within Within Within (sq. mile) tion A peo?le) 10-Mile 20-Mile 30-Mile per Sd. M€/ padius Radius Radius Anderson 338 26,600 79 395 14,200 22,800 Blount 584 38,325 66 0 6,720 23,200 Knox 517 138, 700 238 13,100 46,400 96,000 Loudon 240 18,800 78 6,080 16,900 18,700 Morgan 539 13, 500 25 225 3,625 8,630 Roane 379 12, 50C 33 3,070 9,170 11,110 aDoes not include area within Oak Ridge reservation. b1960 census; does not include communities with population of 500 or more, An estimate was made of the distribution of the population in each of the 16 adjacent 22 l/2° sectors of concentric circles originating at the MSRE. radii of 0 to 1, 1 to 2, 2 to 3, 3 to4, 4 to 5, 5 to 10, and 10 to 20 Seven different distances were considered from the MSRE site: miles. The values obtained, given in Table 4.4, are representative of the population in this area at all times. Very little change is experi- enced due to either part-time occupancy or seasonal variation. The popu- lation density in the area has been reasonably stable for a number of years and is anticipated to remain so. 144 Table 4.3. Number of Employees in Specific Oak Ridge Areas in June 1963 Distance Total Ares from Site Direction Number of (miles) Employees MSRE 35 NSPP Q.4 WNW 6 HFIR 0.25 SSE 40 ORNL 3830 X-10 area personnel 0.75=1.25 NW 3327 Construction personnel 0.75~1.25 NwW 194 7000 area personnel 1.0-1.4 NNE 309 HPRR 1.1 ESE 12 Tower Shielding Facility L.4 S 15 EGCR 2.0 NE 150 Melton Hill Dam® 2.25 S Construction persomnnel, June 1964 25 Normal operation {remotely con- 2 trolled; K-25 {Gaseous Diffusion Plant) 2751 K-25 area personnel 5.0 WINW 2678 Construction personnel 73 Y-12 (Electromagnetic Separations 5,75 NINE 6866 Plant ) Y~12 area personnel 5507 ORNL personnel 209 Construction personnel 450 University of Tennessee Agricultural 6.5 NE 160 Research Laboratory Bull Run Steam Plant® 11.25 NE Construction personnel July 1964 1900 December 1964 1500 July 1965 1100 December 1965 700 Normal operation (one unit) 190 “Estimated from construction schedules, TVA, Knoxville, Tenn., June 5, 1963. Table 4.4. Estimated Population Distribution® Population in Given Sector Radius (miles) NNE NE: FNE E ESE SE SSE g SSW W WEW W W W NITW 0-0.5 0 0 0 Q 0 0 0 0 6 0 35 0.5-1 29 0 0 0 0 0 0 0 789 1,445 1-2 309 0 150 0 12 0 15 0 221 738 18 2-3 0 0 0 24 0 41 20 0 90P 20 90 45 0 0 0 0 34, 0 0 24 41 87 40 20 135 135 135 45 0 0 0 0 45 0 0 &7 g7 90 60 40 180 180 180 45 2,751 0 200 5-10 7,944 20,428 460 7,706 7,706 7,706 83 6,546 1,567 1,564 781 781 781 781 781 781 10-20 5,320 13,318 6,650 56,414 55,914 23,131 6,388 5,660 4,700 4,700 1,563 3,573 16,741 2,542 1,500 4,190 aIncludes Does not c Does not Oak Ridge Plants include Melton Hill Dam, see Table 4.3, include Bull Run Steam Plant, see Table 4.3. ST 146 4.3 Geophysical Features 4.3.1 Meteorology Oak Ridge is located in a broad valley between the Cumberland Moun- talns, which lie to the northwest of the area, and the Great Swmcky Moun- tains, to the southeast. These mountain ranges are oriented northeast- southwest and the valley between is corrugated by broken ridges 300 to 500 ft high and oriented parallel tc the main valley. The local climate is noticeably influenced by topography. 4.3.2 Temperaturer The coldest month is normally January, but the differences between the mean temperatures of the three winter months of December, January, and February are comparatively small. dJuly is usually the hottest month, but differences between the mean temperatures of the summer months of June, July, and August are also comparatively small. Mean temperatures of the spring and fall months progress orderly from cooler to warmer and warmer to cooler, respectively, without a secondary maximum or minimum. Temperatures of 100°F or higher are unusual, having occurred during less than one-half of the years of the period of record, and temperatures of zerc and below are rare. The annual mean maximum and minimum temperatures are 69.4 and 47.6°F, respectively, with an annual wrean temperaturce of 58.5°F. The extreme low and high temperatures are -5 and 103°F, recorded in December 1962 and Scptember 1954, respectively. Tegble 4.5 liste the average monthly tem- perature range based on the period 1931 to 1960, adjusted to represent obgervations taken at the present standard location of the weather station. Information on the temperature gradient frequency and mean wind speed for each month were presented in a recent report2 on the meteorclogy of 1U.5. Dept. of Commerce, Weather Bureau, Asheville, N.C., Local Cli- matological Data, 1962, Oak Ridge, Tennessee, Area Station (X-10), April 23, 1963. “W. F. Hilsmeier, "Supplementary Meteorological Data for Osk Ridge," USAEC Report ORO-199, Oak Ridge Operations, March 1963, 147 Table 4.5. X-10 Climatological Standard Normals (1931 to 1960) Maximum Minimum Aversage Temperature Temperature Temperature January 48.9 31.2 40.1 February 51.6 31.8 41.7 Mzrch 58.9 37.0 48.0 April 70.0 46,3 58.2 May 79.0 54.8 66.9 June 86.1 63.3 T .7 July 88.0 66.7 77 .4 August 87.4 65.6 76.5 September 83.0 59.2 71.1 October 72.2 47.7 60.0 November 58.6 36.5 47,6 December 49 .4 31.3 40 .4 Annual 69.4 4'7.6 58.5 the Oak Ridge area. The seasonal and annual averages derived from this information are presented in Fig. 4.4. 4.3.3 Precipltation? Precipitation in the X-10 area is normally well distributed through- out the year, with the drier part of the year occurring in the early fall. Winter and early spring are the seasons of heaviest precipitation, with the monthly maximum normally occurring January to March. A secondary maximum occurs in the month of July that 1s due to afterncon and evening thundershowers. BSeptember and Octobery are usually the driest months. The average and maxinmum annual precipitation are 51.52 and 66.2 in., respectively. The maximum rainfall in the area in a 24-hr period was 7.75 in., recorded in September 1944. The recurrence interval of this amount of precipitation in a 24-hr period has been estimated to be about 70 years. The maximum monthly precipitation occurs normally in March and has a value of 5.44 in. The average monthly precipitation is given in Table 4.6. FREQUENCY (%) Fig. 4.4. 148 UNCLASSIFIED ORNL-DWG 63-2543 [] oar NIGHT 20 15 10 5 e 0 ' . 20 15 [ SPRING . 10 e 5 /, v 20 15 ] SUMMER 10 l 5 0 //A 20 15 FALL 10 =] 7 5 _,!{Z}.l ’{; ’ 7 // s 47/?: - * 20 15 ANNUAL 10 5 7 7% nr . }7,/ o ‘ -2.4 -7 -0 —0.3+0.4 +14 +1.8 +2.5 TEMPERATURE GRADIENT PER 100 ft (°F) Seasonsal Temperature Gradient Frequency. 149 Table 4.6. X-10 Average Monthly Precipitation Data Morth Precipitation (in.) January 5.24 February 5.39 March 5.4% April 4. 14 May 3.48 June 3.38 July 5.31 August 4.02 September 3.59 October 2.82 November 3.49 December 5.22 Light snow usually occurs in all months from November through March, but the total monthly snowfall is often only a trace. The total snowfall for some winters is less than 1 in. The averagge snowfall for the period from 1948 to 1961 is 6.9 in. The maximum snowfall in a 24-hr period was 12 in. in March 1960. The maximum monthly snowfall, 21 in., also occurred in March 1960. The heavy fogs that occasionally occur are almost always in the early morning and are of short duration. 4.3.4 Wind? The valleys in the viecinity of the MSRE site are oriented northeast- southwest, and considerable channeling of the winds in the valley may be expected, This is evident in Fig. 4.5, which shows the annual frequency distribution of winds in the vicinity of CRNL. The flags on the wind-rose diagrams point in the direction from which the wind comes. The prevailing wind directions are upvalley from southwest and west-southwest approxi- mately 40% of the time, with a secondary maximum of downvalley windsg from 3W. B. Cottrell, ed., "Aircraft Reactor Experiment Hazards Summary Report,' USARC Report ORNL-1407, Oask Ridge National Laboratory, November 1952. 150 Feet < 5 icinity the V in lon of Winds istribut Annual Freaquency D D, J, of X-10 Aresa. Fig 151 northeast and east-northeast 30% of the time. The prevailling wind regimes reflect the orientation of the broad valley between the Cumberland Plateau and the Smoky Mountains, as well as the orientation of the local ridges and valleys. The gradient wind in this latitude is usually southwest or westerly, so the daytime winds tend to reflect a mixing down of the gradi- ent winds. The night winds represent drainage of cold air down the local slopes and the broader Tennessee Valley. The combination of these two effects, as well as the daily changes in the pressure patterns over this area, gives the elongated shape of the typical wind roses. During inversions, the northeast and east-northeast winds occur most frequently, usually at the expense of the southwest and west-southwest winds. The predominance of light northeast and east-northeast winds under stable conditions is particularly marked in the summer and fall when the lower wind speeds aloft and the smaller amount of cloudiness allow the nocturnal drainage patterns to develop. Wind roses prepared from 5 years of data, 1956 to 1960, are shown in Figs. 4.6 and 4.7 (ref. 2). These represent the wind direction, fre- quency, and percentage of calm under inversion and lapse conditions for the X-10 area and are applicable to the MSRE location. Considerable varistion is observed both in wind speed and direction within small distances in Bethel Valley and Melton Valley. It may be sald that in nighttime or in sftable conditions, the winds tend to be gen- erally northeast and east-northeast and rather light in the valley, re- gardless of the gradient wind, except that strong winds aloft will con- trol the velocity and direction of the valley winds, reversing them or producing calms when opposing the local drainage. In the daytime, the surface winds tend to follow the winds aloft, with increasing reliability, a8 the upper wind speed increases. Only with strong winds aloft or winds paraliel to the valleys would it be of value to attempt to extrapolate alr movements for any number of miles by using valley winds. In the well- developed stable situation, however, a very light air movement will fol- low the valley as far downstream as the valley retains its structure, “even though the prevailing winds a few hundred feet above the ground are in an entirely different direction. In a valley location, the wind is 152 Uneclagsified ORNL-DWG 63-2943 WINTER SPRING (8198 obs) LAPSE SUMMER FALL Ve ( 60il obs) wind speed, mph % Calm -4 5-9 10-14 15-19 > 20 % Frequency Q 3 10 15 20 1956 -1960 Data Fig. 4.6. X-10 Area Seasonal Wind Roses. 153 Unclassified CRNL-DWG 63-2944 WINTER SPRING (4515 obs) (3959 obs) INVERSION SUMMER FALL { 4792 obs) /)/' (4434 obs) Wind speed, mph % Calm — @.4 5-9 10-14 15-19 2 20 M % Freguency Q D 10 15 20 M [ [ ] 1956 -1960 Data Fig. 4.7. X-10 Area Seasonal Wind Roses. 154 governed by the local valley wind regime and the degree of coupling with the upper winds.? The wind flow between Melton Valley and the X-10 area was investi- gated in conjunction with the Aircraft Reactor Experiment (ARE) hazards 3 Two patterns of wind flow were assumed to be of significance: analysis. (1) from the 7500 area northwest over Haw Ridge to the X-10 area, and (2) from the 7500 area west to White Oak Creek, then northwest through Haw Gap, and finally north to the X-10 area. The frequencies of these + patterns during the period September to December 1950 were normalized to the 1944-t0-1951 wind record at the X-10 area by a ratio method.? The ) normalized frequencies are listed in Table 4.7. Table 4.7. Frequency of Wind Patterns Between 7500 and X-10 Area Frequency of Wind Pattern (%) Over Ridge Through Gap A1l observations 2.5 0.4 Day (9 am to 5 pm) 4.3 0.6 Night (9 pm to 5 anm) 0.0 0.4 Light wind (1 tc 4 mph) 2.6 0.4 Stronger wind (5 mph and over) 2.7 0.3 A comparison of the pibal* cbservations made throughout 1949 to 1950 at Knoxville and Oak Ridge shows that above about 2000 ft the wind roses are almost identical at these twe stations. This identity in the data makes possible the use of the longer period of record (1927 to 1950) from Knoxville to eliminate the sbnormzalities introduced by the use of the short record at QOsk Ridge. Annual wind roses are shown for Knoxville (1927 to 1950) and Nashville (1937 to 1950) in Fig. 4.8. Pibal observa- tions are only made when no clouds, dense fog, or precipitation is occur- ring and so are not truly representative of the upper wind at all times. *¥Pilot balloon visual observations. - 155 Three years of Rawin* data for Nashville (1947 to 1950) are available that consist of observations taken regardless of the current weather at the time of observation. A comparison of these wind roses for Knoxville and Nashville shows that the mode for winds above 3000 meters mean sea level should be shifted to westerly instead of west-northwest when obser- vations with rain are included in the set. In summer and fall, the winds aloft are lightest and the highest winds occur during the winter months at all levels. The northeast-southwest axis of the valley between the Cumberland Plateau and the Smoky Mountains continues to influence the wind distri- bution over the Tennessee Valley up to about 5000 ft, although the varia- tions within the Valley do not extend above about 2000 ft. Above 500 ft, the southwesterly mode gives way to the prevailing westerlies usually ob- served at these latitudes. Previous investigation of the relation of wind directiocn to precipi- tation indicates that the directicon distribution is very little different 3 This is consistent with the experience from that of normal observations. of precipitation forecasters that there is little correlation between sur- face wind direction and rain, particularly in rugged terrain. Figure 4.8 shows the upper winds measured at Nashville during the period 1948 to 1950, when precipitation was occurring at observatbtion time. In general, the prevailing wind at any given level is shifted to the southwest or south-gsouthwest from west or southwest, and the velocity is somewhat higher during the occurrence of precipitation, with the shift being most marked in the winter. The percentage of inversion conditions of the total hours for the seasons is given in Table 4.8 (ref. 2). Tornadoes rarely occur in the valley between the Cumberland and the Great Smokieg, It is highly improbable that winds greater than 100 mph would ever be approached at the MSRE site. ¥Radio wind balloon observations. 156 UNCLASS|FIED DWG. 16972 KNOXVILLE PIBAL NASHVILLE PIBAL NASHVILLE RAWIN NASHVILLE RAWIN ALL OBS. PREGIPITATION 500m 25659 Obs. 20321 Obs 3341 Obs. I500m Q::\‘?‘ ~ .:33:;;. Ao 23449 Obs. {7940 Obs 3269 Obs 227 Obs. 3000m 14710 Obs 12276 Obs. 3193 Obs. 217 Obs. 6000m 3692 Obs 2964 Obs. 2684 Obs. 170 Obs. 10000m 536 Obs. TI7 Obs 1857 Obs & % Calm Wind Speed M PH % Frequency i-ig_1!-33 34-56 57-85 86-1i4 {15+ 10 0 10 20 30 - — I N PIBAL —PILOT BALLOON VISUAL OBSERVATIONS RAWIN—RAD!O WIND BALLOON OBSERVATIONS Fig. 4.8. Wind Roses at Knoxville and Nashville for Various Alti- tudes. 157 Table 4.8. Summary of Seasonal Frequency of Inversions Frequency of Inversion Average Duration@ Season (%) (hr) Winter 3L..8 8 Spring 35.1 9 Summer 35.1 9 Fall 42 .5 10 Annual 35.9 . M. Culkowski, AEC, Oak Ridge, personal communi- cation to T. H. Row, Oak Ridge National Laboratory, July 8, 1963. 4,3.5 Atmospheric Diffusion Characteristics Sutton's*s® methods of calculating the dispersion of airborne wastes require a knowledge of several meteorological parameters. Values of these parameters of the X-10 site were recommended by the Oak Ridge Office of the U,S. Weather Bureau® and are listed in Table 4.9. They provide & “0. G. Sutton, "A Theory of Eddy Diffusion in the Atmosphere,' Proc. Roy. Soc. (London), 1932. °0, G. Sutton, Micrometeorology, McGraw-Hill Co., Inc., New York, 1953. ®"pinal Hazards Summary Report, Experimental Gas-Cooled Reactor, " Vol. 1, Book 1, pp. 3-6, USAEC Report ORO-586, Oct. 10, 1962. Table 4.9. Meteorological Parameter Values for Atmospheric Dispersion Calculations Parameter Lapse Inversion (weak) Mean wind velocity, p (m/sec) 2.3 1.5 Stability parameter, n 0.23 0,35 Diffusion constant, Cy (m/ ?) 0.3 0.3 Diffusion constant, C_ (m/ 2) 0.3 0.033 158 reasonably conservative basis for the dispersion calculations in Section 8.7.2 and Appendix C; however, it is recognized that extreme meteorologi- cal conditions will occasionally exist and could produce exposures greater than those based on the values in the table. 4,3.6 Environmental Radioactivity Atmospheric contamination by long-lived fission products and fallout occurring in the general environment of the Oak Ridge area is monitored by a number of stations surrounding the area. This system provides data to aid in evaluating local conditions and to assist in determining the spread or dispersal of contamination should a maJjor incident cccur. Data cn the environmental levels of radiocactivity in the Oak Ridge area are given in Tables 4.10, 4.11, 4.12, and 4.13 (ref. 7). These levels of activity are, of course, strongly influenced by the radiocactive wastes discharged by the various facilities at ORNL. The MSRE will normally release only the gaseous wastes xenon and krypton at rates of C.5 and 6.4 curies/day, respectively, and thus should not make a significant contribution to the existing activity levels. 