EIR-Bericht Nr. 411 EIR-Bericht Nr. 411 Eidg. Institut fur Reaktorforschung Wurenlingen Schweiz Reactor with very low fission product inventory M. Taube, W. Heer Sy Wirenlingen, Juli 1980 Reactor with very low fission product inventory M. Taube, W. Heer, in co - operation with A. Indrefjord Swiss Federal Institute of Reactor Technology July 1980, Wurenlingen FEIR - Report 411 Abstract. A fast converter with one zone and an infernal breeding ratio of 1.00, with liquid fuel in the form of molten plutonium - - uranium - and sodium chloride, with a thermal power of 3 GW (th) allows continuous extraction of the veolatile fission products (Br, I, Kr, Xe, Te) by means of helium purging in the core. The non-volatile fission products e.g. Sr and Cs can continuously be extracted in a chemical reprocessing plant at the reactor site. The impact on an accidental release of fission products is rather significant; the amounts released are 50-100 times smaller than those in a reference reactor (LWR with oxide fuel). Because the heat sink 1s relatively large and after heat reduced, the temperature of the fuel does not exceed BOOOC after an accident, which greatly reduces the consequences of an accident. Zusammenfassung. Ein schneller konverter Reaktor mit einer Zone und einem Brutfaktor von 1.00, mit flissigem Brennstoff{ in Form von geschmolzenem Plutonium -, Uranium - und Natriumchlorid, mit elner Leistung von % GW (th) erlaubt die kontinulerliche Enfernung von fliichtigen Spaltprodukten (Br, I, Kr, Xe, Te) durch Durchleiten von Helium im Core. Die nicht-flichtigen Spaltprodukte wlie Sr und Cs werden kontinuierlich in der angegliederten Aufbereitungs- anlage ausgeschieden. Die Folgen bei einer unfallbedingten Freigabe von Spaltprodukten sind recht signifikant; die frelgegebenen Mengeh sind 50-100 mal kleiner als die beim Referenzreaktor (einem LWR mit Oxiden als Brennstoff). Well dile Wirmesenke relativ gross 1st, und die Nachwidrme reduziert ist, Ubersteigt die Temperatur des Brennstoffs nach einem Unfall 500°C nicht, was zu einer bedeutenden Reduktion der Unfallfolgen Tfihrt. INIS DESCRIPTORS MOLTEN SALT REACTORS REACTOR SAFETY FISSION PRODUCT RELEASE LOSS OF COOLAN AFTER HEAT REMOVAL MELT DOWN Contents. 1. Introduction 1.1 Present position 1.2 Safety of Fission Reactor: State of the art 1.3 Saflety of Fission Reactor+ desired improvement 1.4 The present State of Reactor Development and the Criteria for a Safer Reactor. Accidental release of fission products and decay heat removed. 2.1 The Rasmussen Scenario 2.2 The problem Principles of the "SOFT" Reactor 3.1 A Schematic View 3.2 System description 3.3 Mass flows and containments Nuclear calculations 4.1 The Method 4.2 Nuclear properties of SOFT Continuous extraction of fissilion products. 5.1 General scheme 5.2Chemical state of fission products in molten chlorides 5.% Fission product volatility and gaseous extraction 5.4 Problem of delayed neutron emitters 5.5 Gas extraction rate Page 5.6 Reprocessing of non-volatile fission products 5.7 Some technological problenms 5.8 Possible reprocessing technologies 5.9 The problem of external storage of the fission products Heat removal 6.1 Primary circuit: core and heat exchangers 6.2 Secondary circult 6.3 The tertliary circuilt 6.4 Heat removal from the sotred fission products Accidents - problems and solutions 7.1 Decay heat: Power and energy 7.2 Worst accident scenario 7.3 Spontaneous cooling processes in the containment 7.4 The prompt critical scenario Conclusions Literature Page 4o L 44 47 b 4T L3 48 51 51 51 55 61 63 o7 68 1. INTRODUCTION 1.1 Present position The llight water reactor, the most common power reactor today and for the next two decades 1s safe enough to be the basic energy source for society. This however does not mean that the search for improved safety 1s a waste of time. The continued search for still safer systems 1is common in all the technologlies making up our civilization (transportation, chemical technology, domestic fire harzards etc.). Fach type of nuclear reactor also has potential for increasing its safe operation. It must however be remembered that the most important safety aspect of the current reactor sysfems 1is in the real practical experience bullt up over many years. Here the light water reactor is in a privileged position being able to demonstrate an excellent safety record. Tn spite of this crucial fact the search nust continue for reac- types which promise improved inherent safety on the baslis of their different system design. Such a search seems to be gene- rally desirable. The best reason for such a reactor type has been given by Alvin Weinberg (1979, 1980). "For the 15 billion curies contained in a 1000 megawatt reactor we could never say that the chance of a serious accidental releasec was zero. Thus a most important technical fix for nuclear energy would be a means of minimising the amount of land that conceivably could be contaminated in the worst possible accident. For nuclear energy to survive we must reduce the probability of any serious malfunction much below the 1 in 20 000 per reactor-year estimated in the Rasmussen report as well as reducing any possible consequences." "But 1f the world energy system involved as many as 5000 reactors - that 1s, 10 times as many as are now either in operation or under construction - one might expect an accident that released sizeable amounts of radiocactivity every four years. Considering that a nuclear accident anywhere 1s a nuclear accident everywhere I belileve this accident probability is unacceptable. If a man is to live with fission over the long term he must reduce the a priori probability of accident by a large factor - say 100." (The Bulletin of Atomic Scientists. March 1980) The aim of this paper 1s to try and present a concept for a fission power reactor being approximately 100 times less hazardous than existing reactor types, and with much less than 15 billion curies activity. Tt must be stressed that this study 1s concerned with a 'paper reactor' (which is of course the safest type of alll) a concept only. The route from the concept to reallsing an actual power reactor in service, and contributing to our energy problems 1s a long and tortuous road and with financial commitments of tens of billions of dollars. Even this should not prevent the search for such a reactor. The search 1tself is of benefit even for a better understanding and possible improvement to the existing power reactor systems. How is 1t possible to meet the regquirement for a 100-fold improvement in reactfor safety? The following proposals are made below: a) the amount of fission products in the core during a normal operation must be reduced by a factor 100 which significantly reduces the effects of a large accident where the fission products would be released into the environment, b) the decay heat level must be reduced and the internal inherent heat sink, such as the total heat capacity of the fuel and other components must be significantly increased. Additionally the maximum temperature reached by the fuel 1n the case of failure of all emergency