TIE UNIVERSITY OF MICHIGAN INDUSTRY PROGRAM OF THE COLLEGE OF ENGINEERING NUCLEAR POWERED GAS TURBINES FOR LIGHTWEIGHT POWER PLANTS Frederick G. Hammitt Harold A. Ohlgren January, 1957 IP-199

TABLE OF CONTENTS Page ABSTRACT. ii LIST OF TABLES iii LIST OF FIGURES iv ACKNOWLEDGEMENTS 1 1.0 INTRODUCTION 2 1.1 Simplified Closed-Cycle Gas Turbine Power Plant with a Nuclear Heat Source 4 2.0 CLOSED-CYCLE GAS TURBINE PERFORMANCE CHARACTERISTICS'7 3.0 SELECTION OF WORKING FLUID 26 4.0 TYPICAL NUCLEAR POWERED GAS TURBINE POWER PLANTS 28 4.1 A Typical Heterogeneous Gas-Cooled Reactor With a Closed Cycle Gas Turbine for Generation of Electrical Power and Process Steam 28 5.0 A TYPICAL LIQUID METAL HOMOGENEOUS FUELED REACTOR IN CONJUNCTION WITH A CLOSED-CYCLE GAS TURBINE POWER PLANT FOR GENERATION OF ELECTRICAL POWER AND PROCESS STEAM 39 5.1 Efficiency and Cost Comparisons With Alternative Heat Engine Systems 45 5.2 Cost Comparisons 47 5.3 Weight Comparison 48 6.0 BIBLIOGRAPHY 51 7.0 NOMENCLATURE 52

ABSTRACT The suitability of closed-cycle gas turbine power plants in combination with various types of nuclear reactors is examined. Typical examples of both a heterogeneous and a homogeneous reactor power plant are presented., Weight, cost, and performance are compared for various possible fluids over a range of temperature and pressure-. Comparisons are made with alternative heat engine systems. - ii -

LIST OF TABIES Page I "Basic Cycle" Parameters 23 II Heterogeneous Helium-Cooled Reactor Closed-Cycle Gas Turbine Power Plant Performance 31 III Heterogeneous Gas-Cooled Reactor Nuclear Data 34 IV Thermal Data for Heat Extraction from Heterogeneous Gas-Cooled Reactor 35 V Nuclear Data for a Liquid Metal Fueled Reactor Bismuth-Uranium Type 41 VI Thermal Data for Heat Extraction from Homogeneous Gas-Cooled Reactor 43 VII Homogeneous Reactor Closed-Cycle Gas Turbine Power Plant Performance 44 -iii

LIST OF FIGURES Page 1 Simplified Nuclear Powered Gas Turbine System 5 2 Thermal Efficiency at Optimum Pressure Ratio as & Function of the Ratio of Specific Heats 8 3 Thermal Efficiency of a "Basic" Gas Turbine Cycle with Recuperator Effectiveness = 0.93, & Frictional Pressure Loss=0.07 9 4 Thermal Efficiency of a "Basic" Gas Turbine Cycle with Recuperator Effectiveness = 0.75 10 5 Thermal Efficiency of a "Basic" Gas Turbine Cycle with Recuperator Effectiveness = 0.50 11 6 Thermal Efficiency of a "Basic" Gas Turbine Cycle with Frictional Pressure Losses = 0.12 12 7 Thermal Efficiency of a "Basic" Gas Turbine Cycle with Frictional Pressure Losses - 0.20 13 &e Flow Area 8 Flow Rate/vs. Temperature for Air and Helium Regenerative Gas Turbine Cycle 14 9 Thermal Efficiency vs. Plant HP. Air as Working Fluid "Basic Cycle" T1 = 15000F 16 10 Thermal Efficiency vs. Plant HP. Air as Working Fluid "Basic Cycle" T1 -l 1200~F 17 11 Thermal Efficiency vso Plant HP. Air as Working Fluid "Basic Cycle" T1 = 900 ~F 18 12 Thermal Efficiency vs. Plant HP. Helium as Working Fluid "Basic Cycle" T1 = 15000F 19 13 Thermal Efficiency vs. Plant HP. Helium as Working Fluid "Basic Cycle" T1 = 1200~F 20 14 Thermal Efficiency vs. Plant HP. Helium as Working Fluid "Basic Cycle" T1 = 900~F 21 15 Velocity Vector Diagrams for Air and Helium. a. Compressor Diagram 24 16 Velocity Vector Diagrams for Air and Helium. b. Turbine Diagrams 25 17 Nuclear Powered Closed Cycle Gas Turbine (Heterogeneous) 30 -iv

ACKNOWLEDGEMENTS A considerable portion of the work upon which this paper has been based has been conducted under a research. contract of the Engineering Research Institute of the University of Michigan with the Chrysler Corporation who have kindly given permission for its publication. The writers have received advice from faculty and staff in the direction of the work. Also, Eo M. Brower, J. L. Summers, and R. K. Fu, research personnel of the University have contributed to the calculations on which this paper is based.

1.0 INTRODUCTION Nuclear power plants, which employ high temperature nuclear heat sources coupled to closed-cycle gas turbine power generating devices, offer promise for application to light-weight power plants of relatively low power output. In conventional fossil fueled gas turbine plant design in that power range in which gas turbines are of interest, the closed-cycle is at; a disadvantage with respect to open cycle units in that the closed cycle combustor is a bulky and costly item which is required to replace -the simple internal combusti.on heater of the open cycle. In fact, compromise semi-closed units, in which the fuel is burned internally, and continuous make-up air provided to insure adequat;e oxygen for combustion, have been constructed to avoid the req uirement of a closed, heater) With a nuclear power plant, the heat source for the heat engine cycle must be a closed heat exchanger, and -the simplified combustcxr olf an open cycle is not possible in any case. Thus, one of the most- important advantages of an open.-cycle in a fossil-fueled planrt does not apply presently to a nuclear plant. Some of the advantages of the closed-cycle design which apply to both fossilfueled and nuclear plants are: 1e Increased capacity for given size and/or weight 2. Wider range of loads without substantial loss of efficiency since power may be reiduceid- by reducing pressure level without affecting temperatures, pressure ratios or speeds. 3. Reduction of size tends to control difficulties resulting from differential temperature expansions. An advantage of closed-cycle over open-cycle units, applying only to nuc-Jlear plants, is that containment of the working fluid allows direct; cooling of the reactor without the release of radioactive gas to the atmosphere An example of the advantage of nuclear gas turbines in general over nua.lear steam plants is that direcl cooling of the reactor with. the heat engine working fluid becomes feasible at high temperature without prohibitive press;reo A disadvantage is that fairly high temperatures (say 1200 - 1500 F) appear necessary to achieve efficiencies anrd capital costs comparable with steam plan-ts. Some of the disadvantages for the closed-cycle gas turbilne as compared to the open-cycle are: lo Water or air or some other coolant is required in the heat sink which is necessary for any closed cycle. 20 Higher internal pressures involve complicated design of heat exchangers. 3. Complex shaft sealing arrangements may be requiredo 4. Ability to react to rapid speed change requirements is impaired by the closed nature of the cycle~ -2

Escher Wyss, Ltd. has developed a closed-cycle gas turbine with external comnbustion, whereby the working fluid for the cycle exchanges heat with th.e combustion gases so that n.o combustion products come in contact with the internal surfaces of turbines, compressors, and. heat exchangers. The heat rejection for the cycle is through a precooler upstream of the compressorso With fossil-fueled unit;s, the advantages claimed for this mo di:fi catlon are: 1. Contaminated gases do not pass through. the mechanism~ Hence, corrosion is materially reduced. 2. Cases other than. air can be employed. with.;lout d.ifficul;lty. i. Several types of fuels may -be employed in'the comaus'tion chamber. The dl.isadvantages of an. externally-fi.red closed-cy' 1le gas turb.-ine are: 1o 7t.e gas heate,. when. using:fossil fuels, must be very large and made of high-grade materials since hot; gas at low pressure is a poor heat transfer medium. 2. The maintenance of t;he combustion chambe'ro containing heat -transfer tubes will be highE 3 For a gi:ven: maximum temperat'u)re, t - (?~:e:yc(le efficiernc-.V is reduced. by stack losses and degradation o.f ene:rgf in the -heat exchange r. With a nu.clar hea: t sour-ie, such a power pl.ant of fers advantages it...:h z.an...ot be achieved wit;h fossil-fueled firing..Th.e ltae ach.evrement o: a:tu.r.-ear heat source coupled. wi-th a closed-cycl-e gas tuarb:ine is d.epern ient largel-y upon thbe suc-cess by whidch materials can be developed so:th}at. high tempezrati;res9 whidch may'be geRne.ated. by f.ission, can tbe transfe:r~.eA l direc tl`iT -ao a s-ciA.ab a:le'worSking fluid., INuc.lear properties of the f:L.ld, as well as hermo.i. ami properties a-re mportant i.n ordIer to prod-.c_ e, tlie hi.ga:..est percentage o: useW work in terms of the heat power gean.erra tel Y r.l...lea issio:os. P ossitilif+es ex:ist for a c l~osett,:,cy;ic~-le gas turbine wor:kirag fluid e,: t-a't.in heat di.rec;lXy fr:om reactor fuels, provid0in'g s -uih wrorkingf f lu:..d. doe.s o-J t becs. irrafaleii to the e-xt;enrt -t.hat major biological shieldifng is requir,.ed fcc,-,`h<- po wer planr~.. as well as for the nuclear reac toro a. numbe:r of pl.an.ts, ranrgi.ng from a`00 hp resea:ch unit to a 127,`00 kwT plarn f:or cent+ral s'ta'tion duty, have'been built using fossl 1uels as thne ener.rgy sou:rre.2 Recently a number of organizations in conjunction with.!. t+ihbe A'.tomic Energy Commission and the U. So Maritiime Administration have under'taken preliminary studies to delineate the engineerLng parameters and. evaluaate ech.:}ni. cal andk economic:feasibilities of nuclear powered. gas turbines. FEecerti: publications have presented drata and. information on open an. clbosed-c(.cjcle systems with potential nuclear heat sources for application to packaged reactors, and large and small gas turbine reactor plants.