4.3.7 Geology and Hydrology The bedrock beneath the MSRE site is a dark gray calcareous clay shale with a bearing value cf © tons/ftz when unweathered. The overbur- den on this fresh unweathered shale averages 20 ft in thickness and con- sists of a thin vlanket of organic tepsoll, generally less than 1 T in thickness, on top of weathered shale. The latter, if confined, has a bearing value of 3 tons/ftz. The strata of the shale are highly folded and faulted, and the dip, although it averages about 35° toward the south, is dirregular and may vary from horizontal to vertical. The Melton Valley in which the MSRE is located is underlain bty the Conasauga shale of the Middle and Upper Cambrian Age. The more resistant rock layers of the Rome formation, steeply inclined toward the southwest, are responsible for Haw Ridge, which parallels the valley immediately to the northwest. 73. ¢. Hart (ed.), "Applied Health Physics Annual Report for 1962," USAEC Report ORNL-3490, Oak Ridge National Laboratory, September 1963. 159 ! Teble 4.10. Concentration of Radioactive Materials in Air in 19628 / fl N Nunber of Particles (in dis- Number of : o Long-Lived integrations per 24 hr) Total - s | ;fii"n Location Activity by Activity RangesP Number of Iarvicles er . Per , -~ (pefec) . Particles . ;,0q pi3 ‘ : L <10° 10°-10% 10%5-107 >107 a. '%aboratory7Area ' , x 10713 - | . \ HP-1 S 3587 38 128 1.6 0.00 0.00 129 3.1 - HP-2 NE 3025 , 43 122 1.9 0.04 0.00 . 124 3.5 HP-3 SW 1000 - 37 129 2.1 0.10 0.02 131 2.1 HP-4 W Settling Basin 21 91 1.2 0.04 0.00 93 1.6 HP-5 E 2506 51 115 1.2 0.04 0.04 117 3.9 HP-6 SW 3027 . . _ 33 136 1.5 0.02 0.02 137 2.4 HP-7 W 7001 ‘ 40 115 1.8 0.00 0.00 117 2.3 HP-8 .Rock Quarry , 39 132 1.5 0.00 0.02 . 133 2.5 HP-9 N Bethel Valley Rd. 31 145 1.6 0.00 0.00 146 2.3 - HP-10 W 2075 38 126 1.3 0.00 0.00 128 3.1 Average 37 124 1.6 0.02 0.01 125 2.7 ; - Perimeter Ares HP-31 Kerr Hollow Gate 3% 135 1.6 0.04 0.04 137 2.7 HP-32 Midway Gate 37 132 2.1 0.02 0.00 134 2.6 HP-33 Gallaher Gate 32 113 1.4 0.00 0.02 114 2.2 ‘ AP-34 White Wing Gate 3 153 )‘ 1.5 0.00 0.00 155 3.0 : HP-35 Blair Gate 39 168 1.6 0.00 0.02 169 3.3 f HP-36 Turnpike Gate 39. 158 2.2 0.02 0.04 161 3.2 I HP-37 Hickory Creek Bend T 34 © 114 1.6 0.02 0.00 115 2.3 ; Average - 36 139 1.7 0.01 0.02 141 2.8 | | ' Remote Area HP-51 Norris Dam . L A3 139 2.3 0.04 ~ 0.00 141 2.6 HP-52 Loudown Dam - .42 0 130 2.8 0,10 ~ 0.00 133 2.4 HP-53 ° Douglas Dam - .. 44 " " 150 2.6 0.02- 0.00 - 153 2.8 "HP-54 . Cherokee Dam L A0 164 2.4 0.04 - 0,02 © 167 3.0 HP-55 Watts Bar Dam =~ =~ 45 - 157 2.0 0.04 0.00° = 159 2.9 . “HP-56 - Great Falls Dam 46 71667 2.3 0.00 0.00 = 1p8 3.1 ' HP-57 Dale Hollow Dam ' . . 38 .7 171 "~ 1.6 0.00 0.04 - 172 2.9 L Average - 43 . 154 2.3 0.03 0.0L 157 2.8 K 8pveraged weekly from filtérffiéperfidata; , - ) Ppetermined by filtration techniques, .- . ' 5 « v e . - 1 — - e o e e e e B i 160 Teble 4.,11. Radioparticulate Fallout in 19628 s “Number of Particles (in dis- Statl Long-Lived integrations per 24 hr) | . Total Total ,Nu:b on Location Activity by Activity Ranges . Number of Particles er ' (pe/1t2) — Particles per ft2, s - <10® - 10°-10% - 10%-107 >107 Laboratory Area x 10-12 | ) , HP-1 S 3587 ' 15 79 2.1 0.12 0.06 8L 42 HP-2 NE 3025 : 17 88 2.3 0.04 0.06 91 .49 HP-3 SW 1000 15 83 2.0 0.15. 0.06 86 42 HP-4 W Settling Basin 14 73 2.3 0.08 0.04 75 48 HP-5 E 2506 : 14 86 2.0 0.08 0.04 .91 50 HP-6 SW 3027 16 101 2.9 0.02 0.02 104 61 HP-7 W 7001 15 89 2.4 0.02 0.06 92 49 BP-8 Rock Quarry 17 89 2.6 0.00 0.08 9 46 HP-9 N Bethel Valley Rd. 16 88 2.9 0.06 0.12 91 41 HP-10 W 2075 15 100 2.3 0.04 0.00 103 55 ' Average 15 88 2.4 0.06 0.05 91 48 i ‘ Perimeter Ares _ HP-31 Kerr Hollow Gate 17 103 2.13 0.13 0.10 105 &7 HP-33 Gallaher Gate 14 82 2.4 0.10 0.00 85 - 42 HP-34 White Wing Gate 18 104 2.2 0.19 0.08 106 47 HP-35 = Blair Gate 15 124 2.0 0.06 0.04 126 50 HP-36 Turnpike Gate 16 109 3.5 0.08 0.02 112 57 HP-37 Hickory Creek Bend 16 85 2.3 0.04 0.08 87 47 Average , 16 101 2.5 0.10 0.05 103 48 Remote Area . HP-51 Norris Dam 14 86 2.2 0.12 0.04 89 T 36 HP-52 Loudoun Dam 13 70 2.7 . 0.06 0.06 7 29 HP-53 Douglas Dam 13 T 2.7 0.06 0.08 . 80 35 HP-54 Cherokee Dam 14 11 2.9 - 0.13 0.06 84 .35 HP-55 Watts Bar Dam 16 81 2.2 0.14 0.08 83 37 HP-56 Great Falls Dam 14 - 98 2.2 0.06 0.02 100 39 HP-57 Dale Hollow 14 96 2.0 '0.08 0.06 98 33 ' Average ‘14 84 2.4 0.09 0.06 87 - 35 E"m'era.geél' weekly from gummed-paper data. ¢ iy 161 Table 4.12. Concentration of Radioactive Materials in Rain Water in 19622 Activity in Station Location Collected Number Rain Water (uc/cc) Laboratory Ares x 1077 HP-7 West 7001 10.3 Perimeter Ares HP-31 Kerr Hollow Gate il HP-32 Midway Gate 12 HpP-33 Gallaher Gate 10 HP-34 White Wing Gate 11 HP-35 Blair Gate 11 HP-36 Turnpike Gate 10 HP~-3"7 Hickory Creek Bend 11 Average 11 Remote Ares HP-51 Norris Dam 14 HP-52 Loudoun Dam 1L HP-53 Douglas Dam 13 HP-54 Cherokee Dam 11 HP-55 Watts Bar Dam 14 HP-56 Great Falls Dam 16 HP-57 Dale Hollow Dam 11 Average 13 aAveraged weekly by stations. Table 4.13. Radloactive Content of Clinch River in 1962 ‘ . . . Average Concentration of N%clldes of Primary Concern Concentration Radiocactivity _ ne/cc) (MPC )., Location of Total a as Percentage : - (ne/ce) qr20 geldd (0gl37 Rul03-106 (60 .95 yn 95 Radiocactivity of (MPC),, (nc/ce) x 10-8 x 1078 x 108 x 107®% x 10-8 x 1078 x 1078 x 1070 Mile 41.5 0.16 0.14 0.02 0.78 (o) 0.42 1.5 0.90 1.7 Mile 20.8° 0.15 0.02 0.09 21 0.18 0.09 34 4.6 7.4 Mile 4.54 0.34 0.20 0.07 16 0.32 0. 54 17 3.5 4.9 “Weighted average maximum permissible concentration (MPC )y calculated for the mixture by using values for specific radionuclides recommended in NBS Handbook 69. ’ bNone detected. cValues given for this location are calculated values based on the levels of waste released and the dilution afforded by the river. Center's Ferry (near Kingston, Tennessee, just above entry of the Fmory River into Clinch River). c9T 163 These layers dip beneath the shales of the Conasauga group in Melton Valley. The shale layers in the area are in keeping with the general structure of the surrounding area, as reported in a geological survey of the area.® Conasauga shale, a dark red shale, contains thin layers and lenses of limestone that are generally irregular in distribution. How- ever, there are no persistent limestone beds in the upper strata of the shale layers and, consequently, no underground solution channel or cavern to permit rapid and free underground discharge of water. The dominantly clay soils of the Oak Ridge area are generally of low permeability, so the surface runoff of water is rapld. Cbservations in test wells show that the Conasauga shale, although relatively impermeable, is capable of transmitting small amounts of water through the soll a dis- tance of a few feet per week. Furthermore, all the active isotopes, ex- cept ruthenium, apparently become fixed in the immediate vieinity of the point of entry into the soil. It is concluded that such ground water flow as may exist below the surface in the soil surrounding the site will be small and slow (a few feet per week) and that natural chemical fixation will reduce the level of the activity of mixed, nonvolatile fission prod- ucts more than 90%. The Conasauga shale Tormation is extensive, quite heterogeneous in structure, and relatively low in permea’bility.8 The depth and extensive- ness of the formation provide a large capacity for decontamination of any contaminated liguid reaching it, and the slow rate of percolation improves the efficiency of the decontamination by ion exchange and radioactive de- cay. However, the heterogeneity of the formation makes difficult the prediction of pattern and rate of movement of water and hence of the fis- sion products contained therein. Most of the seepage is along bedding plains parallel to the strike. Storm drains for the MSRE site will discharge into Melton Branch. The area storm sewers consist of two individual trunk systems. One trunk 8p, B. Stockdale, "Geologic Conditions at the Ozk Ridge National Taboratory (X-10 Area) Relevant to the Disposal of Radiocactive Waste,” USAEC Report ORO-58, Oak Ridge Operations, August 1951. 164 system empties directly into Melton Branch (see Fig. 4.5) through a con- necting ditch; the second trunk system combines with the cooling tower blowdovn and the flow from the process waste pond before discharging into Melton Branch through a connecting ditch. The storm drainage system is based on a runoff coefficient of 30% for unpaved or seeded areas. The high runoff coefficient applied to the unpaved areas is justified by the highly impervious character of the solls present in the plant area. Downstream from ORNL, numerous uses are made of the Clinch River water by both municipalities and industries, as indicated in Table 4.14. The first downriver water consumer is the K-25 plant, whose water intake is located at river mile 14.4. Representative flows of the Clinch, Hmory, and Tennessee Rivers are listed in Table 4.15. The local flow pattern in the Watts Bar reservoir during the period May through September is sig- nificantly affected by the differences in water temperature of the Clinch River and the Watts Bar reservoir. When the Clinch River is considerably cooler, stratified flow conditions caused by density differences may exist, with the cooler water flowing on the bottom beneath the warmer water. This phenomenon markedly affects the travel time of water through the reservoir and complicates the analysis of flow. 1In addition, during the period of stratified flow, some Clinch River water may flow up the Emory River as far as the Harriman water plant intake. 4.3.8 Selsmology Information cn the frequency and severity of earthquakes in the East Tennessee area is reported in the ART Hazards Report.9 Barthquake forces have not been considered in the design of facilities at CRNL or TVA struc- tures in the Fast Tennessee areca. The Oak Ridge area is currently classi- fied by the U.S. Coast and Geodetic Survey as subject to earthquakes of intensity mm-6 measured on the modified Mercalli Intensity Scale. W. B. Cottrell et al., "Aircraf: Reactor Test Hazards Summary Re- port," USAEC Report ORNL-1835, Oak Ridge National Laboratory, January 1955. Table 4.14. Community Water Systems in Tennessee Downstream from ORNL Supplied by Intakes on the Clinch and Tennessee Rivers or Tributaries® Commmity Population Intake Source Approx?mate Remarks Stream Location ORGDP — K-25 Ares 2,678P Clinch River CR Mile 14 Industrial plant water system Harriman 5,931° Fmory River FR Mile 12 Mouth of Emory River, CR Mile 4.4 Kingston Steam Plant 5004 Clinch River CR Mile 4.4 (TVA) Kingston 2,0004 Tennessee River TR Mile 570 River used for supplementary sup- ply Watts Bar Dam (resort l,OOOd Tennessee River TR Mile 530 village and TVA steam plant ) Dayton 3,500° Richland Creek RC Mile 3 Opposite TR Mile 505 Cleveland 16,196° Hiwassee River HR Mile 15 Mouth of Hiwassee River is at TR Mile 500 Soddy 2,000d Tennessee River TR Mile 488 Chattanooga 130,009¢ Tennessee River TR Mile 465 Metropolitan area served by City Water Company South Pittsburg 4,130¢ Tennessee River TR Mile 435 Total 168,061 “HGCR Hazards Summary Report, USAEC Report ORO-586, Oak Ridge Opera- tions, Oct. 10, 1962. bBased on May 1963 data. ©1960 Report of U.S. Bureau of the Census. dBased on published 1957 estimates. GOt Table 4.15. Flows in Clinch, Emory, and Tennessee Rivers, 1945-1951% _ Flow (cfs) Clinch River Emory River, . Tennessee River Miles 20.8 and 13.2 Mile 4.4D Mile 12.8 Mile 529.9 Mile 465.3 Maximum Mean Minimum Msximum Mean Minimm Maximum Mean Minimm Maximum Mean Minimm Maximum Mean Minim January 22,900 8,960 1620 70,700 14,400 2120 50,000 4450 178 181,000 47,700 15,800 218,000 63,900 23,200 February 27,700 10,100 1230 88,800 15,800 2250 69,000 4550 468 204,000 50,800 17,900 195,000 67,900 19,700 March 12,700 5,850 690 26,700 9,450 1830 15,400 2910 507 100,000 32,400 13,300 148,000 44,200 19,500 April 8,540 3,400 306 13,300 5,620 752 6,600 1800 249 43,700 23,800 5,200 82,000 27,700 13,200 ~ 6,600° 9, 730¢ 25200 34,900° .. " 45,900¢ Mey . 8,080 2,750 208 19,700 4,700 520 13,100 1610 58 38,300 20,000 3,000 95,900 28,800 17,900 June 7,420 2,820 224, 9,280 3,320 262 5,300 396 14 32,100 18,900 8,200 32,800 25,500 18,900 July 7,630 2,930 259 12,800 3,400 281 9,230 360 21 87,500 19,600 6,500 106,000 26,400 15,400 §_ August 8,390 4,520 374 8,760 4,800 378 3,060 177 4 37,600 21,400 9,900 43,300 28,100 17,800 ‘ September 8,450 4,620 341 13,000 4,940 462 5,500 224 2 39,900 22,200, 6,900 54,400 30,100 17,100 | 3, 5304 4,2304 5534 20,4008 | 27,8004 October 9,200 5,130 150 14,200 5,300 150 5,040 93 1 67,100 23,500 9,800 72,800 29,700 16,300 November 12,700 4,430 453 40,500 5,830 556 27,800 1100 2 128,000 25,400 10,300 167,000 34,000 13,600 December 27,000 8,360 569 - 60,300 11,700 593 33,300 2720 24 112,000 41,000 11,800 139,000 53,600 21,100 ®EGCR Hazards Surmary Report, USAEC Report ORO-586, Oak Ridge Operations Office, Oct. 10, 1962. TFilow data were collected prior to construction of Melton Hill Dem on Clinch River. Although current data are not available, flows are not expected to be significantly changed. bFlows shown for Clinch River Mile 4.4 include Emory River flows. “Nonstratified flow period. Astratiried flow period. ®By egreement with TVA, a flow of not less than 150 cfs has been main- tained in the Clinch River at Oask Ridge since August 28, 1943, ' - 167 Both LynchlO of the Fordham University Physics Department and Moneymakerll of the Tennessee Valley Authority indicate that the earth- quakes that occasionally coccur in the East Tennessee area are dquite common in the rest of the world and are not indicative of undue seismic activity. An average of one or two earthquakes a year occurs in the Appalachlan Valley from Chattancoga to Virginia according to TVA records. The maxi- mum intensity of any of these shocks recorded is 6 on the Woods-Neuman Scale. A gquake of this magnitude was experienced in the Oak Ridge area on September 7, 1956, and was barely noticeable by either ambulatory or stationary individuals. Structures were completely unaffected. Distur- bances of this type are to be expected only once every few years in the Oak Ridge ares. The Fordham University records indicate a quake frequency below that estimated by TVA. However, the magnitude of the observed quakes is ap- proximately the same. Lynch indicates that "it is highly improbable that a major shock will be felt in the ares (Tennessee) for several thou- sand years to come." L0Letter from J. Lynch to M. Mann, November 3, 1948, quoted in a re- port on "The Safety Aspects of the Homogeneous Reactor Experiment,"” USARC Report ORNL-731, Oak Ridge National Laboratory, June 1950. 11B, C. Moneymaker, Tennessee Valley Authority, personal communica- tion to W. B. Cottrell, Oak Ridge National Laborstory, October 1952. 168 5. CONSTRUCTION, STARTUP, AND COPERATION 5.1 Cecnstruction Although no special constructicn practices were employed in assem- bling the MSRE, meny special precauticns were taken to ensure a high- quality, clean, and leak-tight assembly. A detailed specification re- quiring better cuality contrcl than that of existing commercial cocdes was prepared for each compenent of the system. Nondestructive inspection techniques, such as ultrascnic testing, dye-penetrant inspection, X-ray examination, and helium leak testing, were employed at each fabricator's plant under the supervieion of ORNL inspectors. Assembly at the reactor site is being examined similarly. After completicn, the separate systems will be leak tested using rate-cf-pressure rise, mass-spectrometer, and isotcpic~-tracer technicues. During the last stages cf constructicn and for several weeks after censtructicn 1s completed, the reactor ecuipment will be checked cut for acceptance by the Operations Department. The individual systems, that is, the cover gzs, offgas, pump cil, cooling water, component air-cocling systems, etc., will be cperated fcr the first time. Remote mzintenance will be practiced during this period also. The successive steps in the startup of the reactor are listed in Table 5.1. The most important phases are discussed below. 