ccoling systems must be low enough to allow the fuel to be trapped in a core catcher, Table 1.1 Characteristics 1) 3) Fission product inventory Decay heat removal Pressure in fuel and coolant (in containment) Coolant with low boiling point (Explosion possibility) Reactor characteristics affecting safety (data for 3 GW(th) ) Existing Reactor (e.g. LWR) ~15 Geurie. Possibility of releasing into the environment: a) volatile Fission Products (F.P.) b} non-volatile F, P. Immediately after shutdown “180 MW(th). The integrated decay heat over some hours is of the order of 1000 Gigajoules ajyvl50 bar in a PWR ~80 bar in a BWR in the coolant. b) internal pressure in the fuel pins The presence cof water in the primary circuit gives rise to the hazard of uncontrolled boiling and production of large volumes of steam, pressurizing the contalnment Desired Reactor Minimum in F.P. inventory but at least two orders of magnitude smaller than in existing reactors for both: a) volatile F.P. b) non-volatile F,P. Minimum decay heat. At least by more than one order of magnitude No pressure: 2) in the coolant b) in the fuel No low boiling point media allowed in the reactor circuits Characteristics Chemically active medium (coclant} Hydrogen Evolution 'China Syndrom! Criticality control External (away from reactor) movement of plutonium Existing Reactor {(e.g. LWR) In a liquid metal cooled reactor an exothermic reaction is possible 4 Na + UO_+2Na,_.0+U ™ 2 2 m et et In Light Water Reactors during an accident the following reaction may occur s Zrmet+H2O+ZrO+H2gaS The evolved hydrogen results in an uncontrolled increase of pressure in the core vessel or containment because of burning in air For the LWR system the report WASH-1400 discussed the probability of the'China-syndrom' A strongly negative criticality,for the case of loss of coolant control rods The LWE system being a producer of plutonium results in transporting plutonium away from reactor (e.g. in irradiated fuel) Desired Reactor No chemically active agent is allowed Ne¢ chemical agent including hydrogen is allowed Elimination of this kind of accident A self regulating system is desired A self sufficient reactor with a breeding ratio of ~1 has no external circulatiocn of Pu Both of fthese criteria seem to be possible in a reactor having a molten fuel with continuous extraction of fission products in the fuel during normal operation. (Ref . 19). In this paper one sclution to this problem 1s discussed in detail: the molten salt reactor. It must be said that such reactor types have been consldered for many years having molten salt as a fuel and with continuous extraction of fission products. Some examples of molten salt reactors: (Ref. 5, 14, 16, 17, 18). a) thermal breeder with molten fluorides developed decades ago by Oak Ridge National Laboratory, existing as an experimental reactor with a power of 8 MW(th), (Ref. 13) b} fast breeder reactor with molten chlorides discussed for tens of years, but still a "paper reactor”. A reactor of the second type is discussed here. 1.2 Safety of a Fission Reactor: state ot the art. Of course the hazard of a fission reactor is not only connected with 1ts fission product inventory. There are other important safety aspects. Table 1.1 shows the 1mpact of other reactor parameters on its safety and ways of resolving these. The question arising out of these criteria 1s can all the necessary improvements be made in a single reactor type? The answer assumed here 1s yes. 1.5% Safetv of a Fission Reactor+ possible improvements. Table 1.1 summarises very briefly the properties of such a 'super safe' reactor. The reactcr proposed in this paper is called SOFT. alt reactor rn site reprocessing ast converter Floo O e ask The properties are discussed below in the order given in table 1.1. Table 1.2 10 3afety of a Fission Reactor: possible improvements Parameter 1) Fission product inventory 2) Decay heat removal %) Pressure 4) Low boiling point media (Explosion possibility) Properties of the desired reactor Minimum F.P. Two orders of magnitude less volatile and non-~ volatile F.P. Decay heat at least one order of magnitude less Works at ambient presssure a) in coolant b) in fuel Absernce of agents having low boiling point reducing also the explosion hazard The solution formed for the SOFT reactor Continuous extractlon of: ajvolatile F.P. by means of gas pumping blnon-volatile F.P. by means of chemical ftreatment and storage in another contalnment Decrease of decay heat energy by: a)continuous extraction of F.P. b) continuous extraction oszng and storage in another containment Use of molten salt as both: a) the coolant b) the fuel in a pressureless system Both fuel and coolant have a bolling temperature of approx lBOOOC thus no explosion is possible Parameter 5) Chemically active media 6) Hydrcgen evolution 7} 'China Syndrome! 8) Criticality control 11 Properties of the desired reactor Absence of chemically active media which could react with fuel e.g. sodium and oxide Absence of substances which could give rise to hydrogen. E.g. water reacting with Zirconium Elimination of any formation of a mass of molten fuel, slumping through containment Fuel inherently 'self contrclling' The solution formed for the SOFT reactor Both fuel and coolant are in the form cf molten chlorides. The thermo- dynamic stability of all constituents exclude any exothermic chemical (or explosive) reaction Neither fuel nor coolant contain hydrogen The significant decrease in decay heat and the presence of large heat sinks (reflector, core catcher) eliminates a 'China Syndrome' Fluid fuel reactors are known to have a large negative temperature cocefficient and Doppler ccoefficient 12 1.4 The present State of Reactor Development and the Criteria for a Safer Reactor The present state of reactor technology results from a 1long and complicated development and marked by not always the most logical decisions. If the following are the four main targets of reactor development: 2) an economically and technically feaslible source of electrical energy, b) economically and technically feasible source of high temperature heat for chemical processes, ¢) high grade utilisation of uranium resources (and even thorium resources), d) satisfactory level of safety, then it may be possible to show with the aid of a diagram (Fig. 1.1) the present state of reactor development. Of course there are other ways of eliminating the characteristics presenting the greatest hazards - high pressure and high F.P. inventory. Figure 2 shows some possibilities. It seems however that only molten salts can solve both these problems. In this paper the accent is on the safety problem 1.e. the effort to reduce the consequences of the worst credible accident - core melting. Paradoxically as this might sound the solutlion to this problem is the use of a reactor with a molten fuel core. Tn this reactor: a) the most important parameter to be optimised is the reduction in the inventory of volatile fission products (which will control the consequences of an accident during the first hours) and fission products such as Sr-90 and Cs-137 (which determine the accessibility of the contaminated area for tens of years) , 13 14 Fig. 1.1 TRENDS IN THE PRESENT DEVELOPMENT OF THE DIFFERENT REACTOR TYPES AND THE REACTOR TECHNOLOGY SAFETY PROBLEMS HIGH HIGHEST TEMPERATURE SAFETY HEAT - HIGH Pressure in the (QTSOFT (G?