Nuclear powered closed-cycle gas turbines can be coupled with a heterogeneous fuel element nuclear heat source from which. the workinLg fluid extracts heat and flows directly t;hrough the closed, cycle gas tur bine system arnd then r;et-u+rns *to the reactor. Alternatively, a second promising -type of nuclea-r reactor fo:r t;his purpose would be the high temperature liquid metal homogeneous fueled reactor of the general type bei.:ng d1eveloped by Brookhaven.Natio:nal Laboratories in which the gaseous working fluid heat exchanges with circtul.a-tfi1n-g fissioning fuels. This paper endeavors'to discuss specific examples of'othi, a heter:ogeneous and a homogeneous reactor in eonjunlt;ion wi4; h a closetI-cycvler gas turbiTne s Although there are numbers of variations for bothne.s and homogeneou.s high t~emperature reactors, specifi>c examples of eatch are dniscis'. ssei _n';erms of t;:he th.lermodynamic- parame ters and the resq1ui_-emrerT*^s for- VFe-r e ffci-e t cony:r sion of heat to energy in sucfih a plar.t A:n a+tt+e'mpt' is rsmade *' o d,e:.1,e'ae the trends of the variations in. efficiency, size,'weightL, anrliJ -os't oaf.closed c, y'cle gas tuarbine systems as affected, by working fluid,.:.p:em ra'..tu:re, p:r essu-r e, a:nd. plant size, an-od. to afford rou.gh compariso:nls betweenr close-ec'vclv- gas'rit b;ites an.d alternative Teat engine tpes which might'be'ti;l. zed., An application as a packaged power plant generat.-ling electi, cal powe anl5d process steam in locations where the elect:rical Power requt me.si.Treme_'j-s a.-... in ho range from 5,000 to 30,000 kw, and where steam riequlrements are in he range of 10,000 to 70,000 lbso per:hour at 250 psig, has been. selecteu: for osc.ss:on, While many other applicat;ions a.:rle possible, a, ni.uclear: powere; i lose' i(Jc;:le; ga's turbine power plant of t,:his t.ype has tremendous techni cal a.nd. -;;or" mic poten,tia rin many remot;e locations of th.e world whe<^re:fossil f-uel. (-cosT s ei:e:hi gh, or where multiple uti.lizations of thermal heat; can.,e made simu!i'ane 1,sy,' An example of such a muiltiple purpose reactor ewolTd be a thernIeal pla:.: or a petroleum relfineK/.y where reg.xquirement-l1s are for c.hiemi cal hea?+- saacp, U i aT, process steam, aend elect+,,rical power. Such> an instalao wot a- inst-ala.o: o as.o p-rio'te a cornvenient sour(ce of nucleaar radiation, if such were of use in'-.., p-?ocesso 1.1 Simplified. c;lose"d-Cy:,lle Oas 11uib7ine Power iPlanr.'t Wi.t>17i a.'ulea.. ea. So:rce A schbemrzatkic.'flow diagram represen-ting a tIy ic -al sst-em for a n ci.ea. heat sourcle coupled(. t.o a closeri-cycle gas t1,_,rbi..ne is given in,. i tgr.xe I. nro.e operation of this cycle i.s as followso',"he wor-kirn. 1-l. Ji.. se- ve's as t;he reactlor coolanat, acntd cIrTerS't;he r-actQor a4: a g.ivnJ- ~:r'FSS.. e a'rk. teF:.mperature, As it flows tJ:roulgh. the rac.:-or' I. eXa..:s,he c.aa; generate.. y thrf fission process so t.hlat'tfh,c oeut, le. gas leaving t[he eat souarce is al;t a t e. mpera.ture sui table for e. JiT'c ieJt ga;s g, rts-lir operat ion,. Tibe.gas seravi.n.g as t-he worki.npg:fn lut is ex'pafi;. a...d ih.hg-,: pressur.e S.turbineI. wl] hich pr. ovides teh, m c..ari.a eergy.o:.d.v lw, a'r i high p:ressure compressor unitso S'tartir.ing eqdi. prme.: t an:.' au>t:ti.iam' p ele c-'trical gene.ratingl equipmenit might well be atb:ache.: to t;hi s saf-at:f e.:" gas leaves t:he high pressure tu.rbine amnid is e:.xpandr,dei a se.con.:! t.ime in a low pressure turbi ne t.o provide the maxzimum of, mec'Itan rl-aia vTo w-k f;r di -lvi.ng an elect:rical generat, ing unit,. Upon leavinASg the low pr-essr,?e:!rf.:-;irne, th.l.e. gas exchana.ges:heat in a recuperator w-'.f.t t:he gas st-,ream. rre —c',lir. io r t0 reacitor.'The gas leavi:ng the recuperator is the.n at t"h.e hOLresr I:.,:ss ume level of,the system; In order to close the Cylde g it: iS rn:essary to-, extract h.eat from the woroking fdluid prlor to compressiaon, izn a p'c tociLer..e gas is cooled to as low a temperature as is practi.canil.ay possiti.e b-y an external coolant such as wTater or air4 O.The cool.ed gas -c her-re r:rs into a low pressure compressor unit~. Tio ac'hi.e.v-e max:imu.m!,::Q'- le, et:fi,.er..,:y, an intercooler between the compressors seems d.rsi.rahbl e."'T'he::roeoiE. gas -4-j

5 - Sr o Li _/ - I __LLJ -.

then enters the high pressure compressor and is compressed to a pressure adequate to overcome the pressure drops through heat exchangers, ducass, valves and fittings, and reactor, as well as the expansion in the turbines~ The high pressure gas again flows through the recuperator, exchanging heat with the gas exhausted from the low pressure turbine so that the temperature of the high pressure stream is raised to a level satisfactory for the operation of the nuclear reactor, and suitable for minimum required heat addition. With such a nuclear heat source, heat sink, coolers, recuperators, turbines and compressors, a number of process va:.iables inrvolving nurlear and thermodynamic parameters must be considered to optimize tb:h. system. The general equation for the thermal efficiency of a closedl.-cy-le gas turbine plant of this type is: 2 tR 1T [1 - (r PR) j c (PR/2 -1) t TR {1 - r^R + rlr 1 -(p PR) + _R (PR ~ -T -1 ) The meaning of the symbols is listed under Nlomenclature. It is noted from this equation that for equal component efficiencies the overall thermal efficiency is not- a function o:f pressure level., power out;put;, or gas molecular weight. It is a function, however, of press;.ure:ratio, -temperature ratio, and the ratio of specific heats*t of whe workicg flul.i i e in;flluence of these various factors as well as th'.e effe-t of puressure and power levels upon component efficiencies, anrd -herce, upon overall thermal efficiency, is discussed. furtJ:her in a later sec!-ior..o - If a compressor temperature ratio, say. —, were used 4 nstead of pressure ra.tio, the above equation becomes: TRrT -T (p)Y > -TRc c2 ( 1/2 TRh Th 1' R +'1R!T 1 - (rpY) TRV +? -l (T 72 + Ic 11) It is then noted that the dependence of lth on y is only through rp. is the ratio between required pressure ratios for compressor and turbmrie to overcome the duct and heat exchanger pressure losses. -6

2.0 CLOSED-CYCLE GAS TURBINE PERFORMANCE CHARACTERISTICS In the design of a closed-cycle gas turbine system there are to be considered not only the choice of fluid and pressure and temperature levels, but also the question of cycle arrangement. Generally, there is the possibility of using high pressure-ratio cycles with a minimum of regeneration, or alternatively low pressure-ratio with maximum regeneration to achieve high efficiency. This choice constitutes an unresolved issue, since conventional plants of either type are being and have been designed and constructed. However, the closedcycle, in which a maximum mean working fluid density is to be obtained at low pressure ratio, and in which the size of heat exchanger components is drastically reduced by the high fluid density, appears to favor the low pressure ratio, highly regenerative cycle. Also regenerative cycles, at their own optimum efficiency pressure ratios, which may be as low as say 3.0, attain the highest thermal efficiencies. For these reasons, the regenerative cycle has been emphasized in these studies. As previously explained, Figure 1 is a flowsheet of the cycle under consideration. Included are heat source, heat sink, single intercooler, regenerator, compressor, and turbine. A possible variation would be the inclusion of reheat. For the cycle of Figure 1, the thermal efficiency is given by equation 1.1 previously shown. Using this relation as a basis, it can be shown that for equal component efficiencies and available temperature limits, a larger ratio of specific heats, under these assumptions, results in: 1. Reduction of peak efficiency 2. Movement of peak efficiency point to lower pressure ratio 3. Increase of the sensitivity of efficiency to pressure ratio. Thus, helium or any monatomic gas under these conditions appears less favorable than air or nitrogen or any diatomic gas, which in turn, are less favorable than a polyatomic gas such as carbon-dioxide. These trends are illustrated in Figure 2. It is noted that the effect increases at low temperature. For example, at 900~F maximum cycle temperature, for the component conditions specified on the figure, there is a decrease of about 2% points out of 22 between carbon-dioxide and helium. At 15000F, the decrease is only about 1 point out of 41. Of course, for a given capital investment in machinery, a higher overall efficiency may be obtainable with helium, for example, than with air because of its more favorable heat transfer and/or duct pressure drop characteristics. Thermal efficiency increases strongly with temperature while volumetric flow decreases. It is for these reasons that temperatures at least in excess of 1200~F seem necessary. Regenerator effectiveness, and heat exchanger and duct pressure losses, are also important parameters to which cycle efficiency is very sensitive. The effect of these is shown in Figures 3 to 8 for both a diatomic and a monatomic gas. The calculations are based on a "Basic Cycle." The component efficiencies and revelant data for this cycle are listed in Table Io It is noted from Figures 3, 4, and 5 that at 1200~F, for example, for the "Basic Cycle" for air, there is a decrease of about 7% points in maximum thermal efficiency (i.e: from 0.345 to 0.275) as the regenerator effectiveness drops from 0.93 to 0.75, and a further loss of about 4 points for an effectiveness drop to 0.50. At the same time the pressure ratio for -7

-8FIG. 2 THERMAL EFFICIENCY AT OPTIMUM PRESSURE RATIO AS A FUNCTION OF THE RATIO OF SPECIFIC HEATS. BASIC GAS TURBINE POWERPLANT WITH REGENERATOR AND INTERCOOLER. RATIO OF SPECIFIC HEATS FOR WORKING FLUIDS VARIED. INLET TEMPERATURE- 1500 F, 1200 F, and 900 F REGENERATOR EFFECTIVENESS=.93 DISCHARGE TEMPERATURE= 90 F DUCT AND HEAT EXCHANGER PRESSURE LOSSES=.07 CODE: -TURBINE AND COMPRESSOR EFFICIENCY-.85 —. —- -TURBINE AND COMPRESSOR EFFICIENCY-.80 - - TURBINE AND COMPRESSOR EFFICIENCY-.75 ___________!T_ qc.85 T= 1500 F 0 ~7 TT NC =*.85, T= 1500F _ 4 qT= 7C =.80, T= 1500 F o z Z I~~~~~~~~~~~-? rT= rc =.85, T= 1200 F -- -- -- ~~ -- ~~-2T —c 7?=.75, T= 1500 F _X T 5C =75, T= 1200 F lT=7 Cc =.850, T= 900 F'?7T =.75 T= 1200 F 1.3 1.4 1.5 1.6 1.7 k- CCV

0 I 2 3 4 5 6 7 8 9 10 0.4 - - 0.3 - - - C..) 3 4 6 -3 9 - PRESSURE RAT 10 FIG. 3 THERMAL EFFICIENCY OF A BASIC" GAS TURBINE CYCLE WITH RECUPERATOR EFFECTIVENESSI0.93, AND FRICTIONAL PRESSURE LOSS =0.07

-100 - 2 3 4 5 6 7 = f = - 0.3. 0... A-, o o 0 I- 1 - 1 I: I i I I I//I_!_ i,.e o,! I PRESSURE RATIO FIG. 4 THERMAL EFFICIENCY OF A "BASIC" GAS TURBINE CYCLE WITH RECUPERATOR EF CTIVE N ESS.75 WITH RECUPERATOR EFFECTIVENESS= 0B75~~~~.