5.2 Flush-8al1lt Operation It is planned to demonstrate the mechanical performance of the as- sembled reactor by & several-month period of testing with & flush salt in the fuel system. Simultanecusly, ccclant salt will be circulated in the ccolant system. Each piece cf equipment will be examined tc determine whether 1t perfcrms as designed, insofar as this can be determined with- out nuclear heat generation. The flush salt will zlso serve to scavenge oxygen and to remove other impurities. Another important benefit of the flush-salt test will be the development of the operating skills necessary Tor satisfactory contrcl of the system variables. Heat-balance methods 169 Table 5.1. MSRE Startup Plans Startup Phase Month Complete Reactor Installation 0 Operator and Supervisor Training 1-6 Equipment Checkout — Without Salt 1-3 1. Heaters, thermocouples, instruments, and logger 2. Cover-gas, offgas, pump-oil, cooling-water, component-air- cooling, and radiator-cooling systems 3. Air-compressor, containment-ventilation, diesel-power, and electrical systems 4. Remote maintenance Flush Salt Operation 3=5 1. Coolant and flush-salt loading, inventory, and transfer methods 2. Coolant-circulating system 3. Fuel-circulating system 4. Bampling and leakage detection 5. Heat-balance methods, heat-removal systems, and temperature control 6. Graphite examination Precritical Shutdown o= 1. Chemical plant performance (hydrofluorination of flush salt) 2. Loading of fuel salt 3. Completion of nuclear instrumentation and control-rod tests 4. Final maintenance Critical Experiments 69 1., Check performance of nuclear instrumentation 2. Load fuel to critical 3. Measure pressure and temperature coefficients of reactivity, control-rod worth, response to flow changes 4. Establish baselines for determination of effects of power on chemistry, corrosion, and nuclear performance Posteritical Shutdown 10 1. Seal-weld membranes on reactor and drain tank cells 2. Final check on contaimment leakage 3. Fill and check out vapor-condensing system Approach to Full Power 11-14 1. Heat balances, heat-removal control 2. Shield and containment surveys 3. Sampling program for chemistry, etc. 4. Measurement of power coefficient, xenon polsoning, fuel 5 permeation of graphite, and offgas composition Reactor control and logger performance 170 and temperature control will be practiced, and sampling operaticns will be tried. After the flush-salt cperating periocd of one to three months, the flush salt will be transferred tc the chemical plant, where it will be treated tc remove accumulated cxides in a trial performance of the chemi- cal plant. While the reactor system is shut down, the fuel salt will be lcaded into cne cf the drain tanks, with about two-thirds of the U?35 es- timated for criticality, and the multiplication of the fuel will be meas- ured at several liquid levels. Alsc during this shutdown, nuclear in- strumentaticn and contrcl reds will be given a final checkout, and remote- maintenance practice will be completed. 5.2.1 Critical Experiments After the reactor system has been heated to 1200°F and neutron count- ing rates have been measured with the reactor empty, the fuel mixture, with about two-thirds of the final U??° concentration, will be slowly transferred to the reactcr. Ccunting rates will be determined with the control rods in varicus positions and a2t several temperatures. If criti- cality is not attained, the fuel salt will be drained and more U2?3° will be added and mixed with the original salt before refilling the core, This procedure will be repeated after every fuel addition. When the counting rates indicate that criticality 1s near, further fuel additions will be made in approximately 100-g increments at the fuel-circulating pump through the sampling mechanisn. When the critical mass has been determined, the temperature and pres- sure ccefficients will be measured, and the ccntrol-rod calibrations will be completed. Next, the effective fraction of delayed neutrons will be studied to learn the fraction lost by circulation. The reactor must be cperated at zerc power for a pericd of one to three months to establish baselines for determinaticn of the effects ¢f power operation on salt chemistry, corrcsion, and nuclear performance. 5.2.2 Power Operation When the baseline data at zero power are judged to be satisfactory, the power will be increased in increments of several hundred kilowatts 171 over a period of six to ten weeks. At each successively higher power, samples will be collected and measurements will be made of heat output, power coefficient, radiation levels inside and outside shielding, xenon poisoning, offgas composition, and possible permeation ¢f the graphite by Tuel. 5.3 Operations Perscnnel The operation of the MSRE facility will be the responsibility of the Reactor Division Operaticns Department. The previcus assignments of per- sonnel in this group Include the ccnstruction, startup, and operation of four other experimental reactors: the Low-Intensity Test Reactor, the Homogeneous Reactor Experiments 1 and 2, and the Alrcrarft Reactor Experi- ment. The experiment will be conducted con a three-shift basis, emplcying four coperating shifts and a day staff for reactor analysis and planning. The organization chart 1s shown in Fig. 5.1. EFach of the four shifts will be headed by a senior-level supervisor for the first few months. A Junior engineer and three or four nontechnical operators, many with the previously mentioned reactor experience, will complete the shift organization. The Reactor Analysis Group will be compesed of four to six engineers with broad experience in fluid-fuel reactors. Its function will be prin- cipally to plan, supervise, and analyze the experimental progran. Over the years this organization has developed training, operating,l emergency,2 and maintenance practices3 that especially contribute to ex- rerimental reactor safety. The same methods and policies will be applied to the Molten-Salt Reactor Experiment. 1R. H. Guymon, "MSRE Design and Operations Report, Part VIII, Operat- ing Procedures,' USAEC Report ORNL-TM-908, Oak Ridge National Laboratory, to be issued. °R. H. Guymon, "MSRE Design and Operations Report, Part IX, Safety Procedures and Emergency Plans," USAEC Report ORNL-TM-909, Osk Ridge Na- tional Laboratory, to be issued. 3B, (. Hise and R. Blumberg, "MSRE Design and Operations Report, Part X, Maintenance Fquipment and Procedures,' USAEC Report ORNL-TM-910, Oak Ridge National Laboratory, to be issued. DEPARTMENT HEAD SECRETARY CLERK UNCLASSIFIED ORNL DWG. 64-5660 OPERATIONS CHIEF ASSISTANT CHIEF TECHNICIANS (2) L A-SHIFT SUPERVISCR ASSISTANT SUPERVISOR TECHNICIANS (3) B- SHIFT SUPERVISOR ASSISTANT SUPERVISOR TECHNICIANS (3) C-SHIFT SUPERVISOR ASSISTANT SUPERVISOR TECHNICIANS (3) D-SHIFT SUPERVISOR ASSISTANT SUPERVISOR TECHNICIANS (3) Fig. 5.1, ANALYSIS CHIEF MAINTENANCE AND DESIGN CHIEF -————I COMPUTER PROGRAMMER NUCLEAR ENGINEERS (4) TECHNICIAN CHEMIST TECHNICIAN CHEMICAL ENGINEER TECHNICIAN METALLURGIST INSTRUMENT ENGINEER MAINTENANCE ENGINEER DESIGN ENGINEER PROCUREMENT AND PLANNING ENGINEER TECHNICIANS (2) Reactor Division, Operations Department Organization for the MSRE. LT 173 In addition to the capabilities of the Operations Department person- nel, the Reactor Division can provide agsistance as needed from its larger Design, Development, Analysis, Engineering Science, and Irradiation Engi- neering Departments. Also, other ORNL research and service divisions are participating in many aspects of the MSRE program. The engineers and technicians assigned to the MSRE operating group are prepared for their duties as reactor supervisors and operators by a combination of formal classroom work and inservice training. The first undertaking is a two-week period devoted exclusively to lectures and in- struction on reactor physics, chemistry of salts, MSRE design, nuclear instruments, etc. Next, the candidates are assigned to work on circulating-salt loops, a research reactor, or engineering developments, depending on the candi- dates' previous experience. The next agsignment is at the MSRE to assist in calibrations, checkout, and startup of equipment. Following these tasks the personnel will be organized into shift teams to begin operation with molten salt in the reactor systems. At the end of the several months period of circulating hot salt, both the reactor and the operators should be ready for the uranium loading and the initial critical experiment. However, during the maintenance period (approximately one month) prior to criticality, all the operating teams will be given a refresher series of lectures and an examination to conclude the eight-month training program. 5.4 Msintenance The maintenance practices to be employed on the MSRE are the outgrowth of several years of experience in repairing and replacing parts of the egrlier fluid-fuel reactors, the ARE, HRE-1, and HRE-2. Basically the system of maintenance makes use of long-handled tools manipulated by hand through special shielding. ATter a fault has been discovered, the reactor will be shut down, drained of salt, and coocled. Plans for the repair will be made and previously prepared procedures medified, 1f necessary, to cover not only the repair details but safety precauticns also. When the system 1s cocl, 174 a hole will be cut in the top sealing membranes at a location directly above the faulty equipment. The maintenance shield will be set in place and manipulated to permit remcval of the shielding blocks below the mem- brane. Next, the faulty part will be disconnected, using tools through the special shield. If the faulty part is highly radicactive and must be removed from the reactor cell, the operators will retire to the shielded maintenance control room from which the building crane can be operated remotely tc remove the part and store it in a safe place. The operators can then return to the area above the cell and install the new part, using the special shield. Viewing windows and remote television are available to assist the operatcors in all these procedures. PART 2. SAFETY ANALYSES 177 6. CONTAINMENT 6.1 General Design Considerations It is required that the containment be adequate to prevent escape of large amounts of radiocactivity to the surrcunding area during opera- tion and maintenance and in the event of any'credible accident. The containment must also prevent the release of.dangerous amounts of other hazardous materials and, in general, serve to protect personnel and ex- ternal equipment from damage. Any equipment that contains or cculd contain multicurie amounts of redicactive material must be surrounded by a minimum of two barriers to prevent escape of the radiocactivity. During coperation of the MSRE, the first barrier, the primary container, consists of the walls of the vari- cug components in the system and the connecting piping. The reactor and drain tank cell enclosures provide the secondary containment. These cells are normally operated at subatmospheric pressure to assure inleak- age. The controlled ventilation areas in the high bay and in the varicus cells constitute a third barrier tc the escape of activity during normal operaticn of the reactor. Alir is drawn from the enclosed areas that are at subatmospheric pressure, passed through absclute filters, and moni- tored for radicactivity before discharge from the 100-ft-high ventila- tion-system stack located south of Building 7503. When the reactor or drain tank cells are opened for maintenance, air flows through the openings at velocities in excess of 100 fpm and sub- stitutes as the secondary barrier, The controlled ventilation area of the high bay above the cell is still the third barrier. However, if the primary piping in the cell 1s opened for maintenance, the air flow through the cell opening becomes the primary barrier and the contrclled ventila- tion area becomes the secondary containment. In the hypothesis of the maximum credible accident (see sec. &.35), in which it is assumed that hot fuel salt mixes with the water in the cells to generate steam, the total pressure in the containment vessels 178 cculd exceed the design value of 40 psig, i1f not controlled. A vapor- condensing system is therefore provided that can rapidly condense the steam and also retain the ncnccondensable gases. During normel operation, the atmosphere in the reactor and drailn tank cells is maintained as an inert mixture (95% N, 5% O,) to eliminate hazards from combustion of flammable materisls in the cell, such as the cil in the lubrication system of the fuel circulating pump. The overall leak rate of the entire primary system is expected to be less than 1 cm” of salt per day under normal cperating conditions. The gystem is ¢f all-welded ccnstruction, and all flanged Jjoints are monitored for leakage; a few Jjoints at the least vulnerable lccations have autoclave- type fittings. Pipe lines that pass thrcough the cell walls (i.e., the secondary containment) hzve check valves or air-operated block valves or both which are controlled by radiation monitors or pressure switches that can sense a rise in cell pressure. The portion of thig piping outside the cell, the portion between the cell wall penetration and the check or blcck valve, and the valves are enclosed to provide the required second- ary containment. These auxiliary enclosures are designed to withstand the same moaximum pressure (40 psig) =s the reactor and drain tank cells. 6.1.1 Resctecr Cell Design [ The reacter cell, shown in Fig. 6.1, is a cylindrical carbcn steel vessel 24 Tt in diemeter and 23 £t in overall height that has a hemi- spherical bottom and a flzt top. The lower 24 1/2 £t was built for the ART in 1956. It was originally designed according to Section VIIT of the ASME Boiler and Pressure Vessel Code for 195 psig at 565°F and was tested hydrostatically at 2300 psig.l The hemispherical bottom is 1 to 1 1/4 in. thick. The cylindrical portion is 2 in. thick, except for the section that contains the large penetrations, which is 4 in. thick. This vessel was modified for the MSRE in 1962 by lengthening the cylindrical section & 1/2 ft. Severzl new penetrations were installed, 1W. F. Ferguson et al., 'Termination Report for Cernstruction of the ART Facility, USAEC Repcrt ORNL-2465, Oak Ridge Naticnal Laboratory, Nov. 21, 1958, SHIELDING ANNULUS ) o TYPICAL FREEZE FLANGE TYPICAL PIPE HEATER 'SECONDARY CONTAINER WALL Fig. 6.1. Reactor Cell Model. . ‘l w§ UNCLASSIFIED PHOTO 38793 SHIELDING ANNULUS 6LT e ot sl bt * 180 and a 12-in. section of 8-in. sched-80 pipe closed by a pipe cap was welded into the bottom of the vessel to form a sump. The extension to the vessel is 2 in. thick, except for the top section, which was made as a 7 1/4- by l4-in. flange for bolting the top shield beams in place. The flange and top shield structure are designed to withstand a pressure of 40 psig, and the completed vessel was tested hydrostatically to 48 psig, measured at the top of the cell. Both the original vessel and the exten- sion are made of ASTM A201, grade B or better, fire-box-quality steel. A1l the welds on the reactor cell vessel were inspected by magnetic particle methods, 1f they were in carbon steel, or by liquid penetrant methods, if they were in stainless steel. All butt welds and penetration welds were radiographed. After all the welding was completed, the vessel was stress relieved by heating to between 1150 and 1200°F for 7 1/2 hr. The top of the cell is constructed of two layers of 3 1/2-ft-thick reinforced-concrete blocks, with a stainless steel membrane between. The top layer is ordinary concrete, with a density of 150 lb/ftB, and the bottom layer is magnetite concrete, with a density of 220 lb/ft3. The blocks in both layers run east and west. To ald in remote maintenance, the bottom layer is divided into three rows of blocks. Blocks in the outer rows are supported on one end by a 13- by 4-in. channel-iron ring welded to the inside of the cell wall. Two 36-in. I beams provide the rest of the support fcr the bottom layer of blocks. These beams have angle-iron and steel-plate stiffeners. The cavities are filled with con- crete for shielding. The beams rest on a built-up support plate assembly, which is welded to the side of the cell at the 847-ft 7-in. elevation. Offsets 6 1/4 by 26 in. are provided in the ends of the bottom blocks to fit over the support beams. Guides formed by angle iron assure proper alignment. Several of the bottom blocks have stepped plugs for access to selected parts of the cell for remote maintenance, The sides of the blocks are recessed 1/2 in. for 14 in. down from the top. With blocks set side by side and with a l/2-in. gap between, a 1 1/2-in. slot is formed at the top. COne-inch-thick steel plates 12 in. high are placed in the slots for shielding. The 1ll-gage, ASTM-A240, type 304 stainless steel membrane 1s placed on top of the bottom layer of blocks 181 and 1is seal welded to the sides of the cell. Cover plates are provided over each access plug. These are bolted to the menbrane and are sealed by necprene O-rings. A l/8—in. layer of Masonite is placed on top of the membrane to protect it from damage by the top layer of blocks. The top blocks are beams that reach from one side of the cell to the other. The ends of these blocks are bolted to the top ring of the cell by fifty 2 1/2-in. studs, 57 1/4 in. long, made from ASTM-A320, grade L7, bolting steel. These studs pass through holes that were formed by casting 3-in. sched-40 pipes in the ends of the blocks. The reactor cell vessel is installed in another cylindrical steel tank that is referred to here as the shield tank. This tank is 30 ft in diameter by 35 1/2 ft high. The flat bottom is 3/4 in. thick, and the cylindrical section is 3/8 in. thick. The shield tank sits on a rein- forced concrete foundation., The reactor vessel cell is centered in the shield tank and is supported by a 15-ft-diam, 5-ft-high cylindrical skirt made of l-in,-thick steel plate reinforced by appropriate rings and stiff- eners. The skirt is Joined to the hemispherical bottom of the reactor cell in a manner that provides for some flexibility and differential ex- pansion, The annulus between the shield tank and the reactor cell vessel and skirt ig filled with magnetite sand and water for shielding. The water contains about 200 ppm of a chromate-type rust inhibitor, Nalco-360. A 4=in.