towards an intrinsically safe and economic thorium breeder reactor

Energy Conversion and Management 47 (2006) 2781–2793

www.elsevier.com/locate/enconman

Towards an intrinsically safe and economicthorium breeder reactor

V. Jagannathan *, Usha Pal

Light Water Reactor Physics Section, Reactor Physics Design Division, Bhabha Atomic Research Centre, 5th Floor,

Central Complex, Mumbai 400085, India

Available online 29 March 2006

Abstract

Thorium does not have intrinsic fissile content unlike uranium. 232Th has nearly three times thermal absorption crosssection compared to 238U and hence requires much larger externally fed fissile content compared to uranium based fuel.These factors give a permanent economic competitive edge to uranium. Thus thorium is not inducted in any significantmeasure in present day power reactors, despite the fact that thorium is three times more abundant in the earth’s crust thanuranium. Uranium reserves vary from country to country and there is also difficulty in having equitable distribution ofuranium. Thus when 235U would get exhausted, perhaps much sooner in countries having limited uranium reserve, therewill be a need to switch over from the today’s open fuel cycle programme based on 235U feed to closed fuel cycle based onPu feed. At that stage thorium and (depleted) uranium would become equal candidates to form the fertile base. All eco-nomic considerations would have to be readdressed. The size and growth of the nuclear power programme based on closedfuel cycle would be dependent on maximizing the fissile conversion rate in those reactors. In this paper we reemphasize theprinciples and the details of the thermal reactor concept ‘A Thorium Breeder Reactor’ (ATBR), in which the use of PuO2

seeded thoria fuel is found to give excellent core characteristics like two years cycle length with nearly zero control maneu-vers, fairly high seed output to input ratio and intrinsically safe reactivity coefficients [Jagannathan V, Ganesan S, Kart-hikeyan R. Sensitivity studies for a thorium breeder reactor design with the nuclear data libraries of WIMS library updateproject. In: Proceedings of the international conference on emerging nuclear energy systems ICENES-2000, September 25–28, 2000, Petten, The Netherlands].� 2006 Elsevier Ltd. All rights reserved.

Keywords: Thorium breeder; Pu-seed; Two years cycle length; No refueling; Minimum control maneuvers; Safe and economic reactor

1. Introduction

India has vast reserves of thorium. Harnessing energy from thorium has been one of the main objectives ofIndian nuclear power programme. Though the mechanical and thermal properties and irradiation behaviourof thorium are better than that of uranium and the physical characteristics like ‘g’ of the man-made fissile

0196-8904/$ - see front matter � 2006 Elsevier Ltd. All rights reserved.

doi:10.1016/j.enconman.2006.02.008

* Corresponding author. Tel.: +91 22 2559 3739; fax: +91 22 2550 5151.E-mail addresses: [email protected], [email protected] (V. Jagannathan).

mailto:[email protected]

mailto:[email protected]

2782 V. Jagannathan, U. Pal / Energy Conversion and Management 47 (2006) 2781–2793

isotope 233U is the best in thermal reactors, the overall economic considerations and the problem of handlinghigh gamma activity of the short half-life isotope 232U (72 years) and its decay products have been hinderingthe use of thorium in a major way. Since there are no simple solutions to these problems it is necessary toevolve a reactor type, which is tailor-made for large scale use of thorium and in which all issues of frontand back end of the fuel cycles are comprehensively addressed. We had proposed a reactor concept called‘A Thorium Breeder Reactor’ (ATBR) which has the potential suitable for large scale induction of thorium[2–6]. ATBR with PuO2 seeded thoria fuel was found to give excellent core physics characteristics like two yearcycle length with nearly zero control maneuvers, fairly high seed output to input ratio and intrinsically safereactivity coefficients.

We would elucidate here the design principles, which were used to achieve these characteristics. The resultsof the fresh analysis with the latest 172 group library in WIMS/D format obtained from IAEA as part of theIAEA Co-Ordinated Research Programme on WIMS library update project would be presented [7,8].