0O'0= SS3N3A1I13_i —3 80.LV3dnlr33.I HlIM 3313A 3NI8lnl SV9,,ISV8,, V JO AON3131J-3 -IVl 3HI'91 OI.LVI 38n$SS3Ud 01 6 8 L 9 9 ~ O r - - - - u0,mO...00 ~t..0 I'I 1G. I 0 = =T=. I. ~' "-~'~oo...;T'El

l'O-S3SSO1 3unSS3&ud "tNOI..l.IJ HJ.IM 310A3 3NI8l.I StV9 1,,01SV8,, V 10 AN3131.. 3 1VW3H1 9'91 01.1v 3fnss3ad 01 6 8 L 9 G I o 0 ~~~~~~~~~~~IIIII I II I ~I I II 0<F T -~-~~... -I,.~~~~~~~, rrl,, - m = _ "_N 1~~~~~~~~~~~~~~. ~~~~~~~~~~~~~~~~-g t___ ~0=z mS -7 -_ I I I I I' I I I i -Zl7:f - -T - I- -- -_ - -[ -'-'~ --- — "/-!

-130 - 2 -a3 0.4. — /_ jt._?' 74 —_- 0.2 W-A- -:-.-d- -. PR ESSURE RAT10 0.WITH FRICTIONAL PRESSURE LOSSESU.. FIG. 7 THERMAL EFFICIENCY OF A.BASIC" GAS TURBINE CYCLE U.~~WT -RCIOA /RSUELOSS0

.22:i 22 1 FIG. 8 FLOW RATE AND FLOW AREA VS. TEMPERATURE FOR AIR.221-~ 22 1 w w w w AAND HELIUM REGENERATIVE GAS TURBINE CYCLE.20 61000 20 100 HELIUM (VOL- FLOW) cr c;:Q~~i Icr18 9 Zi I 1: REGENERATOR EFFECTIVENESS =.93 I o,~a TURBINE - COMPRESSOR ADIABATIC EFFICIENCY.81 z c AIR (VOL SINK TEMPERATURE 90' F 2 16 80 FLOW) Q- 16800D LO AREA _I's Z w~~~~~~~~~~~~~~~~~MS FLOW FAT I.18 o 18 90 ~'JN LUVOLUME FLOM z co N A (VOL AIR (AREA)'812 600 6 o cl AIR (MASS 0 LOW)~N LO) LU0 0 50N -- FLO AR <~~~ 0L HELIUM H'1 c7 MASS FLOW4RATE TT~~~~~~~c Q~~~~~~~~ VOUM FLO RAT 0 < HELIUM (AREA) 00 CE g ci' 0 04O cL200 4 8 2~0 < I L U < LL o c: C C 0 >4 00 4 02 900 1000 1100 1200 1300 1400 1500 MAXIMUM CYCLE TEMPERATURE - 0F

maximum efficiency shifts from about 2.7 to about 5.5. At 1500~F the efficiency losses are similar. The loss of efficiency for a drop in regenerator effectiveness from 0.93 to 0.75 is about 7.5 points out of an initial value of 0.42, and the further loss to 0.50 effectiveness is 6 points. The maximum efficiency pressure ratio shifts from 3.3 for 0.93 regenerator effectiveness to about 8 for 0.50 effectiveness. Figures 3, 6, and 7 show the effect of an increase in frictional pressure losses around the cycle. This is measured as an increase in the required ratio between turbine expansion and compressor compression ratios. For the "Basic Cycle" of Figure 3, this ratio is 0.93. If it is decreased to 0.88, there is an efficiency loss of about 2-1/2% points for air at 15000F, and 3 points at 12000F. A further decrease of the ratio to 0.80 reduces the 1500~F efficiency by an additional 4% points, and the 12000F efficiency by 5 points. Thus, the effect becomes proportionately more serious for lower temperature cycles. In the process, the maximum efficiency pressure ratio is increased although not so substantially as for reduction of regenerator effectiveness. In both instances, the effects with helium are similar. It will be noted, however,`-5;-.at she loss in efficiency for helium with increased frictional pressure drops is much more serious than for air. Figure 8 shows the variation of relative mass flow, volumetric flow, and cross-sectional flo~upath area witch temperatu..re assuming a given allowable Mach number for both air and helium. It is noted that there is approximately a 2 7 to 1 increase of mass flow requirement between 1500 F and 900 ~F, a 1.9 to 1 increase in volume flow (as measured at the maximum cycle temperature) andc a 2.2 to 1 increase in flow path area, measured at maximum cycle temperature, for a fixed Mach number for either air or helium. It is believed that the last index is probably most closely related to tihe size of the various components. While helium holds approximately a 6 to 1 advantage in mass flow over air for a 12000F cycle, there is a disadvantage in volume flow of 1.2 to l However, due to a much higher sonic velocity, the advantage in flow path area is 2.4 to 1. This probably is the most accurate indication of relative machine sizes. The foregoing does not include the effect of power output and pressure level. In an actual plant, the efficiency which may be attained. by the turbomachines is a function at least of Reynold's number, flow path dimensions (leakage effects, etc.), and Mach number. These, in turn, depend upon the power output and pressure level as well as, to some extent, a compromise between mechanical design factors (thermal stress problems, etc.) and optimum flowpath considerations. A generalized study cannot attempt to predict the precise efficiencies which will be attainable for any given' set of design conditions. However, an attempt can be made to indicate the trend of efficiency with operating pressure and power level, and thus, to guide to some extent the choice of the pressure level in the design of a plant for given output. A study of this type has been carried out at the University of eichigan for plants utilizing air and helium respectively for power outputs ranging between 600 hp and 60,000 hp, and pressure levels from 45 psia for a 3:1 pressure ratio open-cycle to 1000 psia for a closed cycle over a range of inlet temperatures. The resulting curves of efficiency versus output are included in Figures 9 through 14. It is noted that there is a fairly sub-15

.52.48.44.40.36 g 700 psio.32.28.24.20 FIG. 9 THERMAL EFFICIENCY VS..16 PLANT HP. AIR AS WORKING FLUID..12 "BASIC CYCLE".12.08 Ti = 15000 F.04 010 103J PLANT HP 104 105

.52,48.44.40,36 45 A 0 PS.32.28.24.20.16.12 FIG. 10 THERMAL EFFICIENCY VS. PLANT HP AIR AS WORKING FLUID.08 "BASIC CYCLE",04 T1 12000 F 0 a 2 103 PLNOH 0 PLANT HP 0

.52.48.44.40.36.32 00.28 45 lp\t.24 400'Ps'.20.16.12 FIG. II THERMAL EFFICIENCY VS. PLANT HP.08 AIR AS WORKING FLUID BASIC CYCLE.04 T1 = 9000 F ~1O2 0 PLANT HP 10 10

.52.44.40 - 36 t- I ~~~~~~~~~~~Pg=45 psia Pg 100 psila Pg = 400 psia ~~~~~~~41 ~~~~~~~Pg =1000 psia Pg= 700 psioa.20 -16.12 1.0's FI~~~~~~~~~~~~~~FG. 1 0.4 ~~~~~~~~~~~~~HEL/UW l 0 18A SIC CYClonG to I'r'1500'0 F LANTV Hp /OQ

.52.48.44.40.36.32 45500.28 ~ - — ~00, Ps,.24.20.16 FIG. 13 THERMAL EFFICIENCY VS. PLANT HP.12 HELIUM AS WORKING FLUID "BASIC CYCLE".08 T, = 12000 F.C4 0 102 103 PLN H 0 PLANT HP

.52.48 FIG. 14 THERMAL EFFICIENCY VS. PLANT HP HELIUM AS WORKING FLUID "BASIC CYCLE".44.40 TI = 900~ F.36.32.28.24.20 e.16 SIG.12 20s\.08.04 0 102 103 104 10H PLANT HP

stantial reduction of efficiency for any power level with increasing pressure. This is attributed to the fact that although Reynold's numbers are increased, with the higher densities achieved by increasing pressure (for a constant velocity and flow rate, Reynold's nZumber is roughly proportional to the square root of pressure, since density is directly proportional and. hydraulic diameter inversely to the square root of pressure), the influence on the efficiency of the decrease in machine size, through all factors, as abstracted from reference 4, is greater. There is also a substantial reduction of efficiency with plant output. It is noted, for example, that a decrease of thermal efficiency from about 0.395 to about 0 35 is shown for a 1200~F air plant at 20,000 hp as maximum cycle pressure is increased from 45 psia to 1000 psia. Also a decrease from 0.395 to about 0.335 is shown for a 1200~F air plant with 45 psia maximum pressure as the design output is decreased from 20,000 to 1000 hp. The corresponding decrease at 1000 psia is from 0.35 to 0.275. Similar results are obtained at the other temperatures, and also for helium. While it is certainly possible to design plants for a range of efficiencies for any of the conditions investigated, it is believed that these results are fairly representative of the trends which will be encountered. The cycle calculations are based on the conditions of the "Basic Cycle" (Table I) except that the turbine and compressor efficiencies are varied according to the operating conditions, and a 3j radiation heat loss is' included. Foir the larger sized units these results are based on the compressor and turbine velocity diagrams shown in Figures 1r; and 16. eilus, the vach number effects are the same for all units. The variation of blading loss coefficients is taken from reference 3. The overall compressor efficiency is taken from reference 4 for conventional atmospheric air machines for unit-s of various size. Corrections according to the blade loss coefficients for the various Reynoldis numbers are th-en applied to the basic efficiency of the applicable air -nit. Since tulrbine efficiencies should. varry accordirng to the same relation-s, it was assumed t.ha-:fo:r the large units, turbine and. compressor e:fficiencies could be considered equal,-* at least from the viewpoint of indicating the trend.o An arbitrary cut-off of component efficiency at.90 was assumed regardless of Rey:n.old's nmnber since it was felt that pract;i.cality consieb:rations role out machines of extremely large dimensions. For the smaller machines, centrifugal compressors or positive displacement units such. as'the Lysholm-type were assumed below 2000 CFMO Efficlencies were extrapolated from reference 4. It is realized. that there is a wide -variation in. efficie.ncy for machines of these types, but again., the results should be illustrative of the trendso Turbine efficiency for the smaller lnits was estimated by direct computation of t;ypcal machines using the blade coefficient data of reference 3, and computing reasonable leakage, windage, and bearing losses. It was assumed that a 5-inch. diamet;er, 5/8-inch blade height represented an approximate practical minimum. If flow rates were insufficient to provide:full flow for such a wheel at the velocities corresponding to the diagram (Figures 15 and 16), it was assumed'that all velocities were reduced proportionately as required. * This is an attempt to balance the higher efficiency inherently attainable with a turbine for a given blading design agains't the likelihood of a reduction in number of stages for mechanical simplicity of this high temperat;ure component. -22