\‘Eggg‘?) coclant or fuel HIS PAP T - No High ;l\ {good +) (bad - i SOFT/r Ligquad Light ‘\K Metal Water High Fast Reactor {pad -) Breeder (=) (=) Fission (+) (=) product inventory Low Molten Gaseous . LIQUID ™\ and (good +) Salt Fuel E@@% FAST}) after- Reactor Reactor BRE@QEB/ . heat (thermal-flucrideyd (-)(+) fast-chloride) HIGH (+)(+) FEASIBLE UTILISATION POWER OF URANTUM Improvement possible No possibility for further improvement b) decay heat resulting from the spontaneous radioc- active decay of fission products and actinides but alsoc of the structural material is significantly lower, even 1f it 1is not enough, ¢) the breeding ratio has been set at 1 in order to minimise the need to transport fissionable material. this makes the reactor a net zero consumer and zero producer of fissionable material, d) the entire system fuel, primary and secondary cooling systems have been desligned to operate at amblent pressure, e) no highly reactive compounds, (e.g. metallic sodium) are present In the system, f) no hydrogen containing compounds (e.g. water), which can evolve free hvdrogen., are present. (see Ref. 14, 15 2 16, 17, 18). 2. ACCIDENTAL RELEASE OF FISSTON PRODUCTS AND DECAY HEAT REMOVAL 2.1 The Rasmussen Scenario (Ref. 12, 2L4), The scenario of an accident used here is taken from the reporst WASH-1400 as the most fatal case: the PWR-1 which can be characterised by a steam explosion on contact of molten fuel with water in the reactor vessel. This accident category includes the following fraction of fission product core inventory release: 16 Rasmussen German risk study (maximum) Xe - Kr 0.9 1.0 1 0.7 6.8 Cs=Rb 0.5 0.5 Te-5b 0.4 0.35 Ba-Sr 0.05 0.052 Ru (Rh,Co,Mo,Te) 0.4 0.38 La (Y,Zr,Nb,Ce,Pr, Nd, Np, Pu, Am, Cm) 3%10"° 2. 6x107° The release of fission products and actinides to the environment results in two different scenarios (simplified here): a) the impact of volatile short lived fission products from the passing cloud in the direct neighbourhood of' the reactor - some hours after the accident, b) the impact of non-volatile long lived fission products. 2.2 The Problenm From consideration of this scenario of the worst reactor accident 1t 1s clear that to improve the safety significantly the proposed reactor must allow the continuous extraction of both classes of fission products - the short lived volatile F.P. and the long lived non-volatile F.P. This 1s the most important aspect. (Table 2.1) For the accident described above to occur in its entirety it 1t 1s necessary that the last barrier, the containment building becomes breached. 17 Table 2.1 The most dangercus Fission Products according to (WASH 1400) Release Critical Nuclides Relative Proposed Counter- Cat egory Organ Dese measure Bg@éfifigfl bone I-132,135,133 w500 continuous from marrow I-131 extraction Kr-88,Te-13%2 of volatile assi passing F.P. by - [ mn cloud lung I-132,135,133 VR4 means of I-131,Kr-388 from ? He-gas Te-1%2,3b-129 . purging reacton lower I-132,135,133 240 hours large 1-131,5b-129 after intestine Te-132,131m 1 . release thyroid I-132,135,13%3% 330 gland I-131,Kr-88 Te-132,3b-129 Long term bone Sr-90,Cs-134 6.7 continucus _fffect marrow Cs-137 extraction T f non-volatile Ce-1Uk Ru-106 © (0-10 yrs) F.P. in of inhaled lung Ru-106,Ce-144 126 the fuel : reprocessing radio mineral Cs-134,137 n18 plant nuclides bone Sr-90,Ce-144 15 /lm breast Cs-134,137 vl,25 f'rom Ru-106,Ce-144 reactor Fig. 3.1 18 1SOFTTREACTOR VERSUS CONVENTIONAI, REACTOR N SOLID < FUEL REACTOR FU ‘ RE, -4 a CoO ""* : b L = 3 . * * ol Al 2 Bl ol x| x o * * % - : U FU 'SORT! CN REACTOR o Uranium makeup ki P ] U 7 P A% % F l S EZ:a Fuel element <+ Fission products (F=z=P=) ~—~—+ Decay heat RE = Reactor ) RP = Heprocessing Power, FP = Fission Product : : CN = Containment ; Criticality FU = Fuel element Co = Cooling of irradiated elemMents 19 This can occur in the following ways: a) steam explosion, b) leakage of pipes, ¢) hydrogen burning, d) overpressure due to heating, e) melting of the core systems, f) loss of fluid giving overpower (for fast reactors). It seems that In the reactor system propesed here almost all of these failure mechanisms (excluding the leakage of pipes) could be fully eliminated. The safety is thus significantly improved. 3. PRINCIPLES OF THE 'SOFT'" REACTOR 3.1 A Schematic View Of course there is no possible way to bulld a fission reactor in which the dangerous fission products are not produced at all, or in which they could be fully destroyed (by transformation) in situ. Therefore the only possibility 1s to shift in time and space: a) fissionable fuel as a critical source and generator of power - both being controllable, b) fission products as a source of radioactivity and source of decay heat, both of which cannot be controlled. Figure 3.1 shows this principle. 4.2 System Descriptilon FFor the sake of simplicity, and by no means as an optimised system the reactor characteristics can be summarised as follows: (Table %.1 and Fig. 3.2) - fast converter reactor: internal breeding ratio of 21; includes a continuously operating reprocessing plant, - thermal power : AGW 20 ) n- L L LA =) > AR B0 AN NS SN _JUN ASS JUK 3 vy & VY § VT D g MNANANANEANAN - m 3] o s 4 % 1 N[ NN VA VA AAee4 e N > L L T T T T T S T T T T 7 \- FPV+FPN 08 = FEmergency cooling 5 h r o = Secondary cooiant, 4007 VYol bh W_I o~ o : A L) —FPVY + FPN — - sc! = Oontainment dlancter = vore, IR = kel lootor [ T 4 4 L | | | | -n T = Heat exchauger S0 MV Ut 1 “ = ZSeconaary cooclant, - MW g = frerpgency cocl 11 cOol1ng mn sslng plant sing system (Hellium) i = . o i FPV = Fisslorn products, volatille i i : - . N T r = S T PN = Fission products, nonvolatile Lo = Uranlum make-ugp Po= Tuun for Tuel = Fuel v o Zz < 21 Table 3.1 A Short Description of the 30FT Reactor Unit Value Reactor Type of reactor - fast Type of fuel - molten salt Type of cooling - external cooling Power - total GW({th) % - electrical gross GW(e) 1.2 - effilciency % 40 Fuel - composition 1 PuClB-B UClB-lo NaCl - density kg/liter 3,3 - total mass kg 247200 - mass pluteonium fisslile - mass plutonium total - outlet temperature ~ temperature, inlet - specific heat of fuel core - volume flow through core - specific power - total mass (i core + heat exch) - total volume (in core + heat exch) - breeding ratio wh fuel kg w? fuel;kg 5.18w% 512800 6.48w% 3016020 650 M7O}AT = 180 0.188 6.65 12.14 353 000 107 1. 