FAST BREEDER REACTORSSTATUS AND PROSPECTS

P. M. MurphyGeneral Electric Company

Breeder Reactor Development OnerationSunnyvale, California

In view of the importance of energy toa nation' s standard of living (Figure 1) itis not surprising that the question of howto provide a continuing supply of low costenergy is a matter of continuing concern tonational governments. It was because of theoverwhelming importance of energy that themore industrialized nations of the worldhave been led to invest in the developmentand application of thermal reactors (Figure 2)for electric power generation. It is forthis same reason that seven of these eightnations are now concentrating their advancereactor development on the Liquid Metal FastBreeder. The concept of a power generationsystem which produces as a by-product morefuel than it consumes is an intriguing one,particularly to nations which are now de-pendent on external sources for their fossilfuel supplies. This report traces the historyof the development of the Breeder system anddiscusses some of its incentives and pro-blems and how they relate to its develop-ment schedule.

The discovery of nuclear fission andthe concept of a chain reaction quicklyled to consideration of how this new formof energy could be used for electric powergeneration. The potential for breeding wasrecognized and led to active interest inthe breeder as soon as the cessation ofWorld War II hostilities allowed moreattention to be directed to peaceful apoli-cations of nuclear energy.

The first controlled fast reactor,"Clementine" was built in Los Alamos in1946 to provide a supply of fast neutronsfor experimentation. It was fueled withplutonium pins, cooled with mercury, pro-duced 25 kw of heat, and remained in servicefor seven years.

Shortly after the Clementine reactorwas put into service, the US Atomic Energy

Commission initiated design and developmentwork on two breeder systems to prove thetechnical feasibility of the concept to gen-erate useful power and to breed. One ofthese, the 'wlest Milton Power Breeder" waspursued by General Electric. The work start-ing in 1946, involved consideration of alter-native coolants, the selection of liquidsodium as the preferred choice, developmentwork on sodium technology and components,reactor core'design, and nuclear physicsexperiments. The nuclear physics experimentswere limited to examination of the epithermalneutron spectrum. That is to sav, thermaland fast reactors were arbitrarily excludedfrom consideration; only the middle groundwas studied.

The second of the two initial programswas assigned by the AEC to the Argonne NationalLaboratory. It started at the same time, pro-ceeded along the same line of liquid metaltechnology, but concentrated on the fastneutron spectrum.

By the end of the 1940's the data indi-cated that either concept would breed butthat the fast spectrum would be the preferredchoice on the basis of its prospect for asignificantly higher breeding ratio. The WestMilton Epithermal Power Breeder turned intothe sodium cooled naval reactor which poweredthe submarine Sea Wolf for two years with nomaintenance on its primary system components.Argonne's fast spectrum breeder concept wasturned into reality as EBR-1 which, late in1951, generated the world's first nuclearelectric power -- 200 kw.

This initial development momentum con-tinued for the next decade. The success ofEBR-l led to the decision in 1953 to build asuccessor, EBR-II with a hundred-fold increasein electrical rating to 20 MW. It went criticalin 1961, started nower operation in 1965, andis currently providing essential fast flux

10

irradiation in support of US fast breederfuel development. Commercial activity infast breeders started with the 66 MWe Fermiplant for which construction was started in1956 at the same time that the BWR and PWRprograms were getting underway with Dresden-l,Yankee, and Indian Point-l.

In the intervening years since the mid50's we have seen the BWR and PWPR businessgrow to the point where it now accounts foran important fraction of each year's newadditions of electrical generation capacity.(Figure 3) The capacity of LWR's already inoperation, under construction, or committednow total 69,000 MW. Fast Breeders, on theother hand, despite their earlier successes,have seen no further commercial application;activity is still limited to development andstudy programs. The reasons for this delayin Fast Breeder development are based onstrategic, technical, and economic consid-erations.