-diam overflow line to the cooclant cell controls the water level in the annulus. The region beneath the reactor cell vessel inside the skirt contains only water, and steam will be produced there if a large quantity of salt is spilled into the bottom of the reactor cell (see sec. 8.6). An 8-in.- diam vent pipe, which has a capacity for three times the estimated steam production rate, extends from the water-filled skirt area under the vessel to the coolant cell, where the steam would be vented. Numerous large sleeves are required through the walls of the reactor cell and shield tank to provide for process and service piping, electrical and instrument leads, and for other access. The 4- to 36-in.-diam pipe gsleeves are welded into the walls of the reactor cell and the shield tank. 182 Since the temperature of the reactor cell will be near 15C°F when the reactor is operating and the temperature of the shield tank may at times be as low as 60°F, bellcows were incorporated in most of the sleeves to permit radial and axial movement of one tank relative to the other with- out producing excessive stress. The bellows are covered to prevent the sand in the shield tank from packing tightly arcund them. Several other lines are installed directly in the penetrations, with welded seals at cne or both ends, or they are grouped in plugs filled with concrete and inserted in the penetrations. The major openings are the 36-in.-diam neutrcn instrument tube and drain tank interconnection and the 30-in.-diam duct for ventilating the cell when maintenance is in progress. The coriginal tank contained several other 8- and 24-in.-diam sleeves, and they were either removed or closed and filled with shielding. The smaller penetrations and methods of sealing them are described in Sec. 6.1.3. ¢.1.2 Drain Tank Cell Design The rectangular drain tank cell, shewn in Fig. 6.2, is 17 It 7 in. by 21 ft 2 1/2 in., with the corners beveled at 45° angles for 2 1/2 ft. The flat floor is at the 8l4-ft elevation. The open pit extends to the 852-t elevation. The cell was designed to withstand a pressure of 40 psig, and when completed in 1962 it was hydrostatically tested at 48 psig measured at the elevation of the menbrane at 838 1/2 ft. The bottom and sides have a 3/16—in.-thick stainless steel liner backed up by heavily reinforced concrete. Magnetite concrete is used where reguired for biological shielding. Vertical cclumns in the north and south walls are welded to hori- zontal beams embedded in the concrete of the cell floor. A 4-in. slot is provided at the top of each horizontal beam so that a steel key may be wedged inte it to hold down the top blocks. The top blocks are arranged in two layers. Both layers of blocks are ordinary concrete (density 150 1b/ft>). The bottom layer is 4 £t thick and the top layer is 3 1/2 ft thick. An ll-gage type 304 stainless | ] s e w " » 183 SECONDARY CONTAINMENT TYPICAL FREEZE § VALVE FUEL SALT DRAIN TANK NO. 1 ig. 6.2. Drain Tank Cell UNCLASSIFIED PHOTO 38768 * TYPICAL SUPPORT .AND WEIGHT CELL ARRANGEMENT FLUSH SALT DRAIN TANK e b At Lt s e 184 steel membrane is placed between the twc layers and seal welded to the cell liner at elevation 838 1/2 ft. 6.1.3 Penetrations and Methods of Sealing Piping and wiring penetrations through the cell walls were designed to eliminate possible sources of gas leakage. All electrical leads passing through the cell walls are in magnesia- filled copper sheaths (Fig. 6.3). The sheaths were sealed to the 3/4-in. pipe penetrations by two compression-type fittings, one inside and one outside the cell. The ends of the sheaths that terminate inside the cells were sealed at the disconnect by glass-to-metal welds. The ends that ter- minate outside were sealed by standard mineral-insulated cable-end seals, such as those manufactured by the General Cable Company. (The seal was formed by compressing a plastic insulating material arocund the wires. ) All thermocouples have Fiberglas-insulated leads in multiconductor sheathed cables. The sheaths were sealed tc the 3/4—in. pipe penetra- tions inside and outside the cell by using soft solder. The ends of the sheaths terminating inside the cells were sealed at the disconnect with glass-to-metal welds. The ends cof the cables ocutside the cells terminate in epoxy-sealed headers. The headers can be pressurized to test for leaks. A1l instrument pneumatic signal lines and instrument air lines are sealed to the 3/4~in. pipe penetrations by two compressiocon type fittings, one inside and one outside the cell. Fach of these lines contains a block valve located near the cell wall; the valves close automatically if the cell pressure becomes greater than atmospheric. Methods of sealing certain lines require special mention, as follows: 1. The 30-in.-diam cell ventilation duct contains two 30-in. motor- operated butterfly valves in series. These valves are under strict ad- ministrative contrcl to assure that they remain closed during reactor operation. 2. The component cooling system blowers are sealed in containment tanks to guard against loss of gas at the shaft seals. 7 i | } | | THREADED PIPE ~ COUPLING ™ | COMPRESSION GLAND, ™ CCMPRESSION NUT A ‘ ! SLEEVE~ Fig. 6.3. QUTER VES3SEL WALL rREACTOR N CONTAINMENT VESSE L WA L L o SAKD-WATER ANNULUSD e - T EXPANSION S BELLOWS FENETRATION PLUG SLEEVE —_ .L” T 25 MIiN OFFSET % NPs (TYPicAaL) M\ CABLE SLEEVE-— CONCRETE ¥ BETWEEN CABLE SLEEVES Typical Electric Lead Penetration of Reactor Cell Wall. \ . 2oz - FPENETRATION SLEEVE FITTINGS TYPICAL [ BECTH ENDS MULTI-CONDUCTOR MINERAL- INSULATELD COPPER-SEHEATHED ELECTRIEC CABLF G8T 186 3. The cell evacuation line contains a block valve, HCV565, which automatically closes in the event radiocactivity is detected in the line by the radiation monitor. 4. The air supply lines for the cell sumps contain soft—seatgd check wvalves. 5. Jet discharge lines from the sumps each contain two block valves in series that automatically close 1f the cell pressure becomes greater than atmospheric. A 1/2-in. connection is provided between the valves to test them fcor leak tightness. 6. The fuel sampling and enriching system is interlocked to prevent a direct opening to the atmosphere. 7. The steam-condensing system used in conjunction with the drain tank heat-removal system is a closed loop, except for the water supply lines, which contain soft-seated check valves, and the vent, which relieves to the cell vapor-condensing system (see sec. 6.2). 8, All cooling water lines entering the cell have soft-seated check valves or blcck valves contrelled by radiation monitors. All lines leaving the cells are provided with blcek valves controlled by radiation monitors. 9. The lubricating-oil system for the fuel circulating pump 1is a closed circulating lccp. Strict supervision is provided during additions of 0il or oil sampling tc assure that the containment is not violated. 10. The leak-detector system tubing operates at a higher pressure than the reactor process systems, sc no cutleakage is anticipated. 11. Several differential pressure cells and pressure transmitters are lccated cutside the cells but are connected to process piping inside through instrument tubing. The instrument lines are doubly contained. The case of the differential pressure cell provides primary containment. The instrument cases are located inside a chamber designed for an internal pressure of 50 psig to provide the secondary containment. 12. A1l helium supply lines connected to process equipment inside the cells contain one or more scoft-seated check valves. 13. Offgas lines that carry gaseocus fission products toc the charcoal beds are doubly contained for their entire length. The offgas lines from the charcoal beds have a common block valve, which closes on detection of 187 radiocactivity in the line. (For details, see Dwg. D-AA-A-40883, Appendix B. ) The gtresses which exist at various penetrations in the containment vessel wall were studied and were found to be within allowable values, with the maximum stresses occurring in the nozzles, @ 6.1.4 Leak Testing After construction of the reactor and drain tank cells was completed, leak tests were performed after the cells had been hydrostatically tested to 48 psig. The cells were held at 20 psig for 20 hr, and a leakage rate of O.25%/day was measured. When the reactor is operating, the cells will be kept at 12.7 psia, and the rate of inleakage will be monitored continu- ously. After each maintenance period, the cell leakage rate will be mea- sured and determined to be less than l%/day (assumed in dispersion calcu- lations, sec. 8.7.2) before the reactor is started. The meny service penetrations carrying air, water, etc., into the cells are equipped with various closing devices, as described above. The cell leakage test does not, of course, test these devices. It is intended that such closures be tested at intervals not greater than 6 months as verification of their integrity. 6.2 Vapor-Condensing System The MSRE vapor-condensing system 1s similar in principle to those of the SM-1A plant at Fort Greeley, Alaska, and the Humbolt Bay power plant. In the event of the maximum credible accident, vapcor will be discharged from the reactor cell into tanks, where the steam will be condensed in water and the noncondensable gases will be retained. The MSRE system differs from the SM-1A and Humbolt Bay systems in important details, however, because the energy release rates are low and the system is adapted tc an existing containment wvessel. Design of the vapor-con- densing system is based on the following:3s % 1. The volumes of the reactor and drain tank cells are, respec- tively, 11,300 and 6,700 ft2. These cells will operate at —2 psig and 150°F, Calculations have shown that only 7000 ft2 of noncondensable 188 gas will be discharged to a vapor-condensing system during the maximum credible accident. The free volume of the system was therefore specified as 4500 ft3, which is enough to contain 10,000 ft? of gas from the reactor cell at 30 psig and 140°T or 13,500 £t3 at 40 psig and 140°F. 2. The vapor-condensing system will contain 1200 ft2 of water at 70°F or less. The full 5 x 10 Btu calculated to be released by the salt can be abscrbed in the water without exceeding 140°F. 3. By the time the reactor cell pressure can reach 40 psig, the steam generation rate will be 16 lb/sec or less. The rate will decrease as the rate cf release of salt into the cell decreases. The relief line from the reactor cell to the vapor-condensing tank was specified to pass 16 lb/sec of steam at 40 psig with a pregsure drop of 10 psi or less. The discharge end of the line is tc be located 6 ft below the surface of the water in the tank to ensure complete condensation of the steam. . t 1s desirable to keep the vapor-condensing system isolated from the reactor cell and to limit its use to accidents in which the pressure rises above 15 psig. Bursting disks are to be installed in the relief line tc the vapor-condensing tank. The disks will burst at about 20-psi pressure difference. Rupture of the disk in the second (smaller) line will not affect the pressure at which the other disk breaks.? 5. Vacuum-relief wvalves will be installed in the reactor cell relief line at the vapcr-condensing tank, This will permit vapors to return to the reactor cell and prevent the pressure in that cell from going below -5 psig when the steam condenses at the end of the accident. 6. The vapcr-condensing system may contain as much as 400,000 curies of gaseous fisgion precducts and daughter prcducts 5 min after the beginning of a maximum credible accident and 40,000 curies after 1 hr. In addition, °L. F. Parsly, "MSRE Containment Vessel Stress Studies,” ORNL inter- nal document MSR-62-15, Feb. 2, 1962. *R. B. Briggs, "MSRE Pressure-Suppression System,” ORNL internal document MSR-£1-135, Nov. 15, 19&1. “L. F. Parsly, "Design of Pressure-Suppressicn System for the MSRE,” ORNL internal correspondence tc R. B. Briggs, Oct. 17, 1961. SLetter from T. E. Northup to R. G. Affel, May 28, 1964, Subject: MSRE Vapor Condensing System — Parallel Rupture Discs. 189 the water will contain some radiocactive golids that will be transferred as entrainment in the vapor. The radiation level at 10 meters from an unshielded tank would be about 2000 rad/hr at 5 min and 200 rad/hr at 1 hr; however, the tanks will be shielded sufficiently to reduce the initial radiation level to below 00 rad/hr. Provision has been made for the con- trolled discharge of the gases to the atmosphere through the filter system and stack. The vapor-condensing system will be installed about 40 £t east of Building 7503, as shown in Fig. 6.4. It will consist of two tanks, the vapor-condensing tank and the gas-retention tank. The vapor-condensing tank, sometimes referred to as the water tank, i1s a vertical cylinder that is normally maintained about two-thirds full of water through which gases forced from the reactor cell in a major accident would be bubbled to condense the steam. Tre tank containg about 1200 ft° of water plus a corrosion inhibitor. The estimated maximum of 5 X 10° Btu of heat that could be released from the fuel salt in the reactor and drain tank cells would raise the water temperature to about 140°F,3 The noncondensable gases are vented to a large gas storage tank. The vertical vapor-ccndensing tank is 10 ft in outside diameter and 23 ft 4 in. high, including the flanged and dished, l/2-in.—thick, top and bottom heads. The shell is 3/8 in. thick and constructed of SA-300 class I, A-201, grade B firebox steel. There are stiffening rings, 1 in. thick and 3 in. high, located on the exterior abcocut 2 ft 6 in. apart. The 12-in. gas inlet pipe in the top head extends 13 ft 8 in. intc the tank to about 6 ft below the normal water level. The tank is installed with the top about € ft below the normal grade level of 850 ft and the bottom at an elevation of about 819 £t. Additional earth is mounded above the tank to provide a total of 11 to 13 ft of shielding. The gas inlet line in the interior of the tank has a pipe cross (12 x 12 X 12 X 12 in.) about 3 ft above the water level to which is connected two 1l2-in. cast-steel-body check valves. These check valves close when the gas flow 1s into the tank but open and act as vacuum-relief valves to prevent water from being drawn in reverse flow through the inlet line during cooldown of the cell after the accident. SPECIAL EQUIPMENT ROOM IN REACTOR BLDG. FLOOR ELEV 852 ft 40 ft FROM REACTOR BLDG. UNCLASSIFIED ORNL-LR-DWG. 87162R2 ELEV 848 ft BURSTING DISK 4-in. LINE1 ELEV 836 ft 12-in-DiAM. ) RELIEF LINE t————3 +— ~ {00 ft —————— XX N 30-in.-DIAM. VENT DUCT FROM REACTOR CELL Fig. 6.4. «f TO STACK VAPOR CONDENSING 1C-ft DIAM. x 23-ft HIGH TANK \) {800 #3 - 1200 1> WATE R BUTTERFLY P o VALVES ‘ Vapor-Condensing System. ELEV. 858 ft 2-in-DIAM. VENT LINE X o — TO FILTERS AND STACK~y GROUND ELEV. FILL GAS 851 ft- 6 in. RE11'_E\I\|I\|T‘ TO FILTER, (><] FAN, AND STACK <] 1> REACTOR > 2 FFT FD-2 AUXILIARY CHARCOAL FFT BED FO-2 X D Hov-$ 544 HCV-573 P> 73, 50-psig DISK FD-2 FFT 40-psig HELIUM SUPPLY Fig. 7.5. Bystem Used in Filling Fuel Loop. 2. Opening HCV-573 relieves the pressure in the drain tank by vent- ing gas through the auxiliary charcoal bed to the stack. 3. Closing HCV-572 stops the addition of helium to the drain tank. During a filling accident all three actions would be initiated auto- matically (at 15 Mw) to ensure stopping the fill. Either of the first two actions stops the fill almost immediately and alsoc allows the fuel in the primary loop to run back to the drain tank. If the only action is to stop the gas addition, the fuel does not drain back to the tank, and the level "coasts" up for some distance after the gaé is shut off. This '"coastup" is a consequence of the pressure drops in the fill line and the offgas line during filling. Because of this, when gas addition is stopped, the fuel level in the primary loop continues to rise until the dynamic head 210 losses have been replaced by an increase in the static head difference between loop and drain tank. In the case of a filling accident, the inherent shutdown mechanism for power excursions is less effective than normally because the tempera- ture coefficient of reactivity of the fuel in the partially filled core is substantially less than that in the full system. This 1s so because, in the full system, the size of the core remains constant, and expansion causes some fuel ©o be expelled. In the partially full core, on the other hand, fuel expansion increases the effective height of the core, tending to offset the decrease in reactivity due to increased radial neutron leak- age. The effective fuel temperature coefficient of reactivity of fuel B with the core 60% full is approximately —0.4 X 10”5/°F compared with -5.0 X 107%/°F for the full core. The temperature coefficient of the graphite is not significantly affected by the fuel level. The most severe of the postulated filling accidents was analyzed in detail. It was assumed that the urarium ir fuel B was concentrated to 1.6 times the normal value by selective freezing of 39% of the salt in the drain tank, and several other abnormal situations were postulated to occur during the course of the accident, as follows: 1. Only two of the three control rods dropped on request during the initial power excursion. 