0Y core - geometry - radius - volume - specific power - dwell time in core - dwell time ex-—core - total circulation time - neutron flux Coolant System - primary heat exchanger - secondary coolant - amount of fuel ex-core - volume (Fig.6.1 page 48) m - mass - volume flow in core - mass flow in core - dwelling time in-core - dwelling time ex-core - total recirculation time - amount of coolant - gpecific heat - density Reprocessing - total amount of fuel in system - total amount of plutonium - total amount of plutonium m 3 m kiW/liter S 2 n/ecm s number Kg kJ/liter.X kg/liter kg kg moles Value spherical 2.616 74.99 40.00 “v11.73 h.8 16.1 5x100° 4 molten chlorides 106 000 353 000 22 874 ~9h 000 23 - thermal power : 3 CW, - fuel : molten chlorides : 1 PuCl5 - 8 UCl5 - 10 NaCl, - the reactor is cooled externally. This means that part of the molten fuel is pumped out of the core through a heat exchanger, - the specific power in normal operation is very low Lo kxW/liter of fuel in the core, - fuel is at a rather low temperature of 65000, - the internal heat sink 1s very high: the fuel itself has a heat capacity of approx 80 GJ(th) per 300 K temperature increase. The negative properties of 'SOFT! - large plutonium inventory: 22.8 tons, 1in no way optimized - low specific power per unit of fissile material, - a new and unexplored reactor type, - serious corrosion problems. 3.% Mgss Plows and Containments This kind of reactor is characterised by the existence of three separate mass flows (see Fig. 3.73%): - the flow of the molten fuel through the external heat 3/S). The ratio of the dwell time outside and in the core is 4.8 =/11.% s = 0.42 exchangers (approx 6.65 m - the flow of molten salt through chemical reprocessing is 0.25 liter/s. The ratio of the dwell times 1n the reprocessing plant (5 day cycle) and in the core is 2.4 nr/5 x 24 nhr. = 0.02 The fission products of concern in the case of an accidental : C . release can be removed by continuous gas purging wifth helium in the core. 24 MA Gx 750 MW (TH) DHCONDARY COOLANT Fig. 3. S8 PLOW LN . z - "SOFT" HEA'T B TURETKE EXCHANGER —3 £ oo e EEFLIETTOHR awell time: a s mohx1l07 s He fe NONVOLAT. GAS PURGING He + F.F. F.P. CONTATNMENT VOILATILE Nucelide Table 4,1 25 eutron Balance of 'SOFT! keactor Core | ~ef'lector Total Product Fisslon Capture Capture |Froduct Pilsslion Uapture J235 0,006 £,.019 0.005 0 0,046 0,019 UL 005 P46 G.713 SL242 0 0,052 0 0,713 0,242 2,052 i1z A O, 140 0.081 0L350 0 O.140 0.051 G.350 Puodn D.021 9.a27 0LC009 o c.021 0,007 0.009 o 8 G G 0 a 0 o Na 0 N G002 CLGel 0 S.003 Fe o 0 . 0.042 i £ o042 Fusdl G.0685 1,022 .00 8] 5,065 0,022 2,004 Puzhz J.501% 0,005 0,005 0 C.olh £.005 0.005 FoR, 0 0 0,001 s { 0 .00l Mo 6] 8 7 3 0 0 0 410 . . o 3 0 s 0 I 0 1] 4] 0 g 0 Ca 0 0 0 0,000 0 5 0 o 0 < 2.105 £.054 0 0 0 Total 1.0C00 0.346 0.533 D097 000 0.346 0.630 =L P, VOLATILE CONTAINMENT 26 L. NUCLEAR CALCULATIONS 4,1 The Method The calculations were based on the British 2240 group nuclear data library FGL5. Twenty group cross-sections for cach reactor zone were derived from FGL5 using the associated cell code MURALE. The reactor geometry was represented by a spherical model with one core zone and a reflector zone. Transport calculations were performed in the discrete ordinate (SM) approximation by the ANISN code to obtain the physical properties of the critical reactor. The effective delayed neutron fraction was derived from basic ENDF-B-IV data using proper flux and adjoint flux welghting and an appropriate reduction of the delayed neutron yields due to precursor losses. The main results of the physics calculations can be summarised as: critical radius 2.616 m average neutron flux 5X1015n/cm25 temperature criticality coef ficient 0.018% /K (see also Table 4.1 and Table 4.2) Effect of doppler contribution - not calculated but according to ref. § 1s very small 27 (barn) o O O o N O . 082 032 L2672 . 801 470 - .885 - 3 .996 - . 736 .52k 191 .115 146 T2 - 237 = 3 L1UT7 - Table 4.2 Calculated One Group Cross Sections Nuclide Concentration VOJJC (102M0m—3) (barn) Uess 0.%1 - 4 '.109 Pu239 h.ooh - 4 h.919 U238 hzo.75 - 4 0.916 Pu2lo 0.54 -~ 4 1.066 C 1.0 - 18 0 Na 5%.83 = 4 0 Fe 1.0 - 18 0O E% Pu2il 0.27 - 4 6.763 S| puzlo 0.54 - 4 0.791 F.P 0.11 - 4 0 Mo 1.0 - 18 0 B10O 1.0 - 18 ¢ K 1.0 - 18 0 Ca 1.0 - 18 0 cr 199.3%9 - 4 0 Uo3s 1.0 - 18 10,403 Puz39 1.0 =~ 18 11.597 U233 1.0 - 18 0.123 Puz2lio 1.0 = 18 0.576 C 1.0 - 18 0 Na 53,65 = 4 0 Fe 420.60 - 4 0 Puzil 1.0 - 16 2h.676 Pu24? 1.0 - 18 0.261 FP 1.0 - 18 0 o & | Mo 1.0 - 18 0 Efl) B10 1.0 - 18 0 E K 1.0 - 18 0 Ca 33.65 - 4 0 Cl 100.9 - 4 0 O O O FE Ny OO O Cc o o O o — (-] o O O N0 O = 154 . 595 .541 L076 .120 - .287 - L 136 - . 920 . 384 .009 .64l . 501 L2hs - .865 - 28 bh.2 Nuclear Properties of SOFT The loss of neutrons due to the use of natural chlorine are 10% which i1s significantly high but far from being prohibitive. (Ref. 8, 10, 16, 20, 21). 5. CONTINUOUS EXTRACTTION OF FISSION PRCDUCTS 5.1 General Scheme The general idea of this reactor type 1s the differentiation of the chemlcal properties of the fission products in the molten chloride medium. (Ref. 10, 16). The rough scheme is given in Fig. 5.1. 5.2 Chemical State of Fission Products in Molten Chlorides It 1s important to know the chemical state of the fission products. The fission reaction of PuCl, can be written as % follows (E1 = fission product element) (index O for elementary state) 1 PuCl, LRIy E1-C1, + V2 E1_ Cumulated Cl-demand is calculated from individual fission 239 product yields (Y ) for fast fission of Pu irradiated for 10 days and product stoichiometry starting with the most stable chloride product being formed. Where necessary (i.e. where one fission product will form several chlorides of comparable stability) a weighted mean was calculated based on the Gibbs free energies at 1000 K. (see also Ref. 6, 17, 22) Table 5.1 fives the cumulated chlorine demand of the fission products forming stable chlorides at 1000 X(100% is set equal to the number of fission events - total C1l supply therefore equals 300%). Because of the apparent lack of free chlorine 29 30 Table H.1 SIMPLIFIED PLOW CF FISDTION PRODUCTE Cumulated Cl-Demand of Fissicon Products forming stable c¢hlorides at 1000 K Compound Yield Chliorine Cumulative s demand ¢hlorine L demand o (%) ¥ e { > BT BaC1 3,502 19,00 19.0¢8 [ 2001 1,050 1.050 20.05 dels Cs0 1%.355 13.355 33,41 ey =1 - WO - - : @ Sl 5. A48T 10,974 L4, 38 L — - Siraioned a Smel 3.737 iy 51.85 plutonium - o . - - - P LaCl, 5,754 17.382 £9.24 E CeCl, 15.586 1.458 L. 20 = Prcl, b.278 12,83 124,03 Nacl, 11.87 35,01 159,64 sy T Yol 3,008 9. 08 168.7% t bl > J N Rh, Hu R Zrcl, * 21.529 £3.05 231.78 METAL 2.6 ) 7 7 T InCl, 5. 0. 06D 1.078 231,85 £dol . 661 1.322 222,18 & rlg Src1 0. 324 0.648 23%.82 e Sbol, 0.674 2.032 235, 86 b e ) NCBELE AgCl 1.880 1.880 237.74 VETALL Mo, Ay, ; Sb . rd 4 Sell 0.088 C.088 237.75 47 1ol e = FLECTRO- %;’ S0, o TeCl 7.654 15.308 253.06 NEGATTVE ’ ” NbC1 0.289 1. 445 254,51 METALS o o MoCl, o4 18.16 53U 288.51 8} . — PacC1 12.66 5.8 297.31 TeCl 4,01 2.7 300.00 Y, Pr, Eu, Pm: Ce; Ta’ 71 Nd, Sm, Gd, Rb, Cs, ' Sr, Ba, * weighted mean 1000 lal o) Tibbs free energy at 1000 K 31 only 65% of the molybdenum and only 357 of the palladium and technetium being produced in the fission process will form chlorides. These cations cannot form stable iodides at . . . 0 : : - «<10"Ll 1.65x10'l 100000 6.95x10'5 9.