TABLE i!' EAS I C C-'(ILEtL" PAPA1ETEBS Descriptio n rDa':a Effective Averbage Adlia'bat'ic''ur i.nr and Compressor Efficiency 0 85:Ree culpel'.a'l;,or 1Eff;e c' lti'v: S S0. 9T:.a' ti.o I'Turbine E.xtpa:nsion iFat,.Lo to Comnpr.essr Comrpression. Ratio 0O9O,7ompresso.t.a ti,,. ^,l O ninim.m. Ga.s.'e:mp9era'[.ure 90 ~F Cv(i.le irvltu'.des re:cu.pecra,,or precoole[; afr.d, si.ngl i...- r-ooler -23

-24AIR - 14 STAGES PRESSURE RATIO PER STAGE = 1, 082 o4.5V.. ACu= 1965 - 5 94.30 u= 385 HELIUM- 16 STAGES PRESSURE RATIO PER STAGE = 1071 ACu=491 2358 u=962.5 FIG. 15 VELOCITY VECTOR DIAGRAMS FOR AIR AND HELIUM. o. COMPRESSOR DIAGRAM

-25AIR - 9 STAGES PRESSURE RATIO PER STAGE = 1. 121 10 7 =~~~0 $ 434.2 4 u= 481o 107 - ACu-694. HELIUM - 21 STAGES PRESSURE RATIO PER STAGE = 1. 050 u 962. -53.8 ACu=1070 FIG. 16 VELOCITY VECTOR DIAGRAMS FOR AIR AND HELIUM. b. TURBINE DIAGRAMS

30O SELECTION OF WORKINlG FLUID As is apparent from the foregoing discussion and the curves of Figures 2 through 14, thermodynamically there are no really conclusi've reasons for ormitting any of the common gases from consideration as closedcycle gas turbine working fluids. There are, however, certain factors which may be worth mentioning. From the viewpoints of availability, cost,; and "state of the art,," air shows great advantages. On the other hand, it exhibits a certain disadvantage because of its oxidizing characteristics at high temperature. IT'hermodynamically, it appears that light gases with low values of the ratio of specific heats are very favorable from the viewpoint.; of minimizing heat exchanger cost s. Preliminary calculations have indicated that from tthis viewpoint hyd rogen is most favorablewith helium next, and carbon-dioxide superior to air or nitrogen. A comparison of pumping work per unit heat transferred for various fluids is gfiven in reference 5, whicsh agrees substanttially with. this conclusion, However, hydrogen might creat-e material problems at high temperature due uo hydrogen embrittlement as well as explosive hazards. Nitrogen or air may create difficulties due'to nitriding. SThe properties of low molecular weight and ratio of specific heats, which make a gas particularly advantageous from the viewpoint of heat transfer, result in a very low attai:nable pressure ratio per stage for turbomachinery, if It is Cr-onisid -ir A that tthe work input pe.r. stage is limi.ted by mechanical considerations according to the relation: P.. 1 WST k-l!M. i PR T1 k. 0 1 755 where tt:he work per stage, WST,is constantr. H:owever, for conven'tional axial flow compressors, it, is not usual to design to the stress limits~ Ra ther +the Mch number may be limiting if acceptable efficiencies for gas turbine plant components are to be obtained, Thus, if it is assumed t;lat Mach numbers atre to be held constant between machines for dif.ferent fluids, t,,he allowable -velocities for a low moleic ular'weight, high ratio of specific heats fluid such. as helium will be very high, and centri fugal stresses may be t:he limi1ting factor. In this case, the number of compressor st ages to achieve- a given pressul.e ratio appears in a typical case to be only slightly greater than for air. As was shown in Figures 3, 4, 5, 6, and 7, the desired pressure ratio for a high ratio of specific heats gas is relatively low. For these reasons, it appears that the number of stages of turbomachinery necessary for air and helium plants,respectively, may be approximately equal, al-thou.;gh at a givern pressure the diameters of the helium machine will be considerably less (see Figure 8). Such an equality in number of stages may well apply fairly closely for any gas, The selection of a fluid on thermodynamic grounds alone would involve a detailed economic analysis of all factors including both mach.inery costs and efficiencies. However, there are other factors to be considered including corrosion, availability, toxicity, handling hazards, etc, as well as induced radioactivity in the case of nuclear plants. If hydrogen, helium, nitrogen, -26

carbon dioxide and air are considered, only hydrogen and helium have nuclear properties so that nuclear inter-reactions'wll not create radiation problems in the power plant. If nitrogen, carbon dioxide or air is t;o be used in a system in which the working fluid extracts heat directly from the reactor fuel, these materials will inter-react with neutrons and present problems of a radioactive working fluid in the power plant cycle. If the system is to be designed so that air, nitrogen or carbon dioxide does not see neutrons or inter-react therewith, it becomes necessary to provide secondary heat transfer loops where a fluid extracts t;he fission heat from the reac tor and flows through suitable external heat exchange equipment to transfer the heat to the power plant working fluid.. In this case, it becomes necessary to sacrifice a portion of the available temperature in the heat exchange equipment. Therefore, in orxder to attain the same temperature of working fluid, it would be necessary to operate fthe reactor a.t higher -temperatures than when direct cooling is used. Since the maximum temperature attainable in rnuclear fuels is limited by the types of materials of constructions available whic:h have satisfactory n.uclear properties, it appears advisable to consider the d.evelopment of nuclear powered closed-cycle gas turbines in which the working;fluid can. be used as the medium to extract the heat di:rectly from the fissioning fuels. Therefore,, ohe of the working fluids worthy of consideration and development- is high-pressure, high.-temperature helium. However, even with, helium there are certain aspects which must4 be considered. Helium 4 does rnot:have a measurable neutron captured. cross-section, and. thus, no coolant activi+ty is produced~ iNatural helium, [however, contains 1 3 x 106 weight fract;ion of helium 3 which has a captured cross-secti;on of 5000 barns. -This reaction is an np reaction in which'tritium is produced., which decays'with a soft beta emission.'this presents no shielding problem, but -:creates a hazard from wviewpoints of biological injectiono Also, commercial helium is reporrted -to contain up to 700 ppm o:f carbon.-dioxide, 80 ppm of argon, and 2800 ppm of nitrogen. Consequently, it might well be nlecessary to purify commelrc:ial helium before its use in a direct-cooling sys-t.emo Aside from i:nduced radioactivity, there is also -th.e possibtility- of t-he coolani gases affecting the nuclear design of the reactors. owevFer, altbhough in some cases th.e thermal neutron absorptioin cross-section per atom is fairly su.bstant-ial (1.8 barns for nitrogen), even at the highest practicable pressures, for zhose gases which have been discussed, the nu-mber of molecules is so low that the main effect is only as a void. -27

4.0 TYPICAL NUCLEAR POWERED GAS TURBINE POWER PLANTS It appears in general that gas turbine power plants in a moderate output range (perhaps 1000 to 60,000 kw) must be provided maximum cycle temperatures of the order of 12000F minimum, if they are to be economically competitive with steam plants. Two general types of reactors possess possibilities for nuclear heat sources which are capable of achieving temperatures in excess of 1200~F. Each of these reactor types have a number of alternative choices in regard to neutron energy range, type of fuel, type of reflectors, type of working fluid,, type of controlss, type of reflectors, and types of moderators. Two specific examples of these general types of reactors are discussed herein. Basically, th-ey have been elected. in order to bring out the variables and optimi za-tioils of an integrated system in terms of those problems which must be resolved through aggressive research and, development. The types of reactors in general which are discussed are: 1. A high temperature heterogeneous gas-cooled reactor in which -the gas is used as the working fluid for the power plant cycle; and 2. A liquid metial homogeneous fueled. reactor in which the liquid metal fuel exchanges heat with the working fluid. of the power plant: The limitaatvions of each of 1these reactor types is corti-ingerv,,4 upon the maximum temperatu, r ar wh.ich can kbe obtainalble in the worki:ng fluid as related. io the str uc tulral mxra.tJenias employed irn tl.F reacto-;Or arnd thte t;tal v-ol.)me anrC wTeight of the re-actor systemQ Wi.Lh dule- c.-onsic-erat,1i on to sucb st.Jru iCural. materials, i t appears that; wiIT.:h presentlr known materials; a working fluid. temperature of 1300F might _,be achieved. in economically feasible hete:rogeneous reactor or es5 in the case of a circ:llating liqui. metal homogeneous rt.eactgor syst-ems, a possib.ilityr eixists that a temperature approachling 1500ooF inr" tjlIe working; flui.d might b.ie attain-ed beFc-lause of th-e elimination. of high. workinz;Ig fbli.i'd presst;.re ir_ thhe reac:,etr. The maximum poessu-re; which. can be aoh:eieved, for a hb.e-.-eo;en —ogeous fueled. reacto;>: in which th"e working flu.id ext;rac-ts fissio:rn. heat; S- y flowing through. the:reac'tor fupel element) is limited primaril.y to the.esi:n of the pressure vessel reuAkired to contain the het,erogeneous:reac;toxr (core, and. the me. b.od) b?7 which th.e wessel wall 1temperat;ures can be maintaine;-d at a temperatuL;-e within the creep and tensile limits of'the materials selecFted. Irn gene:ral, for the requirement of a packaged power plant, with electrical reqtui.rement ri; excess of about 8000 kw9 pressures on the order of 1000 psi appear desirable to limit the reactor heat tzransfer area requirement. When the electrical power requirements are less, corresponding reductions in the pressure of the working fluid can be allowed with sacrifices to rweight and size of t'he heat engine compone:rje.,t. 4.1. A T'11yipical Heteriogeneous Cas-Cooled Reactor With a Closed Cycle.Gas Turbine for Generation of Electrical Power and Process Steam A complreh.ersive review of a great numTb>er of different tyTpes of heterogeneous reactors employing gas as a coolant s heas eer made, othe.results have indicated that a specific type of heterogeneous reactor and power plant combination have possibilities of measured improverments irn he -28