On the strategic side, the initial in-centive for the development of the FastBreeder was based on the philosophical virtuesof a system wvhich would generate power withvirtually no fuel input. This was a compell-ing reason at the time because it was notthen known whether the U.S. was a "have' ora "have not" nation with respect to uranium.(Figure 4) Our wartime nuclear weapons pro-gram had been carried out with importeduranium, only a negligible amount of uraniumore had been found domestically, and noconcerted exploration effort had yet beenstarted. As a further intangible, thereremained the question of priorities if shortsupplies should require that our reservesbe allocated between weapons and commercialelectric power.

By 1957, however, after EBR-II and Fermihad been committed, U.S. developed reserveshad risen above 200,000 tons of U308.(Figure 5) Prospecting had reached a rateof almost ten million linear feet of explor-atory wells per year, and each average footof drilling was verifying the existence ofmore than ten pounds of new uranium reserves.This answered the question of whether wewere a "have" or "have-not" nation and removedthe urgency for further uranium exploration.It also removed the emphasis from earlydevelopment of the Fast Breeder, at least asfar as conservation of fuel resources wasconcerned. The Fast Breeder in the U.S. was

left to stand or fall solelv on its merits asan economic producer of electricity with nocredit to be taken for being a patriotic citizenand holding back on its appetite for fuel.

Tinder these ground rules the Breeder faredrather badly. The technology of the sodiumcooled fast breeder is complicated and it pro-mised to be a long and expensive task to reducethe technology of 1956 to large scale commercialanplication on a competitive basis. To makematters worse, (or better, depending on yourpoint of view) the technology of the LWR'sappeared much more tractable and promised toyield attractive economics at a much earlierdate than would be possible with the Breeder.Next was the matter of size. Sodium cooledbreeders of the plant ratings under consider-ation in the late 50's and early 60's (atthat time a single unit as large as 500 NWstretched the imagination) were at a seriouscapital cost disadvantage with respect tofossil plants and LW4R's of equal rating. Itwas expected, however, that this situationwould improve with time as the electric utilitynetworks grew in size and the maximum ratingsof single units could be increased proportion-ately.

Finally, there was the question of theinitial fuel loads. Although it is technicallyfeasible to use P-235 (the only fissile mater-ial found in nature and hence the only fissilematerial available at that time) for the firstcore loading of a fast reactor, the resultingperformance and economics are well off optimum.This arises principally from the nuclear char-acteristics of U-235 as compared with Pu-239,the preferred fuel.

In brief, 91% of the collisions betweena fast neutron and a Pu-239 nucleus result infissions as compared with 85% for U-235.Further, each fission of Pu-239 produces 2.91new neutrons as compared with 2.47 for U-235.From the 2.91 and 2.47 figures we must sub-tract one neutron to keep the chain reactiongoing, leaving 1.91 and 1.47 neutrons avail-able for breeding (by arranging for its capturein fertile U-238), control, and inevitableparasitic losses thru capture in core structureand coolant. These latter numbers combinemultiplicatively with the 91% and 85% respect-ively to show that plutonium has a 40% advan-tage over I1-235 as a fast breeder fuel withrespect to the neutrons available for breedingper neutrons absorbed in the fissile fuel.This difference in performance is reflected in

11

a fuel cycle cost differential which leftlittle incentive for the development of aU-235 fueled Breeder in competition with theLWR. The plutonium in existence at that timewas weapons-grade material only and was fartoo expensive for commercial use even if itcould have been made available from thedefense program.

Thus the field was left open to the LWR'swith their more readily developed technologyand adequate supply of suitable fuel. The69,000 MW of LWR's now in service, under con-struction, and committed demonstrate how wellthis opportunity was seized.

The situation was far from static, how-ever, with one of the strongest forces forchange being provided by the LWR's themselves.They produce reactor grade plutonium as aby-product and, although this plutonium canbe recycled back into the LWR's to reducetheir requirement for low enrichment uranium,it has more value when used as Fast Breederfuel. The question of how soon and at whatrate this plutonium will become availablemerits some discussion here because it de-termines the timing of Fast Breeder intro-duction and the expected size and rate ofgrowth of the Breeder businesses in plantconstruction and in equipment and fuel man-ufacturing.