@3x101 0 O 1.8]_}(102 Similar calculations were made for other extractable nuclides. The extraction rate cannot be very high not only because of the physical and chemical constraints and the prohibitive amount of helium bubbles in the fuel (effecting the criticality due to changes in total and local bubble volumes) but also because of the problem of the precursors of the delayed neutrons. 5.4 Problem of Delayed Neutron Emitters A significant proportion of the delayed neufron precursors are the volatile fission products. Delayed neutron emitters will be removed along with other volatile fission products by means of gas purging. The nuclides of Interest and theilr respective half lives are approximately as follows : (see Ref.9) Nuclide Neutrons/fTission Half life (s8) x 100 87BP 0.024 54,5 1571 0.176 A 88Br 16.3 0.13%6 1384 5.7 89Br 0.207 b, L 1391 2.0 Vpp 0.087 1.6 Total 0.630 39 138 139 137 I and I have no precursors. "I and 1ts short-lived precursor 137Te (t]/2 = 3.5 8) The effect of gasecus extraction upon delayed neutron emitters inventory can be calculated roughly by fthe same method used in section 5.4 to calculate the degree of extraction of the other volatile F.P. The results of these calculations are presented in Fig. 5.4 and FTig. 5.5. (Ref.17) Delayed neutrons will also be lost in the heat exchanger as the salt melt i1s pumped through this loop. The salt melt will remalin an average of 11.3 s in the core and 4.8 s in the external heat exchanger circuit. During these & s the number of radlo- nuclides is reduced by a factor of 2 exp (-58/t,.) thus total e loss due to 8 decay in the heat exchanger circuilt 1is: 4.3 -4.,8 g Y 0(1 - XeXt> [ ] m [ ] (l—? 2 eXp( t )) V2 t arbitrarily chosen as With a time constant of extraction te 1023 the following results can be obtained : Radio- Relative yield Loss due to Loss due Total "Weighted™” nuclide (normalised to B -decay 1in to ex- loss to losses vy o= 1) external heat traction yield exchanger ratio 7., 0.129 0.313 0.187 0.500 0.965 137I 0.499 0.289 0.076 0.365 0.182 88BP 0.120 0.269 0.046 0. 515 0.038 138I 0.096 0.192 0.018 0.210 0.020 89Br 0.058 0.152 0.008 0.160 0.009 139I 0.080 0.059 0.001 0.060 0.005 90Br 0.028 0.038 0.000 0.038 0.001 Total 1.000 0.250 0.070 0.3%20 These losses of 0.3%2 are significant but not prohibitive. 40 SEEOVAL OF DR A s T e T S Fig. 5.4 0.8,4 e o \ \\ . N ‘s 878Br oA \ \1371\ ; v\ geaN .\ \ \\ \\ O.M-\ \\\ “\ \ . - - \J381 SN e~ =T e — = 898“\‘\ S0 S - —— ——— — 18991 0 90 BI" |2 3 1 4 | 10’ 10 10 10 10° , Fig. 5.5 TOO HIGH DELAYE] PO SSIBLE ! INSUFFICIENT REMOVAL NEUTRON LOSS OPERATING| OF HARMFUL F P. RANGE * FRACTION REMOVED 0.2- delayed neutrom e;mtters lost in, heat \exchanger 41 5.% Gas Extraction Rate In a % GW(th) recactor 9.3 x 1019 fission/s occur producing 3.09 x lO_u mol/s of fission products, 1 x lO_L’L mol/s of which are volatile. The total volume of helium bubbles in fthe core should not exceed O.lmfiB(less than 0.2% of the total core volume) that is 5 moles of helium in order to minimise the rise Iin criticality due to loss of helium throughput. If we assume the extraction process is limited only by diffusion (i.e. the volatiles are completely immiscible in the molten salt which is likely to be the case for Kr and Xe but hardly for I 2 TeCl3 and SbClB) the extraction rate can be approximated from film theory. The model assumes that the gas in the bubble undergoes toroidal circulation. As the gas circulates 1t encounters fresh liquid at the top of the bubble. As the bubble rises the liquid moves downward in relation to the bubble and leaves when 1t reaches the bottom of the bubble. The liquid near the 1interface is usually in laminar flow and the contact time 1s certainly short, thus the ligquid behaves much like the ligquid at the surface of a falling film. Therefore to g first approximation the average rate of mass transfler across the interface 1s: (Ref.2) avg Lo (NA)avg - .__AB 'bAO -+ cont in which CAo 18 the concentration of volatile A in the salt a its diffusion coefficient in the melt and t the AB cont contact time given approximately by tcont = 7 where D 1s the bubble diameter and Vt the terminal velocity. melt, During the normal operation of the reactor the volatiles produced in the fission process are removed as gquickly as they are produced. 42 Theref ore: = L I K S ¢ N .(NA)avg Where F 1s the fission rate, S the surface of 1 bubble and N the number of bubbles in the core. A uniform bubble diameter of 3 mm was assumed and GAB calculated from the Stokes-Einstein equation. For a typical volatile moleculi w%th a molecular welght of ~30 the Yalue of eAB is no2x1077 m-/s. CHO thus equals 7X10_6 mol/m'z> equivalent to the production for 7 seconds. (Ref.2, 23) =2 This investigation shows that: - gaseous extractlion of volatile fission products would be an effective means of reducing the risk to the environment due to an accidental release of part of the reactor inventory in a MSFBR, - the extraction rates which would be necessary to achleve this should be attainable with a reasoconably small through- put that should minimise any criticallty change and allow a reasonable design of a gas purification system, - delayed neutron emitters are removed less than the volatiles that cause concern in the case of a core accident because they have few (or no) volatlle precursors and are short-lived compared to the time constant of extractiorn. 5.4 Reprocessing of Non Volatile Fission Products. The fisslion products experience the conditions in the molten chloride core of fthe recactor and are at“J6BOOC and in the non- volatile form,e.g: - dispersions of metallic particles (colleids?) Rh, Ru, R . ,» BECL, ., - solutions of molten chlorides:CsCl, SrCl They must be extracted from the irradlated fuel by means ofl chemical reprocessing, 1f possible continuously. A rough list of the different fission products which may be formed 1in the molten chlorlide reaction 1s given in Table 5.73. 43 Table 5.3 Extraction of PFisgsion Products in 'SOFT' menen/ IS DI Widla nw ~ (yield per species at 1500 1 bar 100 fissions) at 950 K ( k J ) (K) moli Cl Sr 90 5. 487 SrCl, (1) -520 2300 Rb 86 1.050 EbC1l (1) -275 1654 Cs 134 Cs 137 15. 355 CsCl (1) =275 1575 La 140 5.794 LaCl3(l) -255 2020 Ce 14k 1%.986 CeClB(l) -245 2000 Pr 143 4.287 Pr015<1) -245 1980 Y 90 5.028 YCIB(l) -240 1725 Nd 147 11.870 NdClB(l) -235 1960 Sb 129 0.674 SbC1,(g) - 80 el Te 131m Te 1%2 7.654 TeClz(g) - 62 595 Te 99m L.010 TCClg(l) -17 1474 Ru 105 31. 445 Rumet(s) 0 b170 Rh 105 1.73%6 Rhmet(s) 0 Looo Le 133 Ye 135 21.23%4 Xe (g) 0 166 I 131 ) I 132 I 133 6.177 L, (g 0 >M57.5 T 134 I 155 Kr 88 0.942 Kr (g) 0 7 121 Remarks (1) Iiquid (g) gaseous * The "decontamination factor™ without and with extraction. The letters R and G refer to is defined as the ratio of Decon=- tamination factor * chemical reprocessing and gaseous jru B s B v By oy g inventories extraction respectively AG? = free energy of formation at 1500 K in kJ/mol C1 b 11 boiling point at 1 bar 44 5.7 Some Technological Problems Apart from the fission products and corrosion products an important problem in this molten chlorides reactor is the removal of sulphur. In a flux of fast neutrons the following reaction takes place: (Ref. 8, 18, 20} 35 C £ 17 Ve This sulphur is obviously undesirable and makes the corrosion 55 ( 1 (n,p) 168 = 88 4) problems worse. The continuous chemical reprocessing can be used to remove the sulphur. Nevertheless a consliderable problem remains regarding the resulting corrosion. 