next several years for applications to packaged electrical generating power plants. In the present analyses, an emphasis has been pl.~:ed upon attainment of high thermal efficiencies in terms of major re. ductions in nuclear fuel consumption. Emphasis has been directed toward achievement of a composite reactor design which will initially give working fluid temperatures no less than 13000F, with possibilities of gradually increasing the temperature through developments in new and improved types of structural materials for heterogeneous reactor fuels, Although the system which is described herein is by way of example only, it is believed that the specific reactor discussed might be developed in the foreseeable future to a promising small-size, small-weight, and inexpensive electrical power generating facility which has capabilities also of generating steam at 250 psig. Trhe reactor type herein discussed, includes a heterogeneous fuel element assemblage located in a high pressure vessel and employing helium as a coolant which is also the working fluid for the clpsed-cy.cle gas turbine. It is a thermal, moderated reactor. Reactor design is patterned somewhat after a helium-cooled program originated at the Oak Ridge National Laboratories in 1947, in which it was proposed to extract fission energy by means of helium gas and use such gas as the working fluid in a power plant. It is proposed for this example that electrical power output be 20 megawatts, and that at the design point it be also possible to furnish 10000 pounds per hour of process steam at 250 psig saturated. The cycle is illustrated in Figure 17. The steam is to be procured through bypassing of the high temperature portion of the recuperator. It is possible to obtain varying amounts of steam in this manner. If electrical output is considered constant, then it is necessary that reactor heat load be varied toaccommodate these varying amounts of steam. The approximate capabilities of the system are listed in Table II. In order to minimize the required size of the reactor core, it is necessary that the coolant pressure be a maximum. Premieed upon helium as a working fluid whose outlet temperature is 1300~F, a. working pressure of 1460 psig has been selected. If'the entering gas stream is used to cool the pressure shell containing the core (see Figure 17), then the returning gas stream temperature is limited. by the pressure vessel design.. To allow a reasonable design of carbon steel for this vessel, a maximum returning temperature of about 750OF has been chosen.:For a maximum cycle temperature of 1300~F, such a return temperature is consistent -with a gas turbine cycle, operating at a pressure ratio of 3 (slightly in excess of the maximum efficiency pressure ratio to reduce flow rates) and employing a feasible maximum of regeneration. If a maximum cycle temperature greater than 1300F were employed) it would be necessary to reopen the question of suitable return temperature, and reach a compromise between pressure ratio, regenerator effectiveness, and the requirements of the reactor pressure shell. Under the preessure and temperature conditions Stated above, it appears possible to achieve a design of heterogeneous fuel elements employing highly enriched uranium in the thermal energy range so that the resultant core is relatively small. Calculations which have been made for uranium dioxide (90%'U235) with a stainless steel matrix'and clad in a -29

NUCEUA POWERD CLOSED CYCLE GAS TUININ ETrOUW GA COOLED ~CTOt WITH AUXIUAY STEA erlIRTION roR POCSS UAND lSPCl HEAT 1440 n1 P CLCLOATMOIM FOI OLWU AS WORKII.O L'D 1r00 IFCaTR I NE AID COQIINlN1 WAIATIC IrFrlCI1NCY ~ 14 /10.000 L/nHR POCS TEAI AT ZOO mG AT NUUN FLOW RATE 9 4. Lb/*EC TIElilL IrFIcCInc Y % AT \ Po TIEINdL rFFICIENCY IZEO SIrAM) ~ % NIEC:ITN 1 rlrlE CTIVENS (ZNO T o MA) *g ~.,~"HP T-znt, A, %o - 11 [1 I I nrru~sI~c I I 6 ~ I I r C~lmrm P 11 H ~~~~~~~~~~~~~~~~~~~~~~~~~Lo,,4,,-, I..r' 762 pole?t44~~w COF 4 I ~. _3 I so IF I TO4 4F to If TIIIC EoLCTRIC POW" to MISABTT $ REACTOR ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~O HEAT 41)7 Psla O 11 " I O~~~~~~~~~~~I 751'F ~ ~ ~ ~ ~~~~~~~~~~~~~~~~~ — -- H —..- 3LO P l i.Ft q!Ol Iu~c ~ on rrIF g Is 1 CO~lmLT l TAN 4e P III I 1300'. N [~~~~~~~~~~~~~ TO b~~~~~~~~~~~~~~~~~~~~~~~l ~~~~~~~ ~mm al, I I.,.BI {,~.I #.ll I I~~~~~~~~~t0 ~r ~ ~ ~ ~ ~ 90P AO Iqa,,,,.,.,. N Plllr CM I 00 IF FRED MWA IN W 1REULIT I rtreOLL I1 I rFo~r I Comm up" WATERcoou COO I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~tlUWI ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ A TyNIINE (.rnlm~EI I wrt m) 4F0'F iSO 1 tI tur~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 4n0.r1 LO~ Fig. 17.

TABLE II HETEROGENEOUS HELIUM- COOLED REACTOR. CLOSED-CYCLE GAS TURBINE POWER PLANT PERFORMANCE 20 Megawatts of Electrical Power Process Steam at 250 psig Saturated Helium Flow Rate - 91.3 pounds/second Maximum Cycle Pressure and Temperature - 1460 psig, 1300~F Plant Thermal Reactor Heat Reactor Helium Efficiency (based on Steam Production Power Inlet Temperature electrical output Pounds/hour Megawatts ~F only) 0 62.5 767 0.320 6,100 64.5 750 0.310 10,000 6o.5 63.3 0. 305 50,000 80.0 629 0.250 134,500 107.0 394 0.188 -31

suitable stainless steel, with a suitable moderator such as carbon, beryllium, or zirconium hydride indicate that an acceptable design can be achieved. Recent developments by the Sylvania Corporation have indicated that zirconium hydride offers excellent promise as a moderator in thermal reactors. Assuming that the maximum temperature at the maximum point of neutron flux can be no greater than 18000F, calculations indicate that a flat plate-type of fuel element is required in which the fuel bearing section is 30 mils in thickness when the concentration of uranium dioxide in a stainless steel matrix is 25%. In considering the heat conduction-convective heat transfer problem from the heat generated in the uranium to the working fluid at the maximum point of flux, the stainless steel cladding for the fuel bearing section is determined to be about 10 mils. Arranging the flat fuel plates into a rectilinear assemblage for a fuel element so that proper mass velocities of the coolant flowing between the fuel plates results in effective convective heat transfer, the core size resulting therefrom when employing zirconium hydride as a moderator, is about 3v x 3' x 3-1/2v long. The power level of the reactor is maintained by control rods composed of hafnium and stainless steel which replace fuel lattice positions. It appears possible to arrange the control rods in such a fashion that changes in re-activity result in only small fluctuations of the spacial flux. Employing zirconium hydride as a moderator, and carbon as a reflector, with an adequate thermal shield, i't appears that a pressure vessel to contain these reactor components can be designed to a maximum pressure of 1460 psig when the wall temperature of the reactor shell is no more than 7500F. 4.1,1 Heterogeneous Reactor Control In a packaged tJy~pe of power plant generating electrical power, a fluctuating electrical power demand can be expected. Also, there can be expected a fluctuation in steam requirements for processing or space heating. The reactor design, therefore, must take into considera tion the changes in power output. As a first consideration, it may be worthwhile to consider controls whereby the reactor power level is maintained essentially at some constant power level, while electrical power and steam production are changed by by-passing of turbines and oversizing of heat sinks. However, this approach would require burning of nuclear fuels without useful work, and as a result, higher fuel. costs and losses in thermal efficiency. The increase in the power level of a nuclear reactor is limited only by the rate at which the neutron flux can be increased, which in turn is determined by reactor kinetics and safeguard. Decreases in power output are similarly limited although the socalled "delayed neutrons" prevent a very:rapid shut-down. It is desirable, therefore, to develop controls which automatically reset the power level in the reactor by means of turbine throttle for rapid load change, or change in the working fluid pressure level by pumping fluid into or out of the system for long-term changes. Since a reactor of the type here considered has only a small negative temperature coefficient of reactivity, considerable development must be conducted upon dynamic control rods, -32

shim rods, and reactor scram controls which have capacities to absorb the entire neutron production of the fuel elementso It will be necessary to locate such controls in the fuel lattices so that a fuel element might be displaced when a control rod moves into position. Xenon. build-up tends to provide -this type of reactor with positive temperature coefficients so that rises in temperature suddenly might result in an. excursion. Therefore, a major parameter which requires resolution for this type of a gascooled reactor, lies in the kinetics of the system. Assuming that throttling or pressure level control and reactor heat power can be correlated with the power output of the power plantj, Table III presents the results of calculations for nuclear data for the reactor. Table IV presents the results of the thermal data which has been calculated for a heterogeneous gas-cooled reactor with the nuclear characteristics mentioned. 4.1.2 Power Plant Design Premised upon these general conditions, Figure 17 indicates a schematic flow diagram with the temperature arind pressure conditions listed at various points, for a nuclear power closed-cycle gas turbine power plant generating 20,000 kw of electrical power and. producing 10,000 pounds per hour of 250 psig steam at sthe design point. As shown in Table II., increased steam production can be achieved at commensurate sacrifice of thermal efficiency and increase in reactor power. The closed-cycle helium gas 4tursbine plant employed with this reactor is similar to the "Basic Cycle~" of Figure 1l However, a process steam generating system has been added by dividing the recuperator into two unitis, and incorporating a controlled by-pass around the high temperaitulre portion. through a steam ge nerator~ The quantity of steam to be generat;ed, is cont:trolled by" that portion of'the entire helium stream which flows through`the by-pass line. It is assume1 th.ae the'r.eliuam flow rate and the temperature and pressure conditiorns leaving the reactor, wi.ll remain constant so that the turbo-compressor set will be undistu.,rbed by changing steam requirements, and electrical output will remain constanto Under these conditions, a change in steam generation will affect the helium temperature entering the reactor. %~Lils has been fixed to a maximum of about 7500F because of pressure shell requirements and this temperature applies approximately at the design point (temperatures and pressures at various cycle points shown on Figure 17) where 10,000 pounds per hour of steam are generated. If the steam generation is increased to a maximum (high temperature recuperator by-passed completely) the steam generation is 134,500 pounds per hour, and the heliium temperature returning to the reactor 394~F (see Table II). TUnder these conditions, reactor power must be increased to 107 megawatts~ The log mean temperature difference in the reactor is, of course, increased by the drop in reactor inlet temperature. However, assuming fuel element temperature cannot be increased for reasons of thermal stress, etc., it is probable that the leaving temperature would decrease to some -3-,