2. Two of the three actions available for stopping the fill failled to function. Only the least effective action, stopping the gas addition, was used in the snelysis. This allowed the fuel level to coast up and make the reactor critical after the two control rods had been dropped to check the initial power excursion. 3. The helium supply pressure was 50 psig, the limit imposed by the rupture disk in the supply system, rather than the normal 40 psig. This — pressure gave a fill rate of 0.5 ftB/min when criticality was achieved and produced a level coastup of 0.2 ft after gas addition was stopped. The results of calculations of the power and temperature transients associated with the accident described above are shown in Figs. 7.6 and 7.7. Figure 7.6 shows the externally imposed reactivity transient exclu- sive of temperature compensatiocon effects. The essential features are the initial, almost-linear rise which produced the first power excursion as 211 Unclassified ORNL-DWG 64-6 ret 4 1.0 C.5F Net Reactivity {%) s h ety ..... ,,,,,, 0 50 100 150 200 250 300 350 Time (sec) Fig. 7.6. Net Reactivity Addition During Most Severe Filling Acci- dent. fuel flowed into the core, the sharp decrease as the rods were dropped, and the final rige as the fuel coasted up to its equilibrium level. Figure 7.7 shows the power transient and some pertinent temperatures. The fuel and graphite fiuclear average temperatures are the quantities which ulti- mately compensated for the excess reactivity introduced by the fuel coastup. 212 Unclassified ORNL-DWG 64-605 1360 1320 1280 Temperature (°F) 1240 1200 ) Power ( Fig. 7.7. Power and Temperazure Transients Following Most oevere Filling Accident. 213 The maximum fuel temperature refers to the temperature at the center of the hottest porticn of the hottest fuel channel., The power at initisl criticality was assumed to be 1 watt. The initial power excursion was limited to 24 Mw by the dropping control rods, which were tripped at 15 Mw, This excursion is not particularly important, since it did not result in much of a fuel temperature rise. After the initial excursion, the power dropped to about 10 kw, and some of the heat that had been produced in the fuel was transferred to the graphite. The resultant increase in the graph- ite nuclear average temperature helped to limit the severity of the second power excursion. Reactivity was added slowly enough by the Tfuel coastup that the rising graphite temperature was able to 1imit the second power excursion to only 2.5 Mw., The maximum temperature attalned, 1354°F, is well within the range that can be tolerated. 7.1.5 Tuel Additions Small amounts of uranium will be added to the circulating fuel during operation to compensate for U232 burnup and the buildup of long-lived poi- sons. The design of the fuel-addition system 1s such that only a limited amount of uranium can be added in any one batch, and it is added in such a way that it mixes gradually into the stream circulating through the core. These limitations ensure that the reactivity transients caused by a normal fuel addition are inconsequential. Fuel is added during operation through the sampling and enriching mechanism. Frozen salt (73% LiF-27% UF,) in a perforated container hold- U?3° is lowered into the pump bowl where it melts ing at most 120 g of into the 2.7 ft° of salt in the bowl. The 65-gpm bypass stream through the pump bowl gradually introduces the added uranium into the main circu- lating stream. The net increase of reactivity, once the 120 g of U233 is uniformly dispersed in the 70.5 ft? of salt in the fuel loop, is 0.07% 8k/k for fuel B and 0.03% 8k/x for fuels A or C. This increase is auto- matically compensated for by the servoe-driven control rod. An upper limit on the reactivity transient caused by a fuel addition was estimated by postulating that all 120 g of U??° was instantly dispersed throughout the pump bowl, from where it was mixed into the circulating 214 stream. This produced & step increase in the fuel concentration leaving the pump, and when this fuel began filling the core there was a reactivity increase. The largest reactivity increase would occur in the case of fuel B. (Reactivity effects would be less by a factor of 2 for fuels A and C because of their higher uranium concentrations and smaller concentration coefficients of reactivity., The maximum rate of reactivity 1ncrease for fuel B was only 0.017% 8k/k per sec, which is slower than the rate at which the regulating rod moves. Therefore, with the reactor under servo controel, there would be no disturbance in power or temperature. If there were no rod action the maximum reactivity increase would be 0.09% Bk/k, during the first passage of the more concentrated fuel through the core. Only moder- ate transients in power or temperature could be produced by this much ex- cess reactivity. 7.1.¢ UO, Precipitation The chemical stability of the fuel poses safety problems in a fluid fuel reactor; for if pheses which are unusuelly concentrated in uranium can appear, deposits may cocllect and cause local overheating or, if they shift position, reactivity excursions. The only known way in which ura- nium could become concentrated in the MSRE fuel-circulating system is by gross contamination of the salt with oxygen, with consequent precipitation of UOp. As discussed in Section 1.1, the fuel salt for the MSRE is pro- tected zgeinst oxidation and subsequent precipitation of uranium by the inclusion of Zrr, in the fuvel saglt. The ratio of ZrF, to UF,; in the fuel will be over Z, and mcre than 2 f£2 of water or 9000 ft2 of air would have to react with the 70 ©t° of fuel salt before so much of the ZrF, was con- verted to oxide that U0, would begin to appear. The only way tlLst such quantities of water or air could enter the system in a short period of time would be some failure or accident during maintenance with the system open. During operation, only gradual contamination is at all credible. After any maintenance period and at frequent intervals during operation, fuel salt samples will be examined and analyzed for zirconium and other oxides. Thus oxide contamination should be detected long before the ura- nium began to precipitate, and appropriate action could be taken to pre- vent such precipitation. 215 Although it is extremely unlikely that the precautions against oxygen contamination, the protection of the ZrF,, and the surveillance of samples will fail to prevent UO, precipitation, the consequences of such precipi- tation were considered. The most likely regions for precipitated UO, to collect are in the heads of the reactor vessel, where velocities are lowest. Collection in the lower head would create the greatest potential for a large, positive reactivity disturbance because the lower head is near the inlet to the core. There is no known way in which a substantial amount of deposited UOp could be quickly resuspended, either by physical disturbance or dissolution in the fuel salt. Nevertheless, if some UO; were resuspended in the salt below the core, it could pass rapidly through the center of the core and cause a reactivity excursion. The greatest effect would result if it passed up through the high-velocity (2 ft/sec) channels around the center of the core, Excess uranium introduced at the lower end of a central channel would cause the reactivity to increase to a peak in about 1.5 sec and then de- crease over a like period. Calculations were made using the reactivity coefficients and neutron lifetime of fuel C to determine the consequences of such excursions. The temperature and pressure transients caused by this type of reactivity excursion with peaks up to 0.7% Bk/k are tolerable without any control rod action.* Rod drop at a power level of 15 Mw does not raise the tolerable reactivity addition very much in excursions of this kind, because the period at 15 Mw is so short (about &80 msec) that self-shutdown occurs before the rods have time to contribute very much. Calculations for fuel C indicated that the 15-Mw rod drop raised the toler- able reactivity peak from 0.7% to about 0.9% 8k/k. Rather large amounts of excess uranium would be required to produce the l1imiting excursions. Assuming that all the uranium passed through a central channel in a single blob, over 700 g of U (0.8 kg of UO,) would be required to give a peak of 0.7% 8k/k when the fuel is of composition C. *Tolerable here means that the pressure excursion is less than 50 psig and the peak temperature is less than 1800°F for the most unfavor- able initial power. 216 Equivalent excursions would be produced by about 200 g of U in fuel A or 100 g of U in fuel B. The question arises: how do these quantities compare with the amount of U0, whose separation could be detected by its effect on reactivity? A reactivity effect would be produced 1if the average nuclear importance of the separated uranium differed from that of the circulating uranium. The low-velocity regions where precipitated UO,; may collect are at lower-than- average importance, so the reactivity would probably decrease if UO; were lost. During critical operation a reactivity balance will normally be made at 5-min intervals. (The on-line digital computer connected to the MSRE is programmed to do this automatically, but operation of the computer is not regarded as a necessary condition for nuclear operation.) In this balance, the calculated effects of temperature, power, xenon, burnup, and control rod positions are included. The minirum change in reactivity, occurring over a period of a few hours, that could be recognized as an anomaly is expected to be about 0.1% Sk/k. The reactivity change caused by deposition of UO, where it contributes nothing to the chain reaction in the core is 0.23, 0.45, or 0.066% 8k/k per kg of UO, for fuels A, B, or C, respectively. Therefore the minimum amounts of UO,; separation which could be detected by this means are about C.4, 0.2, and 1.5 kg of UO; for fuels A, B, and C, respectively. It is evident that the reasctivity balance cannot be relied on to de- tect U0, loss below the level at which the concelvable effects of complete and sudden recovery becorme important. Any UO, deposits would probably be quite stable, however, so the probability of sudden resuspension of a large fraction of a deposit is very small. (In the HRE-2, where deposits could be dispersed by the movement of the loose core inlet diffusers or by steam formation, and the dispersed material was soluble, the largest "instanta- neous ' recovery was less than 0.1 of the existing power-dependent deposits of uranium.) Therefore it is probable that if UO,; precipitation were to develop, it would produce a detectable reactivity decrease before any serious reactivity excursion would have a reasonable chance of occurring. This is only added assuranéé, however, for the real protection against damaging excursions is provided by the measures which prevent UO, forma- tion in the first place. 217 Abnormal, localized heating would result if a UO, deposit (or other solids containing uranium) were located where the neutron flux caused fis- sion heating in the deposit. One place where the reactor vessel would most likely be subject to such heating is the lower head. For this reason thermocouples are attached to the outside of the lower head and the dif- ferences between these tenperatures and the temperature of the fuel enter- ing the reactor vessel are monitored continuously to detect abnormal heat- ing. This is primarily a means of detecting a deposit of uranium solids rather than a guard against overheating, since, from the standpoint of damage to the vessel, the heating in the lower head is not expected to be serious, even 1if fairly heavy deposits of uranium collect. The reason is that the specific heat source in the uranium-bearing solids on the lower head would not be intense (3 to 8 w per g of U0, when the power is 10 Mw) because the neutron flux would be greatly attenuated by the salt and the INOR grid below the core. Calculated temperature differences at 10 Mw are shown in Fig. 7.8. (The differences are proportional to power.) The curves are different for fuels A, B, and C because of differences in flux and uranium enrichment. Another region susceptible to local heating due to UO, deposition 1s the pocket between the top of the core shell and the reactor vessel, above the core shell support flange. There are passages at 10-in. inter- vals around the periphery of the support flange which bypass some salt from the vessel inlet for cooling and for providing some rotational flow in this region. Although the vertical component of the velocity in the pocket may not be sufficient to eliminate settling of UO;, the swirling motion should result in an even distribution if deposition should occur. The area for deposition is large enough (180 in.?) that the loss of ura- nium from the circulating salt would be detectable before a UO, deposit could become deep enough to cause intolerable heating in this region. The maximum vessel temperature in this vicinity was calculated by assuming that enough uranium separsted to cause loss of 0.2% 6k/k in reactivity and that all this uranium accumulated as UOs on top of the core shell sup- port flange. The calculated maximum temperature in the vessel wall was 218 Unclassified ORNL-DWG 64-609 U0, Deposit (kg/ft?) Effects of Deposited Uranium on Afterheat, Graphite Tem- Fig. 7.8. perature, and Core React ivity. 219 less than 1600°F with the reactor at 10 Mw and the core inlet temperature at 1175°F. The possibility of settling on the upper support flange was examined experimentally by examining the flange area of the full scale hydraulic mockup after it had been operated with iron filings in the circulating water, but no preferential settling in this area was observed. In summary, this examination indicates that the conseguences of uranium precipitation do not appear to be of a serlous nature from the hazards standpoint. 7.1.7 Graphite Loss or Permeation If a small part of the graphite in the core were replaced by an equal volume of fuel salt, the reactivity would increase. The effect would not be large, amounting to less than 0.003% 8k/k per in.? of graphite replaced by fuel or only 0.13% Sk/k 1f the entire central stringer were replaced with fuel. The reactivity transients caused by loss of graphite due to credible mechanisms presert no hazard. Minor loss of graphite from the core will occur if chips of graphite break loose and float cut of the core. The vertical graphite pieces are fastened at their lower erds and are secured by wire through their upper tips. Thus it 1s very unlikely that a piece of graphite large enough to cause a noticeable change in reactivity could fioat out of the core. Another mechanism whereby fuel replaces graphite in the core is the bowing of the graphite stringers, which causes the volume fraction of the fuel near the center of the core to increase. The bowing is caused by the longitudinal shrinkage of the graphite under fast neutron irradiation. The radial gradient of the neutron flux causes unequal shrinkage on op=- posite sides of a stringer and produces bowing that is convex in the di- rection of the gradient. The shrinkage is very siow, and the increase in reactivity due to unrestrained bowing has been estimated at only 0.6% 6k/k per full-power year. Graphite shrinkage also causes small increases in reactivity because the moderator density is increased and because the lateral shrinkage gradually opens the channels and allows more fuel in the core. These ef- fects are of no concern because they cause less than 0,2% ak/k increase per full-power year and there is no possibility of rapid changes. 