5.8 Possible Reprocessing Technologies Tt is possible to separate the fisslon products from the molten fuel 1in various ways: - extraction from molten chlorides by imiscible liquid metals. This is a well known technology developed especially for the molten salt breeder (Ref. 13) - electrochemical dissolution of metallic uranium. The continuous feed of uranium (depleted or natural) 1in amounts of some gram/secpermits the removal of fission products which have a free energy of formation of chlorides of ~1000K which is smaller than that of UClg. This applies to a good proportion of the semi-noble F.P. (see Fig.5.1 ) - the most attractive method is electrolysis in the molten chloride system. The following very simplified calcula- tion explains this: The total amount of plutonium In the fuel is 96 000 moles. (see Table 3.1). The total amount of chlorides is 1.82 x 106 moles (PuCl,, UClB’ NaCl). For a five day reprocessing cycle (4.3?X105 s)the flow to the reprocessing plant is 4.21 mol/s. Taking into account the electrolysis 1s at a voltage of 3 volts (see Fig. 5.5) and the efficiency is 0.5 then the power needed for the electrolysis 1s: Fig. 5.5 Fig. 5.6 ] Be ETLECTRTICAL Al POTENTIAL TN MOLTEN CHLORIDES rare carths Free energy, 0 10 Atomic number, Z T 90 n ~ 19} 1 at 120 ~ V) LS /mol BOTTING POLINT OF AND CHLORIDES 2 2500 1 = _6408a, g 11384 Sr-d 1000 T-882 - Nayg 1-690-Csue - 679-Rb g — 500 — LORIDES Ty e L1747 La Cl, L1560 Ba Cl, —1413 Na Cl 1381 Ro Ci 732 2r Cl, 47 .21 mol/s x 3V x 96 500 C/mol x Blg = 2,44 MW(e) The amount of electrical energy needed for the electrolysis of the total amount of fuel, related to the total power production is only: 2.4 MW(e) = 0.16 percent A000 MW(th)X(ne ~ 0.5) Continuous electrolysis seems to be feasable both from the econcomic and technical point of view. 1 O The Problem of External Storage of the Fission Products For the reactor type proposed a special problem has to be solved: that of the interim storage of the fission products from the tfime of extraction from the fuel fo the time of transportation away from the reactor building approximately 50 days later. The criteria giving the form of chemical and physical form best suited for interim storage is shown 1in Table 5.4, 6. HEAT REMOVAL 6.1 Primary Circuit: Core and Heat Exchangers Molten fuel reactors have been proposed having differing methods of cocoling the primary fuel : a) internal cooling in which the coolant is pumped into the core in for example tubes, b) external cooling systems in which the fuel is pumped out of the core through heat exchangers. The reactor proposed here has the following characteristics: - the heat removal is based on the pumping of the molten fuel out of the core. The dwell time in the core is ~v11.3% 5,and outside the core ~ 4.8 s, 48 - the outlet temperature of the fuel is 65000 the inlet temperature 470YC. The temperature difference of 180°C with a volume flow rate of 6.65 mB/s permits the removal of 3000 MW{th) (see Fig.6.1) - the secondary coolant is also a molten salt containing chemically inert chlorides of Na, Ca, Ba, Mg. No pressure is needed and no hydrogen containing a volatile compound are present. 6.2 Secondary Cilrcuit The followlng are the requirements for a cooling medium for this reactor: - generating at the proposed 6OOOC, pressureless and stable for long periods of irradiation. - non reactive to molten fuel (no exothermic or endo- thermic chemical reactions, no precipitation or gas evolutilion. Some media which fulfill the above reguirements are: Low melting chlorides mol per cent T T Tmelt (OC) NaCl KCLI Mg612 CaC12 BaC12 20 20 60 - - 396 38.5 - - b7 4.5 bso - 20 60 - 20 Lho 6.% The Tertiary Circuit A rather unconventional choilice has been made for the medium for and form of the tertiary circuit. W] 49 Table 5.4 In what form should the Fission Products be stored? Criteria Chemical =stability (radiation damage) Low volatility Low pressure lligh thermal conductivity {heat transport) Fasy management - storage - solidification Low specific power Small volume See also PMig. 5.6 Fositive oxide, chloride, fluoride. metals, alloys. oxide chlorides solid or molten state metallic - solid metallic - liquid molten salt liquid diluted concentrated Negative carbonate, sulphate agueous solutions metals agueous solutions oxide, solid compounds . solld gaseous concentrated diluted , 50 Fig. 6.1. SCHEME OF HEAT REMOQVAL qp py “_'-‘ FUEL Jiam: O yom — » I . b ] ot . i nelent s Sm 7E0 MW{th) Z U 8 MW/m- HEAT FXCHANGLEHN Fuel volumeo 0vald 4,7 m”’ Boo0 tubes diam. 1 em sine-wave- ~-tibes N Tiam: Ooom Tlameter: Volume @ 75 m Mean velocity @ 0.4 mis Dwell time - 11.3 s Fuel mass 1 247 ¢ ho KW/11¢ Specific pawer Muel 10.% m/s ‘ Total volume 4 2 '/]m ’ 51 The c¢criteria for selection are as follows: - the highest possible thermodynamic efficiency, - the full use of the high temperature of the molten fuel reactor, - minimising of the risk of chemical reactions between secondary and tertiary working agent, - minimising the pressure in the working agent, - use of the same circult materials as for the secondary circulit if possible. A1l these reguirements can be met by using aluminium trichloride as the working fluid. This has been discussed earlier (Ref. 14, 15) 6.4 Heat Removal from the Stored Fission Products An additional problem for this reactor arises from the continuous extraction of the fission products from the molten fuel and transfer out of the reactor containment. For this a second (small) containment must be bulilt to house the stored 'hot' fission products. The power produced will be of the order of some tens of megawatts (th) depending on how quickly the fission products are removed for further treatment elsewhere. This system therefore requires a special cooling system which can be bullt with some simple redundancy relatively inexpensively. The problem of the possible processing of the {ission products and transportation methods have not been discussed here. 7. ACCIDENTS - PROBLEMS AND SOLUTIONS 7.1 Decay Heat: Power and Energy The crucial problem of each fission reactor is the problem of 52 the decay heat removal during a worst core accident. It 1s known that the fission product inventory is: - for short lived F.P. proportional f£o the specific power (that is to the neutron flux), - for long lived F.P. proportional to the energy produced (that is to the neutron flux times length of operation) To carry out a rough estimate the following simplified calculations have been made: a) the decay heat power from a fission reactor has been taken from ANS guides (Ref. 1) (see Fig. 7.1). As a reference the decay heat from the thermal fission 2 of BgPu has been taken, b) for the sake of simplification the decay heat power curve has been presented as an overlapping series of six curves having half-1ife value of 108, 107, 106, 105 104, 103 and 102 seconds. 