Initially the Fast Breeder growth willbe totally dependent on LWR plutonium. Asmore and more Breeders are added they them-selves will supply an ever increasing fractionof the plutonium needed for new Breederadditions. In the near term, therefore, themaximum rate of Fast Breeder additions isdetermined by the rate of production of plu-tonium by a LWR (% 0.18 tonnes ner year fora 1000 MW LWR) and by the initial Pu in-ventory requirement of a Breeder (% 3.0tonnes for an early design 1000 T14 unit re-ducing to IN 2.4 tonnes for an advanceddesign). The quotient of the two numbersshows that the early Fast Breeder additionsin any year can be as much as 6.0% of theLWR capacity which was in service four yearspreviously. The four-year qualification isnecessary, during a period of LWR growth suchas we are now experiencing, to account forthe inherent interval between the date of aLWR startup and the date when its by-productPu first appears on the market plus the timeit takes to fabricate a breeder core loadof plutonium fuel. As improved Breeder

designs are produced this rate can increase to7.4%. After the initial fueling each Breedersupplies its own Pu needs plus a surplus whichcan be used to start additional Fast Breeders.The measure of this surplus is given by thesystem's "Doubling Time" -- the time it takesfor a Fast Breeder to Droduce excess plutoniumequal to the initial needs of another FastBreeder of the same design and rating. Inmaking this calculation it is customary toassume that the annual plutonium production is"invested"in other Breeders as soon as itbecomes available rather than storing it untilthe entire quantity has been accumulated. Forcurrent Breeder design projections havingdoubling times in the range of 8 to 12 years,the annual Breeder growth from Breeder-producedplutonium would be 9.1% to 6.0%.

Figure 6 shows a orojection of U.S. FastBreeder growth on a plutonium-limited basisusing the nresent trend in LWR additions. Theconclusion is clear: as far as the availabil-ity of Dlutonium fuel is concerned, a viableFast Breeder business could be started in thiscountry early in the 1980's.

On a world-wide basis, seven of the eightindustrialized nations which have significantthermal nuclear power programs are also activein sodium cooled fast breeder develonment. Inthe case of England and France the programsare somewhat ahead of ours, a not unreasonablesituation when their incentives are examinedand compared with our own.

The close of World War II found the USwith a unique resource in the form of the OakRidge gaseous diffusion plant; this permittedour civilian power reactors to use slightlyenriched uranium fuel and, incidentally, pricedthe Fast Breeder out of the market until lowcost by-product plutonium became available. Theother industrialized countries of the worldfell into one of two categories. They eitherhad reserves of natural uranium or they didn't.They shared in common a lack of enrichingcapability and a need in each case to selecta nuclear onower development policy which wouldmake their own nation self-sufficient withrespect to fuel to the greatest extent possible.

The pattern which emerged (see againFigure 2) saw the uranium "have" nations(Canada, Britain, France, and Russia) concen-trate their earlv develonment on naturaluranium thermal reactors -- a course which wasconsistent with maximum self-sufficience with

12

respect to fuel. The "have nots" (W. Germany,Italy, Japan) being forced to import fuel inany event, hedged their bets by building bothLWR's and natural uranium reactors. Inneither case did the selected course promiseto be an ultimate solution to their nationalenergy supply problem. For the "have" nations,the high cost of power from the naturaluranium reactors and finite limitations on thesize of their uranium reserves providedadditional incentive for an early conversionto breeders: for the "have nots" the needfor self-sufficiency in fuel supply is stillunfulfilled. The Breeder promises the desiredsolution in each case.

The scope of the overseas effort on thesodium cooled breeder as manifested by theexperimental and demonstration reactors intheir program is shown, together with theU.S. fast reactors, on Figure 7. The figureidentifies the fast reactors which havealready been built, those xwhich are nowunder construction, and those which are inthe advanced talking stage. The UK startedconstruction in early 1955 of its DounreayFast Reactor (DRF). It reached full power(15 MWe, similar in rating to our own ERR-II)in 1962 and has been in experimental serviceever since. Construction is now well alongon a 250 MWe prototype (PFR) expected to gointo service in 1972. By that date theyexpect to have committed the construction ofa 1000 MWe unit for service by 11978.