220 A reactivity increase would also result from fuel salt permeating the pores in the graphite and thus increasing the amount of uranium in the core. There is no known way, however, in which increased permeation could cause a sudden or dangerous increase in reactivity. Tests invariably have shown that the fuel salt does not wet the graphite and that penetration of the fuel into the graphite pores is limited by surface tension, the size of the pores, and the external pressure. Neither of the first two variables 1s subject to rapid or drastic change. As indicated in Table 1.5 (sec. 1.1.3), the measured permeability of salt in the graphite is only 0.2 vol % at 150 psig. The pressure effect on permeation produces little reactivity effect, being less than 5 X 1076 Ek/k per psi and is therefore unimportant either as a means of externally increasing reactivity or as a feedback during transients involving pressure excursions. Another means for increasing the reactivity of the core is sorption of uranium on the graphite surface, None of the many out-of-pile tests has shown evidence for such scorption. However, uranium in amounts as great as 1 mg of uranium vper em® of graphite surface has been observed in a relatively thin surface layer on specimens from capsules irradiated for 1500 hr to fission power densities considerably above those expected for the MSRE. TUranium sorption Las been observed so far only on graphite from experiments in which copious quantities of F, and CF, were evolved by radiolysis of the frozen fuel mixture. (Such radiolytic evolution prob- ably occurred several times during the irradiation experiment since the assemblies were ccoled through reactor shutdowns many times during their irradiation history.) It is possible that the uranium was laid down on the graphite by reactions such as 4UF3 + 2C — UC, + 3UF,, during sudden reheat of the capsule in which the frozen fuel had thus be- come deficient in Fp. According to this explanation the sorbed uranium would not be expected in MSRE operation. It is possible, on the other hand, that the sorbed uranium is due to some unknown (and thermodynami- cally unlikely) effect of irradiation and fission at elevated temperatures and that it can occur under conditions of operation of the MSRE. Uranium in or on the graphite, through whatever mechanism, would increase the 221 reactivity of the assembly, the graphite temperature during operation, and the quantity of afterheat in the graphite. Figure 7.9 shows each of these effects as calculated for each of the three fuel compositions previously discussed, All effects are considerably smaller when salt C (the first salt proposed for MSRE operation) is used, but the reactivity increase, in any case, 1s sufficiently large to be hazardous only if it can occur in a short time. 7.1.8 Loss of Flow Interruption of fuel circulation can produce power and temperature excursions through two mechanisms: a reduction in heat removal from the core and an increase in the effective delayed-neutron fraction. When the fuel is circulating normally, delayed-neutron precursors produced in the core are distributed throughout the entire volume of circulating fuel, and nearly half decay outside the core. If the circulation is interrupted at any time, the delayed-neutron effect produces a reactivity effect of 0.3% Sk/k over a period of many seconds. (The increase in the effective de- layed-neutron fraction would be slow, even if the flow could be stopped instantaneously, because cof the time required for the precursor concen- trations to build up to equilibrium.) A core temperature rise of less than 50°F is enough to compensate for the increased reactivity and, be- cause of the nature of the reactivity increase, no hazardous power excur- sion is produced in any case. If the reactor is at high power when cir- culation is interrupted, the decreases in heat removal from the core and heat transfer to the coolant will cause fuel temperatures to rise and the coolant temperature to fall. The only likely cause of loss of fuel flow is failure of the power to the fuel pump. The results of a simulated fuel pump power failure at 10 Mw, with no corrective action, are shown in Fig. 7.10. The coastdown in flow after the failure of the pump power was simulated by reducing the circulation rate exponentially with a time constant of 2 sec until it reached the thermal-convection circulation rate determined by the tempera- ture difference across the core. (The flow deceleration was based on pump 222 Uneclassified ORNL mperature rise I core at 30 days .4 Temperature (°F) IS Graphite temperatu s {at 10 Mw srature (°F) Tem 5 6y L, k/k) inereasge Reactivity (% - o fLi 0 0.2 0.4 0.6 0. 1.0 ( Uranium Deposited on Graphite (mg/cm?®) Fig. 7.9. Power and Temperatures Following Fuel Pump Failure, with no Corrective Action. 223 UNCLASSIFIED CRNL-LR-DWG 67578 1 1400 o —— ORE OUTLET / ———— CORE FUEL MEAN 1300 .////,, B -t ///i;// hfififii / """ GRAPHITE MEAN 1200 dn ¥ g’ iy - - ——————__| CORE INLET o — L \ 5 100 e % RADIATOR INLET w = \\ 1000 \\\%\\\\\M \\\ \ \ 900 . \ —\ SALT LEAVING RADIATOR e h—____- - -—..‘: 800 12 /\ 10 AN \\QSWER —_ 8 = = - N m 6 T ) K I \\ e - \\ 2 ._-.__-__* 0 0 20 40 60 80 100 120 TIME (sec) Fig. 7.10. Power and Temperatures Following Fuel Pump Power Failure. Radiator doors closed and control rods driven in after failure. 224 loop measurements. ! The heat transfer and thermal convection were pre- dicted from basic data.) The initial increase in nuclear power was caused by the delayed-neutron effects. The decrease in heat removal from the core, coupled with the high production, caused the core outlet temperature to rise. As the fuel flow and the heat transfer in the heat exchahger fell, the continued heat extraction at the radiator caused the cocolant salt temperature to decrease and reach the freezing point in less than 2 min. The behavior in simulator tests at lower power was similar, but the coolant temperature remained above the freezing point 1f the air flow through the radiator was such that the initial power was less than 7 Mw. The effects of & fuel pump power Ffailure are minimized by automatic action of the safety and control systems. At nuclear powers above 1.5 Mw, control interlocks start a rod reverse (insert rods at 0.5 in./sec) as soon as the pump begins to slow down. A rod reverse is called for in any event if the core outlet temperature reaches 1275°F. At an outlet tem- perature of 1300°F the safety system drops all rods. These actions hold down the heat generation and core temperatures. The coolant salt is kept from freezing by other szction. When the pump stops, an automatic "load reverse' lowers the radiator air flow to about 0.2 of the normal 10-Mw rate. The safety system drops the doors to shut off the air completely if the temperature of the salt leaving the radiator reaches 900°F. These actions prevent freezing of the coolant salt in the radiator. Simulator results that illustrate the effectiveness of actions simi- lar to those designed into the control and safety systems are given in Fig. 7.11. Rod insertion was simulated by a negstive reactivity ramp of ~0.075% 6k/k per sec beginning 1 sec after the pump power was cut. The radiator door action in the present control system was not simulated. Instead the simulated heat removal from the radiator tubes was reduced to zero over a period of 30 sec beginning 3 sec after the pump power failure. 10, W. Burke, "MSRE — Analog Computer Simulation of the System for Various Conditions," internal ORNL document CF-61-3-42, March 8, 1961. 225 UNCLASSIFIED ORNL-LR-DWG 67579 1300 CORE QUTLET GRAPHITE MEAN w ° 1200 E FUEL MEA kel x > CORE INLET g o wi ; 1100 = RADIATOR INLET — 1000 RADIATOR OUTLET FISSION POWER > = o wi = O o O 20 40 60 80 100 120 TIME (sec) Fig. 7.11. Effects of Afterheat in Reactor Vessel Filled with Fuel Salt After Operation for 1000 hr at 10 Mw, 7.1.9 Loss of Load The type of incident usually referred to as the loss-of-load accident takes the form, in the MSRE, of interruption of heat removal from the ra- diator while the reactor is operating at high power. The most likely way for this to occur would be for the radiator doors to drop. Failure of the cooling air blowers would cause less of a load drop because natural draft through the radiator can remove up to 3 Mw of heat. The thermal and nuclear characteristics of the MSRE are such that core temperatures do not rise excessively in any loss-of-load accident, even with no external control action. Temperature transients are mild 226 and there is no core pressure surge. This behavior was shown in simulator tests in which the heat removal from the radiator tubes was instantaneously cut off with the reactor initially at 10 Mw. The temperature coefficients of reactivity for fuel C were used and no control rod action was assumed. Tn these tests the core outlet temperature rose less than 40°F, over a period of about 4 min. The average temperature of the coolant salt rose 196°F, with 100°F of this rise occurring in the first 60 sec. The coolant salt volume increased by 1.0 ft3, which, if there were no gas vented from the coolant pump bowl, would cause a pressure rise of 20 psi. The pressure control system would be capable of limiting the pressure rise to less than 11 psi. 7.1.10 Afterheat The problems asgsociated with the decay heat from the fission prcducts (afterheat) are quite moderate in the MSRE because of the relatively low specific power in the reactor. Thus temperatures change slowly and no rapid emergency action is required to prevent overheating. This may be illustrated by considering a hypothetical situation in which the reactor vessel remains full of fuel, with the circulation stopped, after long- term, high-power operation. The vessel contains three-fourths of the fuel salt in the system and gbout three-~fourths of the fission products. The total heat capacity of the fuel, the graphite, and the metal vessel is 8030 Btu/°F. The estimated heat loss from the vessel at 1200°F is about 30 kw. The effects of fission-product heating in the vessel after 1000 hr at 10 Mw are shown in Fig. 7.12. It appears that no serious increase in temperature will result from afterheat if the fuel is trapped in the core. After the fuel is drained from the reactor vessel there will still be some afterheat in the core from fission products which remain in the graphite. These products are there because of diffusion of gaseous prod- ucts into the graphite and because of direct production by fissions in the fuel which had soaked into the graphite. The heating rate in the drained core 1 hr after shutdown from extended operation at 10 Mw is 4.5 kw, most of which is produced by the gaseous fission products and thelir daughters. 22 Unclassgified CRNL-DWG 64-607 600 + ANo’heat - Temperature I Televa bl o + AA‘l, T — To (°F) w O o 100§ 100 & & Q 5 a 40 O " g o < i a 9 o 20 © 2 2 as] 10 0 20 40 60 80 100 120 140 Time (hr) Fig. 7.12. Temperature Rise of Fuel in Drain Tank Beginning 15 min After Reactor Operation for 1000 hr at 10 Mw. 228 After 24 hr the heating rate is down to only 0.44 kw, so heat losses from the graphite will probably keep the temperatures from rising significantly. But even if there were no heat loss from the core graphite (which has a heat capacity of 3350 Btu/°F), the temperature rise would not be hazard- ous: 27°F in one day, 144°F in 30 days. These values were calculated by assuming that fuel occupiles 0.1% of the graphite volume. For this condi- tion, the fission products produced in situ produce only one-third or less of the total heating, so the results are not very sensitive to the exact amount of permeation. If the fuel were transferred to a drain tank shortly after high-power operation, the rate of temperature rise in the absence of cooling would be higher than in the core because the total heat capacity connected with the fuel would be less in the drain tank than in the core, However, the drain tanks are provided with cooling tubes designed to remove 100 kw of heat from the salt at 1200°F. This heat-removal capacity is adequate to prevent & significant rise in temperature. The predicted heat loss from a drain tank at 1200°F is 20 kw. Figure 7.13 shows the expected tempera- ture rise of the fuel charge in a drain tank beginning 15 min after the end of 1000 hr of operation at 10 Mw. (It will take approximately this long to get the salt into the drain tank. The temperature at this time will depend on the manner of shutting down the nuclear power and the ra- diator heat removal, but it will be near 1200°F.) The heat removal rates of 20 and 120 kw correspond, respectively, to heat losses and the heat removal with the cooling system working as designed. The heat capacity of the INOR tank and structure was ignored in computing the temperature rise. As was described in Section 1.2.5, the cooling system for the drain tanks is based on the evaporation of water, but the water and salt are not in contact with a common wall., The evaporation-condensation circuit is completely closed and requires no pumps because the condensate is re- turned by gravity. An emergency reservoir of 500 gal is installed to provide cooling for 6 to & hr. During this time, hoses could be connected to provide a temporary supply for an indefinite time. 229 33 O -~ 4y ol ©v 0 @ — & 5 ] : 5 arnyexadus], 100 ~100 Time (hr) Heating of Reactor Vessel Lower Head by U0, Deposits 7.13, with Power Level at 10 Mw. Fig. 230 7.1.11 Criticality in the Drain Tanks Fuel salt having the uranium concentration appropriate for power operation of the reactor cannot form a critical mass in a drain tank or in a storage tank under any conditions. Only if there were a large in- crease in uranium concentration could criticality occur in a tank. The only credible way for this to occur would bhe for the salt to freeze gradu- ally, leave the UF, in the remaining melt, and thereby concentrate the uranium intc a fraction of the salt volume. Calculations indicate that concentration by more than a factor of 4 would be necessary for criticality in any such case.? Concentration of the uranium is not impossible; gram-scale studies of equilibrium cooling of fuel salt mixtures indicated that the last phase to Treeze was 7LiF-6(Th,U)F4 containing about three times the uranium concern- tration of the original mixture.? However, 1in a vessel as large as the frel drain tank, it 1s unlikely thal such gross segregation would occur. Convection currents within the lurge mass and initial freezing on the many cooling thimbles should keep the segregated salt distributed through the mess. Even so, special measures have been taken to prevent freezing. Heater feilures are not lixzely to cause accidental freezing, since a temperature of 1200°F can be maintained with only about 0.6 of the installed heaters in cperation. Even 1f the heater vower should be cut off, more than 24 hr would be recuired for the heat losses (about 20 kw) to ccol the salt to the liguidus temperature of &40°F. TFission-product afterhesat would extend this tire by many hours. A complete power failure for this length of time ig extremely unlikely because the reactor ares is supplied by two separste feeder lines and there are three emergency dissel-clectric generating units in the area, any of which could supply vower to the heaters. In the event =11 these sources failed, or for any reason it should become necessary to allow the fuel salt to sclidify and cool to ambient temperature, the risk of criticality due ic selective freezing ?J. R. Engel and B. E. Prince, "Criticality Factors in MSRE Fuel Storage and Drain Tanks, " USAEC Report ORNL-TM-752 (to be published). 3 0ak Ridge National Laboratory Status and Progress Report November 1963, p. 15, USAEC Report ORNL-3543. 231 will be eliminated by dividing the fuel between two tanks, neither of which will contain enough uranium for criticality. This can be done because one drain tank will be kept empty for such an emergency. Although criticality in the drain tanks is undesirable and can be avoided as described above, the consequences would not be severe. Because criticality could occur only after at least three-fourths of the salt had frozen, the initial temperatures would be moderate. When the concentrat- ing process slowly reached the point of criticality, the power would rise to match the heat losses without any serious power excursion. (The salt contains a strong (a,n) source; the reactivity would be increasing very slowly through the critical point, and there is a negative temperature coefficient of reactivity.) The establishment of a self-sustaining fis- sion chain reaction in a tank should therefore present no threat to the containment of the fuel. TIFurthermore, the location of the tanks and the shielding around them is such that radiation from a critical tank would not prevent occupancy of the building and emergency work toward restoring heater power. 7.2 Nonnuclear Incidents 7.2.1 Freeze-Valve Failure The freezing and thawing of the freeze valves could conceivably re- sult in a rupture of the piping. This is specifically minimized in the design by meking the freeze-valve gsection as short as possible so that there 1s small danger of bursting as a result of entrapment of liquid be- tween the ends of a plug. Because the salt expands as it thaws, special precautions are taken to apply the heating so that expansion space is always available. A single freeze valve has been frozen and thawed more than 100 times without apparent damage. 7.2.2 Freeze-Flange Failure The flanged Jjoints in the fuel-circulating loop might also fail. The freeze flange was selected as the simplest and most reliable joint avail- able, and the strengths of the bolting and the flange compression members 232 are considerably in excess of those necessary to maintain a tight joint. Typical flanges have been tested by thermal cycling over a period of more than two years, and there has been no indication of unsatisfactory per- formance., The helium-pressurized leak-detection system will monitor con- stantly for leakage and will protect against leakage to the outside. 7.2.3 Excessive Wall Temperatures and Stresses v Overheating of pipe and vessel walls might occur because of external electric heating or internal gamma heating. The heater elements have a melting point several hundred degrees above that of the INOR-8, but it is not considered likely that the THOR-E could be melted by external heating sC long as salt was present ilnside the pipes. 