239 2359 c¢) the decay heat of and Np has been taken into account d) the build-up curve has been constructed for an irradiation time of 500 days and 5 days (Fig. 7.2), e) the effect of gas extraction on the volatile fission products has been assume to be as follows: - for the F.P. with t]/2 = 1025 only a quarter of the total can be extracted by gas pumping, - for the F.P. with tyg > 1@35 approximately one-third can be extracted by gas pumping. ) the effect of chemical extraction on all the long lived fission products with t ;31055 has been estimated to Vo be approximately 99 percent (rather optimistic), 4 5p 67 991" e 20 15 S | 1483 53 Fig. 7.1 APPROXIMATTIOR O 7 54 AY AT | =il 18T TON (Lefererce: TWhy 2 SW(thy days) T Do 1000 L 900+ i ms 800+ . [ - 7 00 WH 3 | 600+ 500 400 300 - ead et Do e g Doerd g r i Jy em— & + + + + e—— o a— 1o b b ime, seconds i Oy : §ITHE AT CXLNEeL of o ; ¢ 3 s 3 i Y e 3¢ ; T / */ O —— T G— — . —. 55 g) Figure 7.3 shows the different afterheat power curves for the SOFT reactor including all the processes and nuclides mentioned above, h) Figure 7.4 shows the afterheat power curves and the resulting integrated heat produced as a function of time. 7.2 Worst Accident Scenarios A number of simplified scenarios for the worst core accident for this reactor is shown in Fig. 7.5. The calculation of the heat and power for each stage of the calculation is given 1n Table 7.1. The following three points have been taken into account a) the scram I1s perfect (in this reactor type not only the engineered fast shut down system, buft also the Inherent negative reactivity (coefficient). A1l the cooling possibilities fail. The fuel in the core and in the external loop begin f£o heat up as a result of the decay heat. A temperature of 9BOOC is reached after a time of 0.7 hours after the scram. This temperature has been postulated as the maxlimum 'allowable' temperature of the fuel because of': - partial pressure of volatile components in the irradiated fuel, - structural material properties (e.g. molybdenum alloys for the core vessel), - management of molten fuel. The end of this step i1s shown in Fig.7.4 and Table 7.1 as point A. b) an interim mechanism has been postulated in that the reflector material belng composed of chlorides of e.g. NaCl, CaCl., MgCl etc. 2 1000+ 1 2 10 10 10 58 Stage After scram all cooling systems fail. ing of fuel Reflector begins o melt Fuel is drained to the core catcher tor is station- ary and is neglected) Guillotine Table 7.1 Scenarlc of the worst acclderts Calculation Heat in Curiulative Elapsed time Decay heat this heat from from start power at stage start (k) given time (3J) (GJ) ' (114 (an)) Mass of fuel 353t Specific heat 0.8MJ/t.X Self heat- Mean fuel temperature 116 119 0.7 38 at start U470+650C 2 Mean fuel temperature at end 9509C Mass of reflector 6A0tL 1 3 23 T\rT L Spec1flcnhe§t ml.O:J/tK 5hg 650 7 14 Heat of fusion ¢.5GJ/t Temperature start 470°C Temperature end 700°C Huel temperature 050°C Fuel mass 353t {(reflec- Catcher mass 2000¢ Catcher temperature 2080 2170 42 6.5 start 20°¢c Catcher temperature end 600°¢ Specific heat 0.9 Fusion heat 0.5 Fuel mass 3R 3 Mean fuel temperature PN at start 5007~ Catcher mass 2000t Catcher temperature 2000 Heat of fuslon C.5GJ/t Core catcher f{inal £ om0 temperature STV REY fuel heat losg: 353x0.8(950-500) 107 3alt heating and melting 2000 (1.0x(600-20) + g Lz 6.t Ralarce of heat 2170 60 LURNTS g, 7.6 A Cuaton e spraying SCHEME OF TH® HEAT >f RN Jore calchier TRANSPORT 16— 3 M TR THE CORTALNENT drainarme and i Cpontancons t ? }}I(ii < remo anp % Guiillotine main tubes 61 and as a result of the temperature increase from M?OOC Lo 700°C and then 950°C, also heat up and melts. (Note the problem of heat transfer from the core to the reflector has not been discussed). Point B on Fig. 7.4 and Table 7.1, c) the drainage to the dump tank 1is negligible, d) the fuel at 95000 drains to the core catcher filled with 2000 tons of solid chlorides with a melting point of approximately 600°¢. (Point C: Fig. 7.4 and Table 7.1) 7.3 Spontanecus Cooling Processes in the Containment In this scenario it has been assumed that all cooling systems (including all redundancy) and all emergency cooling systems of the core of the dump-tank of the core catcher and of the internal contalinment fail. Figure 7.6 shows the principle layout of the reactor after drainage of the fuel to the core catcher. The transport of heat from the molten fuel and the core catcher mixture 1s as foilows: a) the molten mass ofw2350 tons reaches the permissible temperature of 6OOOC after 42 hours from the start of the accident (and scram). At this moment the decay heat is approximately 6.5 MW(th). b) the core catcher has the following characteristics - the upper part free molten salt and a temperature of 600°C (873 X). - the lower part of steel at a temperature of H5OOC (723 K The heat losses due to infra-red emisslon are: - for the molten salt surface (¢ = emissivity) 1 -8 Q :HOOmgx(g =0.5)«(0 =5.67x10 J/(mgKu)x(873)u:6_5 MW 62 - for the steel structure ’" 2 d = 800 m° x (e = 0.8) x (6) x (7237) = 9.9 wmy The total heat emission due to infra-red radiation allows the removal of all the decay heat once this has dropped below 10 Mw. ¢) this heat has to be removed from the alir and the following paramecters have been taken: Temperature Heat Deng%ty Pressure Fffective o) Capacity (kg/m~) (bar) density e (K. Joule/kg) 507 1.01 1.16 1.1 1.27 200 1.03 0.74 1.7 1.26 For the simplified calculation of the heat transfer coefficient for natural correction in air the following simplified relationship has been used (in BTU, ft,OF) i 1 n = 0.22 At/B (according Xern 1950: hc = 0.38 At" ) ffor At= lOOOC the heat transfer coefficient : no= 7 W-mng_1 (from Kern 1950 : h= 7.9 W-m_gK—l> ffor the bulk temperature difference : _ - : _ : 0 AT _(Tair Tsteel) 200-30 1707°C the heat flux will be : 2 - - - / sip Tx170 = 1.2 KW/m The desired heat removal from the outside wall of the containment (dia 56m, for half of surface) ! - . ],C 2 Hdecay heat 6.2 "g = 1.5 kW/m 5000 m d) the air as a heat transport medium Heat capacity of 1 m5 air = 1.02x(1.16)x(200-50) - 1 ‘;T ki K (J— 63 7.4 The Prompt Critical Scenario Since in this study only steady state calculations have been made the problem of the prompt critical excursion is not discussed seriously. However a rather naive attempt is made toc obtain a rough idea of the magnitude of events under such a scenario as follows: a) the best known data for a prompt critical accident for the liquid metal cooled fast breeder SNR-300 has been taken as the reference (Ref. 11) b) where the SOFT reactor has significantly different characteristics such as: temperature of fuel during normal operation: SNR - 11547 SOFT - 550°¢ a heat capacity of total fuel inventory for o T = 1OOOOC; SNE - 0.21 GJ/100 K SOF'T - 28.2 GJ/100 K - fthen the values for the SOFT reactor have been taken. Table 7.2 shows the calculations. The reactor proposed here has therefore the following characteristics: a) the continuous removal of fission products and the adjustment of the fissile/fertile inventory continuously during operation. This minimises the amount of excess reactivity which must be compensated for by the control rods, which reduces the risk of rapid control rod with- drawal causing an uncontrclled excession, 64 Table 7.2 Prompt Critical Accident *) assumed by analogy. Parameter Method of Unit SNR~-300 SOFT calculation Reference - -~ KfK=2845 This paper Dec.1379 Nominal power (Po) - GW(th) 0.76 3 Accident scenario - - F& with compaction Ramp KT K }5 $/s 80 80* Reactor period " ms 3.2 3, 2% Duration of excursion " ms 23,3 23 . 