TABLE III HETEROGENEOUS GAS-COOLED REACTOR NU'CLEAR DATA Description Data Fuel elements - type Flat plate Fuel material Highly enriched Uranium 90% U235 Fuel matrix 25% U02 - Stainless Steel Thickness of fuel bearing section 30 mils Thickness of cladding 10 mils Total thickness of plate 50 mils Number of plates per element 17 Dimensions of fuel bearing core 3v x 3' x 3'6" long Number of elements 400 Core diameter 3 Core height 376" (same as fuel elements) Nuclear characteristics - Calculated fuel loading as 90% U235 135 kgs.7Neutron Energy Epithermal Average neutron flux I of average 1.35 x 1015 neutrons per square cm. per second Maximum neutron flux - T of maximum 2.12 x 1015 Ratio of max. flux to average flux (radial) 157 Ratio of max. flux *to average flux (transverse) 1o 40 Cold start-up K/K 0.40 AK/K at shutdown 0.05 Calculated neg. temp. coefficient rK/K/~F 4 x 105 Calculated fuel burnup 25% Ratio of U02 to stainless steel 0.60 Ratio of zirconium hydride moderator: to fuel at reactor startup 1400 at reactor shutdown 2300 Maximum temperature in fuel bearing section 1800~F Maximum temp. of stainless steel clad.ding l6.5QFj Shim rods stainless steel-hafnium Control rods stainless steel-:hafnium Dr~ive mechanisms electrical servo mechanisms * Includes loading for critical mass and one years inventory.

TABLE IV THERMAL DATA FOR HEAT EXTRACTION FROM HETEROGENEOUS GAS —-COOLED REACTOR Description Data Design point reactor heat power 66 megawatts Coolant flow rate 91.3 pound.s/sec. Design point inlet temperature of 738~F helium to reactor Outlet temperature of helium from reactor 1300 ~F Inlet pressure of helium 1460 psig Outlet pressure of helium 1442 psig Estimated coolant velocity 400 ft. per sec. Design point specific power 9,250,000 BTU/hr/ft3 of core Estimated convective heat transfer coefficient for helium in reactor 1,000 BTU/hr/sq.ft/~F Temperature of reactor shell 775~F Thermal shield material Carbon steel Moderator Zirconium hydride clad in stainless steel Reflector material Carbon blocks, 12 thick, surrounding core and adjacent to walls of vessel -35

extent. Then to maintain electrical output, some increase in flow rate would become necessary. It has been assumed that these secondary effects are relatively small. This is certainly the case for more moderate steam production up to, perhaps, 50,000 pounds per hour. At this steam rate, the reactor power is only increased about 20% over the design point~ The turbine and compressor efficiencies to be applied have been selected from the results of the previously discussed study for the applicable fluid, temperatuxre, and. pressure. In this case, the effective average of adiabatic turbine and compressor efficiencies is 0.84. Friction pressure losses, heat sink temperature, and pressure ratio are as assumed for the "Basic Cycle," Table I. A recuperator effectiveness of 0o95, based on zero steam production, was assumed because of the extremely high pressure~. Determi.nation of the economic feasibility of this value would require an involved. optimization study. A net heat loss of 3% of reactor power through insulation was assumed. At the design point, the helium at 1460 psig and 750~F enters the reactor and flows along the vessel walls and the thermal shield, so that the reactor vessel wall temperature will not exceed 775~F. The helium gas enters a plenum chamber, and. flows downward through and between the flat plate fuel elements,'which are provided with channels 50 mils x 3 inches wide so that helium gas velocities of about 400 ft. per second are achieved. qThe calculated ratio of maximum-to-average flux transversely along the flow of the coolantJ in the fuel channel is about lo57. The temperature of the helium leaving the system will be 13000F. It is estimated that the pressure drop for the heliLum through the reactor itself is about 20 psig. 4.1.3 Power Plant Shielding Requirements There are four main. sources of radioactivity which may present problems in shielding'the closed-cyl.e gas turbine power plant. These are: (1) Induced radioactivity in helium and its impurities; (2) Induced radioactivity in particulate matter carried in the gas stream; (3) Recoil of radioactive atoms due to knockoff; (4) Fission product escape from the fuel elements into the working fluid. A brief discussion of each of these may be worthwhile. 4.1.3.1 Induced Radioactivity in Helium arid Its Impurities This problem was discussed in some detail in an earlier section of this paper. 4.1o.32 Particulate Matter Particulate matter in the form of dust, dirt, and scaling of equipment on start-up and shutdown might introduce particles into the gas stream which have high capture cross-sections, and offer gradually increasing problems in the power plant equipment. It -3&.

might be necessary to provide in. the d.esign, a system for continuously purifying a portion of the helium gas for the removal of particulates with suitable dust collectors, electtrostatic precipitators, etc, 41l.3.3 Recoils One of the ways in which radioactivity can build up in the working fluid st;ream is -through recoil atoms, There- are two types of recoil: recoil of an atom due to the emission of a particle; a:nd recoil due to a c.ollision withh a high energy particle. I:f the surfaces wit.hin the reac-tor are stl;eel or s'tainless s'teel, the most important reactioins are, of the first type in wh:ich. eit her an n-y7 reac'tion converts manganese 55 t:o manganese 56, which is a gamma emi tter, or iron 56 to manganese 56 through an n-p reactiorn. The first, reaction is possible because although most steels contain only i-2% manganese, the cross-section for t+his reaction is rela'tively high (12 barns). The recoil momentum is so low that; only t;hose a;,t+oms within about 10-Q cm of the surface can escape. Whi.le the cross-section for the second reaction is muich smaller, the number of iron atoms is larger, and, the recoil of the vrooron is su.fficientl t;o remove atoms wit'hin a3ou.t; 10-"'' m of tlhe sxurface,. It appears that t.:he proto:n're'oi.l of the i.oro ai oms ia s several times more i.mp ao;, andd.hence, the'type of steel used is rela:, i vel, urnii.mpor'tantr,'It is dio.f fficult t;o say whf-re th..ese rad.ioact:ive athoms will go. X-,rom a main.t:enance viewpoini':, the mrrargaese -'{:6 is not seri. ous, since it de:cays to reasonable valules w'i'thin abou:% 30 hlours,'The second general type o;f eac'tion (recoil du.e to coll..ision with. h igh3 energy pa-"': aiicles) occurs d ue 1to collision wi -t h.. i.gh eriEgy.e7;rons, arnd. will cause radioacti.ve ir'ot. (ch:t:ommxim, n.ickel,, and (cob.alt a.taoms to enter Pthe gas st:rzeam. Sonme of ~;:hese atomns contrairn long life isotopes, so c'r.?.aft -'?.he bac:k:.r:'ou;: levels w.ill. gradually bui.ld up in'th.ie pri.ma-ry' loop eq.uipmer.n'-. Ji.owever,. d.irect maint.enanct should. still be possit le alt:lough a reduction. of working time may7 Te neecessar-y, 4.1.3.4 F'ssion iProduucts Radioactivity may enter tJhe working fluid s-tream due to diffusion of fission product gases through. the:fuel element, or fractuaies ocurring in tlhe fuel elements themselves. Methods by which fission products can escape into tthe working fluid are: (1) ruptiure due to shock and thermal cycling; (2) pin holes in cladd ing; (3) diffusion; (43 melting of a fuel element. The cladding on a fuel element can rupture from thermal cycling and thermal shock.'The probabilities of ruptures occurring must be reduced to acceptable levels by proper fuel element manu:facture and operating techniques. Monitors must be providied, in the gas stream to -3 7

detect such ruptures, and permit corrective action to be taken before activity becomes excessive or dangerous~ Defects in cladding can allow gaseous fission products to migrate into the gas stream. Therefore, carefully controlled inspection and testi.ng of fuel elements prior to installation in the reactor is requisite. Gaseous fission. products can enter the working fluid stream by diffusion at high temperatures. As temperatures exceed 18000F, the rates of diffusion of gaseous fission products are unknown, and research and development are necessary. For the type of heterogeneous reactor described herein, it is conceivable that a fuel element or a group of fuel elements could. have tempera4-ture rises so rapid that melting would occur before the cont+rols could, operateo In the event of such a caast;rophe - two problems could result: (1) a sudden increase in radiation levels outside of the biologi.cal shields witht resultant depositions throughout the piping and components of the closed. cycle gas turbine; and (2) the necessity forx undertaking major decontamination of the facilities to permit direct maintenance. The decontamination of'the helium loop through the reactor system and,through the en'tire power plant would offer serious problems. Major developments must+ be undertaken to assure'that the mos-t credible accident would not result in tlhe mrelting of fuel e-le mer.nts,, Evaluation. of't1h..e co.,dtions ariLd p:rograms,of inves;igations and demonstration, which. a re re 2uisite for early achievement of a h.ete-ogenreous heat power reactor associated with a closed cycle gas 4;-turbTin.e in which'the reactor coolant serves directly as working -fluid can'be ach.ieved with. a nominal program of research, development, and demonst:r ation. In this ctase, -the major de velopment problems appear to lie in the nuclear reactor, and in use of helium at; a satisfactory temperautree and, pressure in a closed cy(cle gas turbine plant. t Some of the problems which require resolution are.: (1) development of suitable high temperature fu1el elemen't.s capable of opera'ting for long periods of time when. the fuel temperature is at 1800 ~F maximum; (2) Development of a fu.el-to-moderator combination which:reduces -the inventory of nuclear fuel'to a minimum, and reduces the cr itical c:onfiguration so that a high pressure reactor vessel car. be achieved; (') development of positive controls so that xenon buildup and changes of working fluids d.o not decrease temperature coefficients. The control problem for a'heterogeneous gas-cooled, reactor is relatively simple, if the react-or power level can be maintained independent of the fluctuatilng requirements of power output in terms of electrical and steam energy, In the event it is desired to achieve output control which resets the reactor power level, a major program of development for heterogeneous gas-cooled reactors may be required. -38