1If salt were not present, the pipe wall might melt before the heater element melted, but in this case only a small fraction of the activity would be released. Wall tem- peratures will be measured by hundreds of thermocouples and will be con- stantly scanned to prevent excessive heating. Also during prestartup testing, the controls of sll external heaters and the resistance-heated drain line will be provided with mechanical stops set to limit the power input of each circuit to values that meet the heat input requirements with a small excess capacity. During reactor operation, various components will be exposed to high gamma fluxes that will cause garmma heating of the components. If this heating produces large teuwperature gradients, excessive thermal stresses may arise. The effects of germa heating on the core vessel and the grid structure were investigated because these structures will be in regions of the highest gamma flux. The gamma heating of the core vessel will result in a temperature difference of only 1.3°F ascross the vessel walls and will produce a cal- culated thermal stress of 300 psi, which is not serious. The temperature difference across each grid of the support grid structure will be 3.8°F, and the resulting thermal stress will be 850 psi, which 1s also not seri- ous, Camma and beta heating of the top of the fuel-pump bowl will result in a maximum thermal stress of about 8000 psi at the junction of the bowl 233 with the volute support cylinder, with cooling air flowing at 200 cfm across the bowl. The Junction temperature under these conditions will be about 1000°F. Considerably higher stresses will occur at cooling air flow rates other than 200 cfm during power changes. The highest stress, about 19,000 psi, will occur in changing from O tc 10 Mw, with the cooling air flowing at 50 efm. With a gas pressure of 5 psig in the bowl, the maxi- mum combined circumferential and meridional pressure stress will be about 300 pei in the pump bowl and about 400 psi in the volute support cylinder at the Jjunction. The resulting maximum combined stress with cooling air flowing at 200 cfm will be 8400 psi, and, with ailr flowing at 50 cim, 19,400 psi. In a normal thermal cycle after a complete shutdown, the temperature of the reactor will vary between 70 and 1300°F. With such a range, there are possibilities for excessive stresses as the piping expands and con- tracts. Particular attention has therefore been given to providing a flexible layout. Analyses of the extreme conditions have indicated that the maximum stress caused by expansion or contraction is only 7050 psi. Instrumentation is provided to observe the normal rate of heatup or cool- down, which will not exceed 120°F/hr. Thus it does not appear reasonable that primary contaimment fallures will occur as a result of excessive tem- peratures or thermal stresses. 7.2.4 Corrosion Another possible cause for failure of the primary container is cor- rosion. As reported in Section 1.1.2, the corrosion rates experienced with the INOR-& alloy have been very low (less than 1 mil/yr) for periods as long as 15,000 hr. All available evidence indicates that it is ex- tremely unlikely that corrosion will be a cause of piping failures. TNumerous corrosion tests have been completed with fuel mixtures of the type to be utilized in the MSRE. Results from 37 INOR-8 thermal- convection loops, 17 of which operated in excess of one year, have shown complete compatibility between INOR-8 and the beryllium-based fluoride systems. Experiments conducted in INOR-8 forced-convection loops for one year or more similarly have shown low corrosion rates in fluoride mixtures 234 of this type. The operating conditions of these experiments are shown below: Forced- Thermal - Convection Convection Variable Loops Loops Fluid-metal interface 1300°F 1350°F temperature Fluid temperature 200°F 170°F gradient Flow rate ~2 gpm ~7 fpm Metallographic examinations of INOR-8 surfaces following salt exposure in these loop experiments reveal no significant corrosion effects in time periods up to 5000 hr. At times longer than 5000 hr, =a thin (less than 1/2 mil) continuous surface layer develops at the salt-metal interface. Some typical weight-loss data are presented below: Weight Loss Time (hr) mg/cm2 mg/cmz/mo 5,000 1.8 0.26 10,000 2.1 0.15 15,000 1.7 0.08 Sixteen in-pile capsules exposed for 1500 to 1700 hr each* and five in-pile fuel-circulating lOOps5 which ran a total of 3000 hr have been examined Tor evidence that corrosion under irradiation is different from that out of pile. Particular attention was paid to the possible effects of free fluorine. Mo evidence was found which indicated that high ir- radiation levels altered the normal corrosion pattern. It has been suggested that the high radiation levels in the control- rod thimbles and in the reactor cell might cause formation of sufficient 40gk Ridge National Laboratory, "MSRP Semiann. Prog. Rep. Jan. 31, 1963, " USAEC Report ORNL-3419, pp. 80-107, and "MSRP Semiann. Prog. Rep. July 31, 1963," USAEC Report ORNL-3529, Chapter 4. 5D. B. Trauger and J. A. Conlin, Jr., "Circulating Fused-Salt Fuel Trradiation Test Loop," Nucl. Sci. Eng., 9(1): 346~356 (March 1961 ), 235 nitric acid (from the Ns, 05, and II,0 in the cell) to create corrosion problems. A calculation for the most pessimistic conditions indicates that the concentration of nitrogen oxides in the cell atmosphere might reach 1% in 4000 hr, the longest anticipated period of continuous opera- tion. The actual concentration should not be this high. The HRE-2 was operated under similar conditions without observable damage from HNO3, so no real problilem is expected. Corrosion in the MSRE will be followed continuously by means of the salt chemistry, which will indicate nickel removal. Salt samples for this purpose will be removed and analyzed daily. At approximately é-month in- tervals, the surveilllance specimens referred to below will be tested. 7.2.5 Material Surveillance Testing Surveillance specimens of INOR-8 and type CGB graphite will be placed in a central pogsition of the reactor core (as shown in Fig. 1.6) to survey the effects of reactor operations on these materials from which the re- actor and moderator are constructed. The specimens will be made from the material stock used to fabricate the reactor primary system and moderator. Separate but identical control specimens will be exposed to a duplicate thermal history while submerged in unirradiated fuel salt in an INOR-8 container to differentiate between the effects of temperature and the effects of drradiation. ©Specimens will be removed from the reactor and examined after six months of operation and then again after periocds of operation that should result in meaningful data based on the results of the first set of specimens. New specimens will replace the material re- moved. The analysis of INOR-8 specimens will include metallographic examina- tion for structural changes and corrosion effects, mechanical properties, and a general check for material integrity and dimensional changes. The graphite specimens analysis will include metallographic and radiographic examination for salt permeation and possible wetting effects, mechanical properties tests, dimensional checks for shrinkage effects, chemical analy- sis for deposits on the graphite, electrical resistivity measurements, and a general inspection for material integrity. 236 The results from surveillance specimens will be available in plenty of time to guide decisions on the allowable safe exposure of the graphite and INOR at varicus locations in the reactor. 7.3 Detection of Salt Spillage The escape of activity from the primary container would be detected by radiation monitors. If the spillage were into the secondary container, the activity would be indicated by monitors on a system which will con- tinuously sample the cell atmosphere. Leakage into a service line (e.g., the cooling water) would be detected by monitors attached to the line just outside the cell. In either case, the action of the monitors would be to stop power removal and to insert the control rods. The salt would be drained unless the leak were 1n a drain tank, in which case it would be transferred to another tank. 7.4 Most Probable Accident Although the chances of any failure are extremely small, one of the above described nonnuclear incidents can be considered as the most likely cause of leakage Trom the primary containment system. Based on previcus experience with fluid-fuel reactors, it is believed that the most probable type of leakage would be a slow drip or spray at a rate of a few cubic centimeters per minute. It is further believed that the leakage would be detected and the reactor shut down before more than 3 or 4 liters had es- caped. By this time, about 20,000 curies would have been released into the secondary container. If 10% of the solid fission products (approxi- mately 1400 curies), 10% of the I, (approximately 480 curies), and all the xenon and krypton (approximately 3200 curies) were dispersed in the cell atmosphere, the concentration of activity inside the container would be 107 curies/cm3, which would alarm the cell air radiation monitors and shut down the reactor. This activity (neglecting decay, for simplicity) would be pumped from the cell by the container vacuum pumps as they main- tained the cell pressure below atmospheric by compensating for the 10“6%/min 237 normal inleakage (at 13 psia). The activity would be discharged through line 917 to charcoal bed 3 and then to the absolute filter-stack system. Assuming that the carbon bed and the absolute filters are reasonably effective, the disposal of the activity would not be a large problem, and concentration could easily be kept below the MPCa. Even without the char- coal bed the concentration of iodine would probably be tolersble (see Appendix D). After an accident such as that described above, it would be necessary to open the secondary container to repair the fallure, and the activity would be released at a rate considerably greater than the normal rate of cell inleakage. In this circumstance, the container inleakage would be increased manually, and the concentration of activity in the stack would be increased to levels which would permit meximum discharge without ex- ceeding permissible exposures downwind. In any case, the container would not be opened for repairs until the activity concentration inside was low enough to prevent overexposure in case the cell ventilation system failed. 238 2. DAMAGE TO THE S=ECONDARY CONTAINER The possibilities for demage to the secondary container were con- sidered. As indicated below, damege from missiles is unlikely, and pro- tection is provided against the puildup of excessive internal pressure. Site studies have indicated that no problems exist with respect to earth- quakes and floods, and protecticn hzs been provided against the conse- quences of arson. The analysis cof tThe maxinum credible accident, which - involves simultanecus release of the meolien fuel and water intc the sec- ondary container, led to the incorporaticn of a vapor-condensing system . fer limiting the container pressure to 40 psig. The conseguences of release of radicactivity frcom the secondary container atfter the maximum credible accident were analyzed Ior two situations: with the stack fan off and with the stack fan still running and passing the effluent thrcugh ebeclute filters. ALl doses were found 1o be low encugh to allow evacua- ticn of the bhuilding or arez in reascnable Times without overexposure. £.1 Mgsile Damage Damage by missiles does not appear to be likely. The maximum pres- sure expected in the reactor system is lese thar 100 psig, and the INOR-8 strucTural material 1is very ductile at the ncrmal cperating temperature. no very large pressure excursicn can be envisicored without assuming that the core inlet and exit liznes are both frozen. Although missiles with significant velccities are not thought to be credible, the reactor vessel ig protected from missiles by the stainless steel thermal shield that completely surrcunds 1t. 2.2 Excegsive Pressure £,2.1 Salt Spillage The spillage of the salt at 3 high temperature does have possibilities for raising the cell pressure to high values. A rapid spill intoc the cell of all the salt in both the fuel and coclant systems would heat the cell 239 atmosphere sufficiently to produce a 2.4-psig final pressure.l If the fuel were released as a fine spray, the maximum pressure would be 16.4 psig. The worst situation would be the simultaneous release of the salt and the inleakage of the correct amount of water to allow the generation of steam without subsequent cooling from additional inleakage of water. This accident is considered tc be the maximum credible accident and is discussed in detail in Section 8. 6. 8.2.2 0il ILine Rupture The fuel-pump lubrication system contains a maximum of 28 gal of oil, which, in the event of an oil-line rupture, could come into contact with the hot pump bowl and the reactor vessel. With an atmosphere containing the normal 21% of oxygen in the cell, the oil would burn and produce an excessive pressure in the cell, cr it would form an explosive mixture that might later be ignited. To ensure against containment damage by these possibilities, the oxygen content of the cell will be kept below 5% by dilution with nitro- gen. Nitrogen will be fed into the cell continuocusly to maintain the low oxygen ccntent. 8.3 Acts of Nature 8.3.1 Farthguake As reported in the site description, Section 4.3.8, earthquakes are not a problem at this site. g8.3.2 TFlood The topography of the MSERE site indicates that flooding is highly unlikely. 1S. E. Beall, W. L. Breazeale, and B. W. Kinyon, "Molten-Salt Reactor Experiment Preliminary Hazards Report, " USAEC Report ORNL CF-61-2-46, Oak Ridge National Laboratory, February 1961. 240 8.4 Sabotage Severe damage to the reactor by sabotage would be difficult; arson is probably the best possibility. The consequences of fire are minimized by fireproof structures, a sprinkler system, and an alarm system that auto- matically transmits to the fire-protection headquarters at X-10. Water for this system is supplied to the building sprinklers by the main line from X-10 and two 1.5 million-gallon reservoirs near the site. Only a person with intimate knowledge of the reactor would be capable of inflict- ing damage that might result in reactor hazards. This possibility is minimized by adequate personnel policiles and security regulations, particu- larly with respect to visitors. 8.5 Corrcsion from Spilled Salt Tn an accident invelving contact ¢f water and flucoride salts in the containment vessel, fairly rapid generation of hydrofluoric acid is ex- pected. Part of the I will be dispersed as vapor in the cell and part of it will dissclve in the water. Corrosion damage resuliing from dis- solved IF is gre=tly deperden® upcn the temperature and the concentra- 4 hr, 262 With the fan on, k¥ = 1.875 hr=!, and for t = 0 to 4 hr, c = a [0.0002405(1 — "1.8750) —-o.oooo534§] For iodine, ¢ = 30.2(L — e"1-8750 _ g 2205t) . For noble gases, ¢ = 90.6(1 — e 1-875% _ g 2205t) . For solids, ¢ = 164.2(1 — e"1-875V _ 5 2205¢) | With the fan off, k = 0.004 hr™*, and for t = O to 4 hr, 1l c = a [5.35(1 ~ 70.004Ty _ 0.025€] For iodine, c = 7.95 % 10° (1L — e70-004% _ 3 937 x 1073¢) For noble gases, 2.385 x 10° (1 — e70:004t _ 3 937 « 1073¢) 0 1l For solids, 0o lh 4.32 x 10% (1 — e70.004% _ 3 937 x 1073¢) . Data on the activity in the building are presented in Tables A.1 and A.2. Calculation of Exposures and Building Escape Times The internal exposure was determined from the following expression: 108 [ E == j; ¢ at Table A.1l. Cell leak rate: Fan discharge rate: lodine released in cell: Noble gases released in cell: Solids released in cell: Activity in Building with Fan On 1% per day or 0.0004 per hr 15,000 £t?/min or 1.875 building volumes per hr 2.5 % 10% x 10% x 50% = 1.25 x 10° curies 3.75 % 10° curies Solids reledsed through filters: 6.8 X 10° x 10% = 6.8 x 10° curies 0.1% of solids to filters . Fraction of Todine Noble Gases Solids Time . . After AcbIvity Re- . . . Accident leased from Release Activity in Release Activity in Release Activity in (hr) Building Rate Building Rate Building Rate Building per Day (curies/day) (curies) (curies/day) (curies) (curies/day) (curies) 0.25 3.5 x 103 436 9.7 1308 29.1 2.4 53 0.50 5.4 x 1073 675 15 2025 45 3.7 82 0.75 6.4 x 1073 800 17.8 2400 53.4 e & 97 1.0 6.8 x 1073 855 19.0 2565 57 4.6 103 1.1 7.0 x 1073 874 19.4 2622 58.2 4.8 105 1.2 6.8 x 1073 855 19.0 2565 57 .6 103 1.5 6.6 x 1073 828 18.4 2484 55,2 4.5 100 2.0 5.8 x 1072 729 16.2 2187 48.6 3.9 88 2.5 4,75 x 1073 594 13.2 1782 39.6 3.2 72 3.0 3.6 x 1073 455 10.1 1365 30.3 2.5 55 3.5 1.4 x 1073 176 6.8 528 20.4 1.0 37 4.0 1.3 x 10-3 160 3.55 480 10.65 0.9 19.3 5.0 1.9 x 1074 24,1 0.535 72 1.60 0.13 2.9 6.0 3.0 x 1072 3.8 0.0845 11.4 0.24 0.02 0.46 7.0 4.7 x 107 0.6 0.0130 1.8 0.039 0.003 0.071 8.0 7.2 x 1077 0.1 0.0020 0.3 0.006 0.0005 0.011 £9¢ Table A.2. Activity in Building with Fan Off Cell leak rate: 1% per day or 0.0004 per hr Building leak rate: 10% per day or 0.004 per hr Todine released in cell: 2.5 x 10% X 10% X 50% = 1.25 x 10° curies Noble gases released in cell: 3.75 x 10° curies Solids released in cell: 6.8 x 10% x 10% = 6.8 x 10° curies ) Fraction of Todine Noble Gases Solids Time e After Activity Re- . . .o . R Accident leased from Relcasc Activity in Release Activity in Release Activity in (hr) Building Rate Bullding Rate Building Rate Building i per Day (curies/day) (curies) (curies/day) (curies) (curies/day) (curies) 0.25 9.7 % 107° 1.21 12.1 3.63 36.3 6.58 65.8 0.50 1.86 x 1072 2.33 23.3 7.0 70 12,7 127 0.75 2.72 x 1072 3.39 33.9 10.2 102 18.4 184 1.0 3.5 x 1077 4 .37 43,7 13.1 131 23.8 238 1.5 4.9 x 10-? 