France, although she started eight yearsafter Britain, has rapidly closed the gap.The Rapsodie reactor (24 MWt) has been oper-ating successfully for two years. A 250 MWeprototype (Phenix) is due for service in1973 only one year behind the British. A1000 MWe unit is planned for startup by1978-79.

Russia started work on breeders in theearly 50's. A series of reactor experiments,BR-1 (zero power, in 1955), BR-2 (100 kWt in1956) and BR-5 (5 MWt in 1959) generated noelectricity but gave a technical base tobegin construction in 1965 of BN-350, a 150MWe dual purpose electric power/desalinationplant at Chevchenko. Originally scheduled for1969 operation, BN-350 is now expected tostart up in 1971. The Russians are also re-ported to be starting construction of a 600MWe breeder at Bieloyarsk. Germany has underconstruction the 20 MWe KNK reactor experiment

and is well along on Dlans for a high flux testreactor (FR-3) and a 300 MWe demonstration plant,the latter intended for oneration in 1974.Work in Italy and Japan has not yet progressedto the reactor experiment stage although studyprograms and sodium technology and hardwaredevelopment activities have been carried onover a number of years paced by their ownlower level of thermal reactor installations.

In the IJnited States, develonment work onthe LMFBR has been increasing in tempo overthe past half-dozen years. It is now the high-est priority item in the US AEC's civiliannuclear power program. Equally important,however, is the industrial support which theprogram is receiving from the reactor ecuipmentmanufacturers and the U.S. electric utilitieswho have formed a number of cooperative groupseach aimed at the design, development, con-struction, and operation of a demonstrationbreeder in the 250 to 500 MWTe size range. TheAEC has recently announced its willingness toparticinate financially and technically in atleast one, and nossibly up to three, demon-stration nlant programs and has requestedproposals from industry groups. It is presentlyreviewing a number of such proposals. Thisprogram would have one or more DemonstrationPlants going into service in thL mid-to-late1970's, a schedule consistent with largecommercial plants going into service on aplutonium-l imited basis in the mid 80's.

3000U.S.

2000 CANADAGNP

(Dollars perCapS ta)

W. GERMANY- * UK

1000

50 100 150 200

COMMERCIAL ENERGY CONSUMPTION (Millions of Btu Per Capita)

FIGURE 1: GIP VS ENERGY COASUMPTIGA

13

FIGURE 2

YEAR

FIGURE 3: NUCLEAR PLANT ORDER HISTORY

1952 1955

YEAR

FIGURE 4: DEVELOPED U.S. U300 RESERVES

14

Present AdvancedPower Reactor ReactorChoice Choice

U.S. LWR LMFBR

U.K. Gas Cooled Nat. U LMFBR

France Gas Cooled Nat. U LMFBR

Italv LWR LMFBRGas Cooled Nat. U

Germany LWR LMFBR

Japan LWR LMFBRGas Cooled Nat. U

Canada 0)20 Cooled Nat. U __

Russia LWJR LMFBRD20 Cooled Nat. U

30

20

NUCLEAR GENERATINGCAPACITY ORDERED(We x 106)

10

ANNUAL NUCLEARGENERATINGCAPACITY (GWe)

25

20 DRILLING

of Feet)

10

1970 1980 1990 2000

YEAR

FIGURE 6: FAST BREEDER GROWTH

FIGURE 7

Operating

(Under Construction)

((Proposed))

15

U308 ADDED TO 60RESERVES(Thousands ofToss) 40

20

0

DRIuLLIW

ORE |f-

RESERV56 \

-,!-/tI-I

High Flux LargeReactor Test DemonstrationExperiments Reactors Plants

USA Clementine ((FFTF)) ((GE))EBR-l ((M))EBR-2 ((AI))FermiSEFOR

UK DFR (PFR)

France RAPSODIE (Phenix)

Italy ((PEC))

Germany (KNK) ((FR-3)) ((SNR))

Japan ((JEFR)))

Russia BR-1 (BN-350)BR-2 (60n)

1952 19ss 1960 1965 1970

YEAR

FIGURE 5: URANIUM ORE RESERVES ANn PODDUCTION TRENDS