5.0 A TYPICAL LIQUID METAL HOMOGENEOUS FUELED REACTOR IN CONJUNCTION'WITH A CLOSED-CYCLE GAS TURBINE POWER PLANT FOR GENERATION OF ELECTRI CAL POWER AND PROCESS STEAM A homogeneous liquid metal fueled reactor employing bismuth with highly enriched uraniumior high temperature operation, offers a great potential for early achievement of a highly compact, high temperature nuclear heat source in conjunction with a closed-cycle gas turbine. Uranium metal is appreciably soluble in bismuth above approximately 400~C in. varying concentrations. Solubility of uranium, and also plutonium and thorium, in bismuth as a function of temperature is given in Figure 18. The data for these curves is abstracted from reference 6. As noted from Figure 18, at a temperature approaching 1,000~C, a uranium bismuth solution is obtainable where the ratio of uranium to bismutlh at the limit of solubility is about 10-15%. A uranium bismuth reactor system has been developed basically by the Brookhaven National Laboratories. Numerous published reports have been made available for this system. A number of private industries are presently undertaking the study of applications and modifications to a uranium-bismuth liquid metal fueled reactor for power generation. Recently, the Atomic Energy Commission arranged a contract with the Babcock and Wilcox Company for a demonstration. reactor called "The Liquid Metal Fueled Reactor Experiment - LMFRE." Studies which are presented in this paper suggest that through proper developments of materials, it might be possible within the foreseeable future to achieve a reactor operating at a ~temperature approaching 1,000~C. By employing inter-nal arid exterrnal reflectors, it is possible to achieve a power reactor which is small and compact. Tt is reported that such a reactort. in terms of a sphere, may have a critical radius on. the order of 7-8 inches.6 Recent developments in new types of carbons of high density and uniform structurye might serve as container mate:rial for such a. reacto:r. For the purposes of this paper, preliminary calc-ulations have been made for a system which circulates a uranium bismuth system to a critical configuration, thence to a heat exchanger where the fission energy is ex-tra cted to a gaseous working fluid, and the cooled uranium-bismuth is pumped back to the reactor (Figure 19 ). The specific system stud.ied 1.consists of uranium dissolved in bismuth at 15500F (reentering reactor), so that the maximum permissible concentration of U235 is limited by this temperature. TIhe urani'm bismuth solu.tion is circulated through a critical configuration in a cylindrical shell which is 18" in diameter by 2' high. Interspersed within the critical configuration are rods of carbon for internal moderation. Surrounding'the core is an external reflector of carbon. On the outside of the external reflector is a thermal shield which is gas cooled. Surrounding the thermal shield is a reactor vessel. Because of the inherent safety of a uranium bismuth core, it is visualized that the only controls necessary are controls for shutdown or scram. The nuclear data for the reactor under these conditions are given In Table V. -159

20 FIG. 18 URANIUM, PLUTONIUM, AND THORIUM SOLUBILITY IN cn 18 BISMUTH ffi / PLUTONIUM z z < 14 URANIUM 12 I-0 z W 10 0 w az ORIUM 4 2 C 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 TEMPERATURE - F

TABLE V NUCLEAR DATA FOR A LIQUID METAL FUELED REACTOR BISMUTH-URANIUM TYPE Description. Data Dimensions 3.5' dia. x 4'0 high Total fuel loading 33.5 kilograms* Design point power level of reactor 78.4 megawatts Average neutron flux $ aver. l:x1015 neutrons per sq. cm. per second $ max. - c aver. in spherical core 1.50 Type of internal reflector - moderator Special carbon Thermal shield Laminates of iron and carbon Fission product removal Continuous de-gaser - 90% removal per pass Neutron energy Epithermal * Includes loading for volume of system, and one year's inventory. -41

CALCULATIONS rNO COR 14LIUIAI AS N FLUI0 NUCLEAR POWlIED CLOM[D CYCLE "l TUAlINt T1l Ull A COMPREISO DImt.C EICIEt * O HOGOGINOUS SA"l COOUL R1rACTOI WITH AUXILIARY Str A HLIU O RAi MA * #. L/ OEA#C THIEMAL IrFFICIENCy.S 4 AT DSGPOT POMEAWTTS C IC OWER TIHMAL FFICIEINCY (ZEO, IEA}) * 40.SA O llTTS %LCTRIlL POR ECUPERATO EFFErCTIVIEIIS (ZO $TRLAM) ~ O 1IS,000 L/HR IlOClrSrAM AT P PM, SAT "WOUS FISSION m Ps IOoUCT As _OlS _ r., 1 7ilOGO I'111J41L 1 _~~~~~~~~~~~~~~~~~ BO' E Fig 19co CQNTROL MPeiI i II P OS iPiHni CM Pie 00. Bolt IF 90 IFr all sIf CARSOMN VI tLLECTl(C POWER1 L.P TURSIII I K IENEIAT I~T LI(QUID IMEIA FUEL TO $AS J ]o MEAT E CUAN 14MIrR Ir X TErlmA4.- - _ REFLErCTOR REACTOR MEAT 569 PSI sOREr 4 IIdITs weo IF S4OlLL~t01n PM "I PSoo GM Mo f ItS~~~~~~~~~~~~~~~~~~~~~~~0 *F IuLK go. It 1 I;~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~l Iql II I 040 P0I t ~ ~ ~ ~ ~ ——!Isi)N IEC10FEMION Rir C U TOR gmill 9011,101 NogI NO as, e PM No I $44uctrw~om, I PoI _,o- tm-, J,, aI in IF Im FEET WT MAN O l irlr TO PROCESS AND NEATIJNN 71111%~~~~~~~~~~~~~~~~~~- re.sos LB/"# FROIC;ES STE"AI CMUESTIE TAMK FWDO WA.w I~TUtl FM WAE PUMP C01, -- J edW TURBINME (I mrNo cm~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~- W —. Fig.:L9

The liquid metal fuel leaves the reactor core at about 1700 F, and flows into a specially designed liquid metal fuel to gas heat exchanger. The design which is presented herein consists of a U-tube bundle. The liquid metal flows through the shell side, and the high pressure gas through the tubes. A circulating pump of the vertical overhung shaft type is mounted directly into a channel of the heat exchanger. This pump circulates the fluid leaving the heat exchanger to overcome the pressure drop in the system. The calculated thermal data for the reactor is given in Table No. VI. TABLE VI THERMAL DATA FOR B4EAT EXThTACTION FROM HOMOGENEOUS (.,AS —- COOLED REACTOR DESCRIPTION DATA Design Point Reactor Heat Power 78.4 megawatts Bismuth-Uranium Design Point Coolant Flow Rate 10,60C GPM Design Point Inlet Temperature of Coolant to Reactor 1550 ~F Outlet Temperature of Coolant from Reactor 1 700 ~F Slpeci fic Fower/liter of Bismuth 1,000,000'BTU/hr/ft3 Design Point; Specific Power/Reactor Volume 7,000,000 BTU/hr/f t3 In the event it is possibtle to at;ainr a max:imum.reactor terperatulre of 1700F entering a liquid metal-to-gas heat exchanger, and to exhaust the liquid metal fuel at; 1550OF, it appears possible to design a suitable heat exchanger so that helium as a working fluid will enter tihe heat exchanger at 1015 psig and. 85 0"F, and will exchange heat with the liquid metal fuel so that the working fluid enterilJng the power plant is at 1500~F and at 1000 psig. The flowsheet for the gas t;urbine portion of the cycle is the same as that described for the heterogeneous reactor plant, and is shown in Figure 19 wit h. appropriate t~emperatures and pressures at the various cycle points. Because of the higher efficiency made possible by the increased turbine inlet temperature, somewhat higher levels of electric power and process steam generation have been considered. The design point has been set at 30 megawatts electrical and 15,000 pounds per hour of process steam at 250 psig saturated. Again, it is possible to obtain varying amounts of process steam, if the electrical output is held constant, by adjusting the portion of the helium by-passed to the steam generator from the high temperature portion of the recuperator. The steam output is increased, then, at cost to reactor power. The upper limit for steam generation for this arrangement is 185,000 pounds per hour if the high temperature recuperator is completely by-passed. The capabilities of the system in this respect are listed in Table VII.

TABLE VII HOMOGENEOUS REACTOR CLOSED-CYCLE GAS TURBINE POWER PLANT PERFORMANCE 30 Megawatts of Electrical Power Process Steam at 250 psig Saturated Helium Flow Rate - pounds/second Maximum Cycle Pressure and Temperature - 1015 psig, 1500OF Plant Thermal Reactor Heat Reactor Helium Efficiency (based on Steam Production Power Inlet Temperature electrical output Pounds/hour Megawatts ~F only) 0 73.3 894 0.410 15,000 78.4 852 0.383 70,000 96.6 701 0.310 185,000 134.8 387 0.223 In line with tdhe studies on probable compressor and turbine efficiencies previously discussed, the effective average of adiabatic turbine and compressor efficiencies was assumed as 0.85. The increase in. one point over the assumed value for t-he heterogeneous reactor cycle is du.e to the increased. volumetric flow rate due both to reduced pressure and increased temperatlre. The mass flow for the two plants is nearly the same. T'Ie greater power output of the homogeneous reactor plant is due to the higher thermal efficiency with the higher temperature. As in the previous cyc-le, a recuperator effectiveness of 09', based on zero steam proluc~tion, was assumed. Again., a 31 heat loss due to imperfect insulation was assessed.> Although the limit on helium temperature for the st;leam leaving the recuperator was set at approximately 750~F for the heterogeneous reactor cycle a maximum temperatu-ae of 894~F has been allowed. in this case. The reason:for the greater leniency in this respect, is that the:high pressure gas is heated in heat exchanger tubes and does not enter a large pressure vessel as it did for the [heterogeneous reactor systemo lThis lack of a high presslre vessel is felt to be one of the most cogent reasons for the possibilit<y of obtaining very high temperatures with this type of reactor. A liquid metal fueled reactor can generate heat power up to temperature levels which are limited only by the development of suitable container materials for the liquid metal fuels which are at approximately ambient pressure. For power output above approximately 20 megawatts, it appears necessary to provide external heat exchanger surface for the efficient extraction of the heat.'The liqulid metal fueled reactor offers promise for con.tinuous addition of fuel without shutt~ing the reactor down, and the removal of the major port;ion ofP the fission products during the course of operation. The reactor appears to be one which is self controlling, and sets a maximum power limit wihou the use of special controls, The maximum power is established by the concentration