6.07 60.7 18,2 182 33.0 330 2.0 6.0 x 107° 7. 48 . & 22 .4 224 40.6 406 2.5 6.8 x 1077 8. 54 85.4 25.6 256 46 . 4 46l 3.0 7.4 x 1072 9.30 93.0 27.9 279 50.5 505 3.5 7.6 x 1077 9.49 94.9 28.5 285 51.6 516 4.0 7.9 X 1077 9.88 98.8 29.6 296 53,7 537 5.0 7.86 x 1072 9. 8% 98 .4 29.5 295 53.5 535 8.0 7.8 x 1072 9,71 97.1 29.1 291 52.8 528 10.0 7.7 x 1073 9.63 96,3 28.9 289 52.% 524 16.0 7.5 x 1072 9.40 9%, 0 28.2 282 51.1 511 24,0 7.25 x 1077 9.08 90.8 27.2 272 49.4 494, 79¢ 265 where E = internal exposure in uc, ¢ = activity in building in curies, V = building volume in m° (13.6 x 10°), B = breathing rate in r’/hr (1.8), t = time in hr. The exposures and escape times are given below for the noble gases, iodine, and fission-product solids. Noble Gases The noble gas escape time 1s based on an external exposure of 25 r. Using the formula r/hr = 935 X pc/em® X Mev (assume = 1), the radiation level is calculated and the exposure 1s obtained by integration. With the fan on, k¥ = 1.875, and the concentration of noble gases in the build- ing in pc/cm? is 20.6 —————— (1 - 728750 g 2018) 13.6 x 10° The radiation level in r/hr is 6.24(1 — ¢™1-8750 _ g,221t) . Integrating the radiation level with respect to time, the exposure in roentgens is e~1.875% t 6.24 |t + Té,.?T— - O.llO5t2 0 s and the exposure at t hours after the accident in roentgens is e-1.875% _ 4 6.24 [% + TR - 0. llO5t2:| With the fan off, k = 0,004, and the concentration of noble gases in the building in pc/cm’ is 260 2.385 X 10° 13.6 x 10° (L — ¢-0.004% . 0,003937%) The radiation level in r/hr is 1.64 x 10° (1 — ¢70-004% — 0,003937t) Integrating the radiation level with respect to time, the exXposure in roentgens 1s t - 0.0019685t5] , 0 e~0. 004t . 5 1. 1.64 x 10 [J + 5 00% and the exposure at t hcurs after the accident in roentgens is 5 . e-0.00é'fiZ -1 5 1.64 » 10 T+ ————-,T-fi'éz——— — 0.0019685t Todine The exposure per inhaled mwicrccurie of lodine is given in Teble A.3.Y Based on Table A.3, the permissible tolal iodine exposure, that ig, the exposure for a 300-rem dose, is 300,000 312.¢6 = 960 pe With the fan on, t 132 x 30.2 j; (1 — e~1-875% _ §.2205t) dt : e-—-l.875t -1 2 = 3980 |t + ——1——5-75——— - 0.11% il Exposure 7, J. Burnctt, "Reactors, Hazard vs Power Level," Nucl. Sci. Fng., 2: 382 {August 1956). 267 Table A.3. Iodine Exposures Ratioc of Activity Exposure Dose per Todine of Isotope to per Unit of Tsotope Total Iodine Unit of Total Iodine Activity Activity Activity (L-hr decay) (mrem/uc) (mrem/ue ) pisl 0.115 1484 170.8 I3 0.170 53 9.0 Iti23 0.249 399 99.4 i34 0.244 25 6.1 7133 0.222 123 27.3 Total 312.6 With the fan off, t Exposure = 132 X 7.95 x 10° J; (1 — e~0.004T _ 3 937 x 10734%) at -0,004t = 8 e ~ _—31_ 2 = 1.05 X 10 (t + 500 0.0019685t ) . Sclids The maximum permissible inhalation intake (MPI;) of solids for a 25-rem bone exposure is 221 uc based on an exposure of 113 mrem per in- haled microcurie of fuel irradiated for one full-power year.1:2 With the fan on, t Pxposure = 132 x 164.2 [ (1 —e™>-875% _ 0,2205¢) at 0 e-la 875t — l 1.875 | 2.17 x 10% (t + ~ O.llt2> 2Third Draft, "American Standard for Radiation Protection at Reactor Facilities," by ASA Subcommittee N-7.5 (June 1963). 268 With the fan off, t Exposure = 132 X 4.32 x 10° j; (1 — e70-004t _ 3 937 x 1072%) dt i 6“0.004—t -1 5.71 X 108 ("G + W — 1,97 X lO—Btz) Calculation of Effect of Meteorological Conditions on Diffusion of Activity Calculations of the diffusion of activity from the stack were made using Sutton's formula for diffusion from a continuous point source:>s % 2 x:%em(_m)} #C_C pxt” cex®7H Yz Q = emission rate, curies/sec, = crosswind diffusion coefficient, (meters)?/2, C, = vertical diffusion coefficient, (meters)nlz, U = wind speed, meters/sec, ¥ = downwind distance, meters, y = crosswind distance, meters, n = height of plume, meters, n = stability parameter, = activity concentration, pc/cm’, for unit emission. The following parameter values were used as recommended by the U.S. Weather Buresau, Oak Ridge office: 37U.S. Department of Commerce, Weather Bureau, 'Meteorology and Atomic Fnergy," USAEC Report ARCU-3066, July 1955. 7. J. Di Nunno et al., 'Calculation of Distance Factors for Power and Test Reactor Sites,” p. 5, USARC Report TID-14844, March 23, 1962. 269 Value Under Lapse Value Under Inversion Parameter (Weak) Conditions Conditions o 2.3 1.5 n 0.23 0.35 Cy 0.3 0.3 Cy 0.3 0.033 All calculations were made assuming y = O; that is, the dispersion was directly downwind of the source. Activity levels from building seep- age were made assuming h = 0. The values obtained for X at various dis- tances for both inversion and normal conditions are listed in Table A.4. Table A.4. Diffusion Pactor, X, Versus Distance Downwind %, Diffusion Factor® Distance Downwind Inversion Conditions Normal Conditions (m) Ground Release Stack Relesgse Ground Relesase Stack Release 100 2.1 x 1072 <10~10 8.0 x 107 2.9 x 1077 160 9.8 x 10~3 <10710 3.5 x 107% 1.1 x 1077 200 6.3 x 1073 <10710 2.3 x 1074 2.3 x 1077 325 3.0 x 1073 <10-19 1.0 x 1074 3.7 x 107° 500 1.5 x 1073 2.1 x 1077 4.7 x 1077 3.0 x 107° 750 7.7 x 1074 3.2 x 107° 2.3 x 1077 1.8 x 1072 1,000 4.8 x 1074 3.5 x 1077 1.4 x 1077 1.2 x 107? 1,200 3.5 % 1074 4.5 % 1077 1.0 x 1077 9.1 x 107° 2,000 1.5 x 1074 6.2 x 1077 4.0 % 107° 3.8 x 107° 3,000 7.7 x 107 4.9 x 1077 2.0 x 107° 2.0 X 107° 4,000 4.8 X 107° 3.6 x 1077 1.2 x 107° 1.2 x 107° 5,000 3.3 x 107° 2.7 x 1077 8.0 x 1077 8.0 x 1077 10, 000 1.1 % 107° 1.0 x 1076 2.3 x 1077 2.3 x 1077 % X release rate (curies/sec) = activity concentration in air (pe/ce). Calculation of Dose Rates Outside Building Dose Rate from Building The building is assumed To be a point source with radiation of 1l-Mev average energy E. The dose rate R (in r/hr) at 1 ft is given by the 270 formula’ R = 6CE , were ¢ 1s activity concentration in uc/cm?. With the fan on the maximum activity in the building is 183 curies at 1.1 hr. With the fan off the maximum activity in the building i1s 932 curies at 4 hr after the maximum credible accident (see Tables A.l and A.2). The radiation level is as- sumed to vary inversely with the square of the distance. Dose Rate from Radicactive Cloud The maximum release rate with the fan on is 3501 curies per day at 1.1 hr after the maximum credible accident (see Tables A.l and A.2). With the fan off the maximum release rate is 93.2 curies per day at 4 hr after the maximum credible accident. The dose rate R (in r/hr) is calculated from R = 935 cE . The release rate in curies per second multipiied by the diffusion factor, X, from Table A.4 is the activity concentration in uc/cm?. Calculation of Total Integrated Dose The total integrated thyroid snd bone doses are calculated for stack 3 4 release” and for building seepsage. Total Activity Released For stack release with the fan on the activity released from the building in the first 4 hr is t j‘ (curies in building at © hours) X (release rate per hour) . 0 Radiation Safety and Control Manual, p. 3-3, Ozk Ridge National Laboratory, June 1, 196l. 271 For iodine (see p. 266) the release in 4 hr is 4 | 30.2 [ (1= e 8750 ~ 0.2205¢) at X 1.875 = 96.5 curies After 4 hr, no additional iodine leasks into the building from the cell, and the 3.55 curies in the building at 4 hr (Table A.1l) is the only ad- ditional release after 4 hr. The total iodine release with the fan on is therefore 100 curies. For solids the release in 4 hr is 4 164.2 j; (2 — e=1-875% 0 2205t) at X 1.875 x 0.001L (99.9% retained on filters) = 0.53 curies s to which 0.1% of the 19.3 curies in the building at 4 hr is added to give a total solids release of 0.55 curies. For release by building seepage with the fan off the activities are calculated in the same manner. Tor iodine the release in the first 4 hr is & 7.95 X 10° J; (1 — e70-004% _ 3 937 % 1073%) at x 0.004 = 1.06 curies From Table A.2 it 1s seen that 98.8 curies of iodine are in the building at 4 hr, and it 1s assumed that this is all. released, although a large vrercentage will undoubtedly plate out on the building walls. The total iodine release is therefore ~100 curies. For solids the release in the first 4 hr is t 4.32 x 10® x 0.004 [ (1 —e70:004% _ 3,937 x 1072¢) dt = 5.43 curies . 0 From Table A.2 it is seen that 537 curies of solids are in the building at 4 hr, and therefore the total release is 543 curies. Maximum Permissible Doses Thyroid: 300 rem (ref. 4); 1484 mrem per microcurie of I131, Bone: 25 rem (ref. 4); 113 mrem per microcurie of solids. 272 These doses are based on an irradiation time of 365 days, using re- vised exposure values representing current estimates of dose per inhaled component microcurie. 1,2 Stack Release For For For For For 2@ X breathing rate TIDmax T X e X wind speed X stack height iodine, -4 2 X 100 x 5 x 10 = 2.04 x 107° curies 3.14 X 2.72 X 2.3 x (50)? normal atmospheric conditions, 2,04 X 107¢ X 1.484 X 10° = 3 rem . inversion conditions, 3 x 242 = 4.6 rem 1.5 solids, ) Q=% 2 x 0.55 X 5 X 10 = 1.12 x 1078 curies 3.14 x 2,72 x 2.3 X (50)° normal atmospheric conditions, 1.12 x 1078 x 0.113 x 102 = 1.3 mrenm . inversion conditions, 1.3 X 24% = 2 mrem . i_J 273 Building Seepage Q X breathing rate T X wind speed X oy X Oz Inhaled curies = Q = total curies released from building Breathing rate = 5 x 10~% m?/sec Wind speed (normal) = 2.3 m/sec Wind speed (inversion) = 1.5 m/sec 1 o, (vertical concentration deviation) = — C d(l—n)/z 2 | L (1-n)/2 o, (horizontal concentration deviation) = — C,d V2 Cy (vertical diffusion coefficient) = 0.3 normal = 0.3 inversion C, (horizontal diffusion coefficient) = 0.3 normal 0.033 inversion Table A.5., Doses Resulting from Building Seepage Quantity of Activity and Dose Distance Lrom Activity Normal Conditions Inversion Conditions Building (m) Le rem e rem 1.00 Todine L 6 66.3 1075 1600 100 Solids 242 27 .4 5840 660 200 Iodine 13.2 19.5 348 517 200 Solids 13.8 1.56 405 45.8 1000 Iodine 0.75 1.12 24,1 35.8 1000 Solids 4,07 0.46 131 14.8 | [ - Uow o \, i P & | i i i, ! | 1 I J i ; | [ by i / t . o s \ e e e A A A S b AR R+ i e et e Flow Sheet‘; No. - D-AA-A-40880 40881 . 40882 40883 40884 40885 40887 40888 40889 275 Appendix B PROCESS FLOWSHEETS Procéés flow sheets follow for the systems listed below: 7 System . Fuel System Coolant System | Fuel- Drain Tank System Off Gas System and Containment - Ventilation , Cover Gas SyStem 011 Systems for Fuel and‘CQQl- ant Pumps o Fuel-Précessing System Liquid Waste System Cooling-Water System O Unclassified ORN'L-DJG 63=7'752 276 | ' - | | 1 L MOVJUIAD T 'ON WISNIQNOD dNYL Nivid 1'0ON HISNIONGD NNVL NIVNG TR SO S BLOWER HOUSE T30 NIVUO LNVI00D MOOW LNIMJINDI IMI3dS T T SR S NILEAS NQ dnnd LNYI00D N JANe "TREATED WATER COOLER 2 g m‘ o ] TO LIQUID WASTE el L TO WPOR CONDENSN AT NORTH. ENO OF BLDG. USED) COOLING TOWER PUMP NO.1. EXISTING TREATED WATER PUMP NO. 2. WATER ROOM PROCESSING SYSTEM GAS SYSTEW & COMTAINMENT VENTLATION SYSTEM INSTR, APPLICATION DMGRAM SYSTEM FOR FUEL & COOLANT PuwPs £, DRAN TANK SYSTEM SNINNKY TT3D §012VIM NONS 1733 w0LIV3IE 0L TYRNIHL HOLOVIE MOM4 dNnd 13NS KoM N3LSAS ONISNIONOD HOUIYA OL T 0N 837000 biv M3 ¥OLOVIN NO¥F 10N ¥3T000 uIV T13D ¥OLOViIN NOM3 HiV TT3D NNVL NIVYa WOUd R S— — M0 _IEen1 — dMNd LHYIO0D - VS 04 SWIW00D dRfid LNVYI00D £V9 804 ¥IW0D NOOY ININDINDI E e E— WNVL F1SYM 01 SATINKNY LNINNBISHI HYIIINN OL YI00D HIV Y12 WNVL NivEd 01 HOLUW dnind INVIOLD NGHS HOLON dNNd ANVI0OD OL ON H3100D di¥ T30 HOLOVIM OL Ol 837000 sy 3D BOLOVAN OL dind 13ns OL HS TTYNH3HL H0LOV3H OL TREATED WATER PUMP DATA COOLING TOWER PUMP DATA PROCESS FLOW SHEET O-AA-A-40800-8 JOB 433-1,0 277 Unclassified‘ ORNL~-IMG 63-7751 , FLTER nr{ SERVICE TUNNEL ’ . VENTILATION STACK — e !2)——_. ’ - “ln puct fLOOR"‘"‘”@"—-—“_—- RADIATOR STACK COOLANT DRAIN CE |- (352~ (3550 SEMVICE ROOM (£358) ' SERVICE TUNNEL—(D59) - V- 21 _3-40-$~ _~4"C.}, ~3-40-3 _ _viss & 8- 40-$ =~y I VENT fi“j CHARCOAL | l BED CELL I CATCH BASIN 8-40-3 REACTOR CELL [ ANNULLS . - l ACCESS _DOOR ‘ TO CREEX FROM REACTOR AND . L . J "DRAIN TANK CELL PR )——— : 3 aEaEe——" - , ] vize 29 @ . 1 CV32e ' ‘ { CAUSTIC ; | _EUMP ROOM - (® ADDITION FUNNEL @, . cvize ¥/ i : 3-40-3 FROM MAINTENANCE ~(329) - | SHOP WOT DNMIN ) e cpw 4-40-3 W d-2-40-23 f i @ ‘ l ‘ — FROM HOT smxs.'k cvaze ! s . dr2-a0-88 : ) . | vize 24055 : " : - o c 8 - e (:‘ 3-40- | PIT_PUMP | ‘ Ld . W 4 | | v's wo*am: - "”'..—'l 8 — 1 ls DRATN TANK Joj-e04 _ p LIQUD WASTE STORAGE TAMK wT TANK JET | {FUEL PROCESSING cviee I §-40-93— 1 ’ i 302 STAINLESS STEEL 3 I8 qearaciTy 1,000 GAL, ] SLECTRICAL seEAvICE AREA VPos cvens CELL SUMP N *..;m OECONTAMINATION CELL DECONTAMINATION TANK 0 | 2-40-88 (B e :i—oo-ss—-\' LIQUID WASTE CELL “Hlry— L D ' SYETER TANK SYSTEM. PUEL PROCESIMS SYSTEM ICOOLING WATER SVYSTEM LIGUID WISTE SYSTEM INST. APPLICATION DiA. [D-A-B-40308 OFF A% SYSTEM B CONTARMENT VeNT L & . e R OM DR NATYORAL LASDRATORY LIQUID WASTE SYSTEM PROCESS FLOW SHEET M.S.R.E. s0m 43340 | |o-ar-a for steel, C, = specific heat of solid, 0.12 Btu/lb:°F, an estimate was made of the amount of vessel heating that would take place during the time salt was spilling and the vapor-condensing system was 1in action. Then, taking t. as the vessel temperature before the accident (150°F) and t’/ as the mzximum steam temperature during the event (262°F), the vessel temperature (ta} at the end of the accident (i.e., 6 = 300 sec) was found. For the value X = 1.32, obtained by solving Eq. (3), the value of Y was found from the X-Y plot of Fig. 10 of ref. 2 to be 0.7. Based on these values, Eq. (1) was solved and t, was found to be about 180°F at the beginning of the cool-down of the cell atmosphere. As an approximation of the unsteady state of the steam temperature during cooling, Y was calculated at 5°F intervals of wall temperature in- crease., For these calculations, the value of X was taken from Fig. 10 of ref. 2. The heat content per pound of steam remaining in the cell was then determined and the equivalent pressure was read from the Steam Tables. The temperature and pressure of the cell obtained for each interval are given in Table E.1. Although the input of heat from fission-product decay is not taken into account for this short period, it becomes significant (approximately 10% Btu) over the 4-hr period assumed in the release calculation. However, with the salt on the floor of the cell and in contact with the steel, which 297 Table B.1. Cell Pressure at Various Times After the Maximum Credible Accident Temperature Interval (°F) Time Cell cell . Presgure Temperature ——————— (min) : o (psia) (°F) t T b a 180 180 0 37 262 180 185 0.8 32 254 185 190 1.5 28 247 120 195 2.3 24 238 195 200 3.4 20 228 200 205 5.6 16 215 205 210 14,4 14 210 is submerged in water,’ a modest heat flux of 10,000 Btu/hr.ft? would re- move the heat thfough the bottom of the tank. (In the experiment involv- ing the release of 500 1b of salt into a dished vessel, cited in Section 8.6, heat fluxes greater than 100,000 Btu/hr.ft? were found.) The 4.4 ft3/sec (max) steam generated under the hemisphere would be vented through the &-in. pipe installed for this purpose. *Memorandum from L, F. Parsly to E. S. Bettis, Jan. 23, 1962, "Con- sequences of a Salt Spill into the Bottom of the MSRE Containment Vessel." 299 ORNL-TM-732 Internal Distribution 1. R. K. Adams 4%7. T. L. Hudson 2. R. G. Affel 48. R, J. Kedl 3. G. W. Allin 49, 8. S. Kirslis 4. A. H. Anderson 50. D. J. Knowles 5. R. F. Apple 51. A. I. Kraxoviak 6. C. F. Baes 52. J. W. Krewson 7. 8. J. Ball 53. C. E. Larson g. S. E. Beall 54. R. B. Lindauer 9. M. Bender 55. M. I. Lundin 10. E. S. Bettis 56. R. N. Lyon 11. F. L. Blankenship 57. H. G. MacPherson 12. R. Blumberg 58. C. D. Martin 13. A. L. Boch 59. H. C. McCurdy i4. E. G. Bohlmann 60. W. B. McDonald 15. C. J. Borkowski 61. H. F. McDuffie 16. H. R. Brashear 62. C. K. MceGlothlan 17. R. B. Briggs 63. H. J. Metz 18. ¢. H. Burger 64. A. J. Miller 19. J. A. Conlin 65. W. R. Mixon 20. W. H. Cook 66. R. L. Moore 1. L. T. Corbin 67. H. R. Payne 22. W. B. Cottrell 68. A. M. Perry 23. J. L. Crowley 69. H. B. Piper 24. D. G. Davis 70. B. E. Prince 25. G. Dirian 71. J. L. Redford 26. 8. J. Ditto 72. M. Richardson 27. I. A. Doss 73. H. C. Roller 28. J. R. Engel 4. M. W. Rosenthal 29. E. P. Epler 75. T. i. Row 30. A. P. Fraas 76. H. W. Savage 31. E. N. Fray 77. A. W. Savolainen 32. H. A. Friedman 78. D. Scott, Jr. 33. C. H. Gabbard 7. J. H. Shaffer 34, M. J. Gaitanis 80. E. G. Silver 35. R. B. Gallaher g81. M. J. Skinner 36. J. J. Geist g2. T. F. Sliski 37. W. R. Grimes 83. A. N. Smith 38. A. G. Grindell 84. P. G. Smith 39. R. H. Guymon 85. I. Spilewak 40. 8. H. Hanauer 86. R. C. Steffy 41. P. H. Harley &7. H. H. Stone 42. P. N. Haubenreich 88. H. J. Stripling 43. G. M. Hebert 83. J. A. Swartout 44, P. G. Herndon 20. A. Taboada 45. V. D. Holt 91. J. R. Tallackson 46. A, Houtzeel 92. R. E. Thoms 111-170. 171-172. 173-187. H. MWW EaROQ . M. H. Q== I\"i . 300 Tolson 100. G. D. Whitman . Trauger 101. H. D. Wilils . Ulrich 102-104. Central Research Library Webster 105-106. Y-12 Document Reference Secction . Weinterg 1207-10%9. Laboratory Records Department . West 110. Labcratory Records, RC . White Ixternal Distribution Roth, Research and Develcpment Division, CRO Reactor Division, CRO Division of Techniczl Information Extension, DTIE