3% (texc) Tncrease of peak e?j.B[&Z plax 1500 1500% power Fao Increase of average roughly Pav/ 740 TU0+ power estimated Pec Total energy (Pav} texc) PoS 7.97 7,97 Total energy relative to solidus KfK GJ(th) 6.1 2%3.9 (Q) Mass of m01%3? fuel KfK t (percent) 6.5(99%) 225(100%) I leat capacity () literature Md /kg. K 0.3%3 G.80 Temperature increase Qfiymelthp) K 2840 133 (aT) Operating fuel nor o temperature (T ) KfK (1684 C 1154 650 Average fuel nor o temperature after (T + AT) C 3994 783 accident Fuel‘temperature KFK °p 3920 (accident) Fuel temperature o maximum {(accident) KfK C 5060 Pressure after KK bar 90 3 accident Compaction Doppler constant Characteristic time for excursion (prompt generation time) Delayed neutrons Neutron lifetlime Reactivity Power at prompt critical Total energy release Fuel in core time Maximum burn-up Fluence Flux 65 Reference Unit SNE-300 SOFT - pcssible 7% Wirtzrp.105 Y - 0.004 A Loy 10 7 e - 0.003 0.00 1 1077 1077 «(1-8 )-1 - 2l 600 O = = 00_3 1 P t2 P(t)=POexp(§Tjr——) KfK day Wby 5 Kf¥ MWd/t £9000 not applicable due to con- tinuous reprocessing KfK n/em” 13x10°° ’ 15 15) ) =7 | - W/fi‘!fih{f%s (qj av 3. 410 )@jav >¥10 c) £) 66 the temperature reactivity coefficient 3K 1is dependant on two effects: Kol The removal of the fuel out of the critical zone due to density decrease (temperature coefficient of expansion equals mlxlOS/K) and secondly the increased leakage of neutrons, the inflow of cooler fuel occurs with time constants of at least several seconds and does not strongly affect the control characteristics, under normal operation the core contains ~0.2% of helium bubbles. Changes 1n gas flow changes fthe average density of the molten fuel. A depressurization of the fuel system which would expand the bubbles by a factor 2 to 3% would cause a reactlivity decreasey the wvoliding by local boiling in this type of homogeneous core 1s not credible even without pumping, the fissile material in the on-line processing plant amounts to less than 1% of the reactor inventory. If all this amount could be returned to the reactor within one minute the excess reactivity would be increased by 0.007% per second (by rough estimation), The worst credible event occurs probably due to a breach of the primary sytem boundary. This results in a release of large amounts of irradiated fuel - even the total amount into the containment building. The rather simple core catcher in the form of ~2000 tons of salt (see Fig. 5.2) 1g sufficient to control this accident, a peculiar property of this reactor is the possibility of inhomogeneity in the fissile material (PuClB) for example due to hydrolysis and/or reaction with oxygen resulting in the formation of insoluble Pu02 . The probabllity of such an event is very small since no water 1s used in the entire system, 67 Fig 7.7 Temperature increase during prompt criticality SNR-300 y il Bun e e e s e e = S OF T = SSRGS EEN SIS TN W NN S 0 : a | i + 0 5 10 15 20 25 time’ millisecond 8. CONCLUSIONS. The reactor concept has the following strong polints: - the steady state operating inventory of pofentially harmful fission products is significantly reduced, - high thermal stability against excusion, - removal of the decay heat 1s guaranteed even in case of a fallure of all emergency ccoling systems provided the outside containment spraying system works, for a power of less than 0.5 MW. however; ~ the technology of molten salt reactors 1s not yet sufficiently developed, - the corrosion problems will probably be difficult to solve, - the amount of Pu present in the reactor 1s very high, (but not optimized) [ (o) 10 11 15 14 15 16 17 63 REFERENCES ANS, (1978). ©Decay heat power in LWR. ANS Standard, 5.1 Bird, Stewart and Lightfoot (1960). Transport Phenomena, Wiley & Sons, New York. Burris L., and Dillon T (1957). Esftimation of fission product spectra in discharged fuel from fast reactors, ANL STH2., Crouch E.A. L. (1977). Atcmic Data Bull. Data Tables, 19, 4rv7-532. Engel T. R., Rhoades W. A., Grimes W.R., and Dearing J.¥. (1973). ¥olten-salt reactor for efficient nuclear fuel utilization without plutonium separation. Nucl. Techn. 1g, 30. Gutman V. Ed., (1967). Halogen Chemistry, Academlc Press, L.ondon. Heer W. {198C). Physikalische Berechnungen zum Salzschmelz- reaktor SOFT, TM=-22-80-26, Swiss Feder. Inst. Reactor Research, Wurenlingen, June. Furrer M. (1977). Chemical behaviour of radiosulphur obtained during in-core irradiation. J. Inorg. Nucl. Chem. 39, 1085, Keepin C.R. (1965). Physics of nuclear kinetics, Addison- Wesley, Reading. ~ Harder B., Long &., Stanaway W.P. {(1969). Compatibility and reprocessing in use of molten UCL, as reactor fuel. In "Symp. Reprocessing Nucl. Fuels"., Ed. P. Chiotti, USAEC, Con €90801. Royl P. (1979). Untersuchungen zu stérfallen in Kernkraftwerk Kalkar, KfK 2845, December. Reactor Safety Study: WASH-1400. Appendix VI, US NRC, Washington 1975. Rosenthal M.W., Haubenreich P.M., DBrigss R.B. (1972). The development status of molten-salt breeder reacfor. ORNL-4812 August. Taube et al: (1965). On the concept of salt-boiling fast-breeder reactor with AlCl3 as cocling agent. Nukleonika, 9-10. €639, Taube et al: (1967). New belling salt fast breeder reactor concepts. Nucl. Eng. Design, 5, 10G. Taube M. (1978). Fast reactors using molten chloride salts as fuel. Swiss Federal Institute for Reactor Research, Wirenlingen, Report 332, Taube M., (1974). The possibility of continuous in-core gaseous extraction of volatile fission producticon in a molten fuel reactor, EIR-Bericht Nr. 257. 18 19 24 69 Taube M. and Ligou J. (1974). Nolten chlorides fast reactor cooled by molten uranium chloride, Anr. Nucl. Sci. Eng. 1, 277. Taube M., Heer. W., Trdjefjord 4., (1980). A molten salt converter reactor with a very low fission product inventory. 2nd Intern. Conf. on Emerging Nuclear Energy Systems. 8-11 April, lausanne, Switzerland: Atomenergie 36, 57. (198C). Tarovieci E., Taube M.(1975). Chemical behaviour of radicosulphur 35¢1(n,p)35s during in-pile irradiation. J. Nucl. Inorg. Chem. 37, 2561. Taube M. (1976). The transmutation of Sr-90 and Cs-137 in a high flux fast reactor with a thermalized central region. Nucl. Seci. Eng. 61, 212. Veryatin U. D. and others. (196%). Thermodynamic properties of inorgaric compounds. Atomisdat, Moscow. (in Russian) Watson G., Evans R., Grimes W., Smith N. (1962). Solubility of noble gases in molten fluorides. J. Chem. Eng. Data. 7, 2. - Deutsche Risikostudie Kernkraftwerke. Verlag TUV, K&ln, 1979.