of the uranium in the bismuth. Provisions can be made so that special scrams and shutdown controls are provided in the event emergency so requires. Such a scram control can remove the contents of the liquid fuels to a-critically safe geome try in the form of dump tanks or inventory tanks. It appears possible that by suitable heat exchange, helium at 1000 psig and at 1500~F could be achieved. At these levels of temperature and. pressures, and an electrical generating capacity of 30,000 kw, a closed-cycle gas turbine power plant appears to be an ideal system. As shown in Table VII, it is indicated that a thermal efficiency (zero steam generation) of 41% can be achieved by optimum selection of components and arrangements of systems. 5.1 Efficiency and Cost Comparisons With Alternative:Heat Engine Systems The applicability of the gas turbine cycle to nuclear power plants depends upon the thermal efficiency, capital cost, and weight which may be ac:hieved by this type of heat engine installation. versus the various alternatives which are available. The most obvious alternative is the steam plant. A comparison of steam plant efficiencies which may conceivably be obtained versus temperature for central station sized plants is shown in Figure 20 (all efficiencies consider zero stack loss for a nuclear plant). phe steam plant data is based on actual plant performance up to 1100~F. The high temperature portion of the curve is based on the predictions of reference 7. The gas turbine efficiencies are bassed onrl componenrz efficiency est;imations as previously described irn this paper. T';Phey may well be above the economical optimums. It is noted that even so, the steam plant efficiency exceeds the gas turbine at, any temperature up to 1500~F. 1However, -whereas the advantage is about+ I: points out of 39 at 9000F, iti is only about 4 point-s out of 49 at 1500F. For smaller outputs in t-,he range discussed in t;his paper, ioe. O up to 30,000 kw, it is believed that the economically justified. optimlnm for steam plants will be reduced. considerably,, so that if high temperature (say 1200 - 15000F) iean be made avai.latble, the gas -turbine plant may be the more efficient~. If the a;railable temperature is no more than 1000~F it appears that the steam plan.t will be considerably superior tot.Lh in efficien.cy and capital cost. As a fur-ther alternative to steam, if very high temperatures are avails able, there is the possibility of a binary Rankine cycle thermodynamically similar to the mercury-steam cycles which have operated for years in certain central station applications. Thhere is the possibility that a liquid. metal ight be boiled directly in a reactor or a portion of a circulating liquid stream flashed to vapor (Sodium is a possibility, mercury seems poor because of its high neutron cross-section; others such as potassium, rubidium, zinc, etc. may be suitable). the vapor would then operate a liquid metal turbine, condense against an ordinary steam system (or gas turbine) and be recirculated. The upper curve in Figure 20 has been plotted for such systems, assuming reasonable component efficiencies. It is noted -that at high temperature, the efficiencies are considerably in excess of either the steam or the gas turbine (only slightly below the Carnot line)o -45

- -.8.7 C,.6 Go.5 gU CpSUL.4!ai X +X t o | / I~f 1I) CALCULATED REGENERATIVE, REHEAT v.3 MAXIMUM EFFICIENCY POINTS 2) X CALCULATED STEAM CYCLE POINTS, (ASSUMED VARIATION FROM DIFFERENT PRESSURE LEVEL AT SAME TEMPERATURE) 3) co CALCULATED STEAM CYCLE POINTS, REFER-.2 / ENCE 7, VARIATION PER NOTE 2 4) * APDA STEAM PLANT, REFERENCE 8 5) 0 BNL STEAM PLANT, REFERENCE 9 6) y PHILO SUPERCRITICAL STEAM PLANT, REFERENCE 10.1 7) v CALCULATED HG-H20 OR NA-HG-H20 POINTS 0 /0 600 700 98) r KEARNY HG-H20 PLANT 500 600 700 800 900 1000 1100 1200 1300 1400 1500 TEMPERATURE - OF FIG. 20 MAXIMUM FEASIBLE EFFICIENCY VS TEMPERATURE VARIOUS HEAT ENGINE CYCLES COOLING WATER AT 70~ F

5.2 Cost Comparisons For the small-output (5 to 30 megawatts) nuclear powerplant of the liquid metal fuel reactor type, it appears that thermal efficiency is not of over-riding importance since fuel costs may be only a fairly small portion of the total. In a typical case which we studied., for example, the cost breakdown appeared very generally somewhat as below for a steam or gas turbine unit coupled with a homogeneous liquid metal fueled reactor, Fuel Burn-up - 30% Capital Costs Reactor and TtUranium Invent-ory - 40% Heat Engine - 30% Thus, it appears that the capital cost of the heat engine is as important as the efficiency. It is very difficult to obtain comparative cost dalta for the various systems. However, according to preliminary estimates which we have obtained from various manufacturers'it appears that in the range of 5000-30,000 kw, steam plants, exclusive of the boiler, cost on the order of $100/kw for a 900 F plant. Open cycle gas turbine plants appear to be considerably more expensive at about $250/kw for a plant of comparable efficiency. It would appear that the cost of hig:hl pressure closed-cycle plants should be consied-e;rably less since it is obvious +- ha+,) th;e weighbt; of material must decrease rapidly with increasing pressure~ For example, the weight of an axial flow turbo-machine for a given prFessu..e e level, flulid, and power output might be considered to be propor tional to (l/p)li - Iis is based on an examination of the casing a nd, its conside.ration as a pressure vessel. It is assumed that there will be a decrease in length prT-oportionate to the decrease in diameter. The weight of the rotor and of the heat exchange eq!:ipment should. a-t least varYr in the same (rd ieci-,on.'For machines of si.milar type in a somewhat similar size range, it seems q4.ite reasonable that- cost should reduce more or less wi-t;h, weigh:t, a:fter dzevelopmental costs have been assimilated.. If this be the case, t4he high pressure closed-cycle plant seems to hold promise of eventual considerably reduced capital cost, even as compared with steam plants. rThe results of a preliminary study which we have concduct-ed shows cost comparisons for liquid metal fuel reactor plants coupled both to steam plants and gas turbine plants for 20,000 kw output. If a very conservative temperature level. for this type of reactor is considered, of about, 900F in the heat engine working fluid, and an int.rermediate liquid metal loop utilized, then with present-day prices the steam plant appeared favorable (Figure 21). The comparison is shown between steam and open cycle gas turbine under'these conditions, with more than a 2:1 advrantage for steam. However, if the temperature level can be raised to exceed 13000F, and if high pressure closed-cycBle plants are assumed with the gas turbilne cost reduced f~rom present p:rices proportionate to machinery weight, then according to our preliminary study the gas tburbine showed atn overall advantage in power cost over the steam plant of about p1.5:1 For comparison, an estimated power cost for a sodium-gas ttulrbine combination plant i~s shown. The estimated power cost at, this particuLar outaput

is greater than that of the present-day steam plant by about 25%. At higher outputs, the very high thermal efficiency of the sodium cycle becomes more effective in overcoming the high capital cost of the machinery. 5.3 Weight Comparisons It appeared from our investigations to date, that the total plant weights for a homogeneous reactor steam plant, and a homogeneous reactor open-cycle gas turbine plant in the 5-30 megawatt range for the same types of duty, are roughly comparable. However, the weight of the closed-cycle plants is considerably less (a factor of about 3.5 at 30 atmospheres for 20 mw). It appears that weight-wise, a helium plant should show an advantage of about 1.3 to 1.0 over air, although there appeared to be little significant overall cost gain. The sodium boiler - gas turbine cycle appears to weigh about the same as the steam plant. These results, and the previpously discussed cost results, are extremely tenuous in nature, and are based on nutmerous assumption and over-simplifications which require detailed examination. -48

-49Z H -- M 8_j < vo N ar-W o - z -n< O cr Z 1 cr. < C L Or), 0 o 0o.. m z 0 ~~~~~~so II I II IItw'"' 0 1 Ii,,,,,, (D o 0 (D OnLo -OSODIUM VAPOR TURB INE-GAS TURBINE COMBINATION - 15000 F O w 6 L HIGH PRESSURE CLOSED CYCLE GAS TURBINE-13000F W 0- PROJECTED MACHINERY PRICES o CIRCULATING LIQUID METAL COOLANT a. WLL (0 _ 0o: OPEN CYCLE l GAS TURBINE _cl::3 9000 F-CIRCU — 0 LATING LIQUID METAL COOLANT o2 ur oDCD If V STEAM - 9000 F CIRCULATING LIQUID METAL COOLANT ) 0 r) N 0 - 0 POWER COST INDEX

6.0 BIBLIOGRAPHY 1 Kent;s Mechanical E:ngineer s Handbook, Power Volume, 12th Ed,, 1950, John Wiley and. Sons, Inc, pp. 10-23 to 10-25o 2. "Nuclear Gas Turbines," Mechanical Engineering, July, 1956, ppo 606-612o 3. Nguyen Van Le, "Report. on Loss Coefficients in Turbine Blade Passages," Gas Tu:rbine Labora-tory, Massachusetts Institute of'Technology, Cambridge, Massachuse tts 4e Kent's Mechanical Engineers' Handbook, Power Volume, 12th Ed., 1950, John Wiley and Sons, Inc., Table 9, pp. 10-38e 5. Gilliland, E R. o, et al, Chapter 10 of The Science and Engineering of Nuclear Power, Vol. II, Edited by Clark Goodman, 1949, Addison-V -sley Press, Inc. 6: Chernick, J' "Small Liquiid Metal Fueled. Reactor Systems," Nuclear Science and Engineering, 1, pp. 135-155 (1956)6 7. owns, J. E., "Margins:for Improvement of the Steam Cycle," ASME Paper Nio. 55-SA-76. 8. Atomic Povwer ievelopment Associates,"Developmental Fast Breeder Power Reactor" (Brocrhure) 9. Sengst4akin, D, J., a:ad.ur.xrham, E,, "T1,- he Liqouid. Me4tal Fuel Reactor, entral-Station. Power," Preprin.t 39, shiuclear Science and. EFngilneering Conrgress, December, 19;55o 10. Fiala, S.'. y, "First Commercial Supercritical Press.ure SteamElectr1ic Generating Unit for Philo 1Plant,' AS:ME Paper No. 55-A-137. -50 —

7.0 NOMENCLATURE Symbol Meani ng ~-tlh Thermal Efficiency rtic Adiabatic Compressor Efficiency'IT Adiabatic Turbine Efficiency Tip Ratio Between Turbine Expansion and Compressor TR Ratio Between Minimum and Maximum Absolute Cycle Temperatures TRc Ratio Between. Compressor Outlet and Inlet Absolute Temperatures'PF (4Cycle Compression Ratio k - l/k k Ratio of Specific Heats WSIT'Work per Stage of a Compressor 11 Absolute Inlet Temperatu re to Compr.e s sor K Effecti've Mul't;tiplica'tion Factor f:or Reac-iO:;or AK Change in E:ffectiv'e Multiplication Fact-or:for Reac-' tor -51