A fuel cycle length of two years operation with no on-power fueling would help in easing out the problemsof refueling and also would provide the much needed lead time for reprocessing, prior cooling and refabrica-tion of fuel assemblies. In this context, we know that long fuel cycle duration would necessarily mean a largeexcess reactivity inventory and control thereof. This is normally achieved by increasing the feed enrichmentand correspondingly large control inventory. A variety of reactivity control means like soluble boron, controlrods and burnable poison rods of Gd or boron type are normally employed. In such a situation any removal ofabsorption such as inadvertent withdrawal of control rod or boron dilution would have the potential to causesome reactivity initiated transient. Further the neutron absorption in these absorbers results in loss of neu-trons and is not gainfully utilized.

One unique feature of nuclear energy, unlike energy from fossil fuels, is the possibility to produce new fissileatoms through neutron capture in fertile atoms when some fissile atoms are being consumed, i.e., nuclear fuelhas a means of rejuvenating itself while being spent. We would be able to prolong the nuclear energy extractionprocess if a large fraction of the fissile atoms that are consumed can be replenished with new fissile ones in thesame reactor and within the same fuel cycle duration. This would require delicate balance of the fissile depletionand production rates. For this exercise to be meaningful the process should last for as long a time as possible. Ifwe can eliminate nearly all external absorbers like control, boron and Gd and load equivalently large mass offertile material like thorium in the core, it would initially act like a pure neutron absorber but eventually turninto fuel after acquiring sufficient amounts of the in situ bred new fissile atoms. Elimination of external absorb-ers for reactivity control would mean that the core excess reactivity should be zero or very small all the time andfor as long a duration as possible. In other words the core composition of fissile and fertile materials should bedesigned such a way that the reactivity loss by fissile depletion is nearly balanced by the reactivity gain of newfissile formation at all times. This does not happen in the present day power reactors because the fissile depletionis normally at a much higher rate than the rate of production of new fissile ones. This is because the absorptionreaction rate in the fissile atoms invariably dominates over the fertile capture reaction rate. The imbalance ismuch more for higher feed enrichments. The rates may become comparable only after considerable depletionof the initial fissile feed, but the fuel may have to be discharged much sooner from reactivity considerations.Natural uranium reactors have the best conversion ratio. But the fuel residence time is as low as six monthsto one year and one requires continuous refueling for the equilibrium core. Since the core excess reactivity isvery low one cannot load any other fertile material like thorium in the equilibrium core using natural uraniumfeed. In enriched reactor systems, the reactivity fall due to fuel depletion cannot be matched by intrinsic fissileformation within the same fuel rods. Even in fast reactor systems where the ‘g’ value of Pu is close to three, thefertile to fissile conversion happens at a lower rate than fissile depletion in the main core. Fast reactors becomebreeder mainly through cumulative capture reactions in external blanket regions. Thus there is a need for a par-adigm shift in the type of reactor design where the reactivity loss due to fissile depletion can be intrinsically com-pensated by fissile formation in the same reactor.

As seen above, there are a number of factors that govern the fertile to fissile conversion rate. We shall dis-cuss as to how to maximize this rate and then attempt to match it with fissile depletion rate in the same reactor.

The integral fissile production rate in a reactor depends on the fertile inventory, flux level and the effectivecapture cross section of fertile isotope(s) for a given neutron spectrum. When fertile material does not containany externally fed fissile seed, one can achieve the best fissile conversion rate. In a thermal reactor system if

Fig. 1. Uniform distribution of fertile zones amidst fissile zones. Shaded—fertile blanket assemblies; blank—seed fuel assemblies.

V. Jagannathan, U. Pal / Energy Conversion and Management 47 (2006) 2781–2793 2783

unseeded thorium can be loaded in large measure one can achieve three times higher fissile production ratecompared to uranium, due to the intrinsic difference in their capture cross sections. The asymptotic fissile accu-mulation potential for thorium is nearly 14 g/kg of 233U in thorium. The concentration of 233U remains at thislevel for subsequent irradiation for very long time. In fast reactor system seedless thorium and depleted ura-nium can both be considered for high fissile production rate. The asymptotic fissile (fresh seed) accumulationpotential in a fast system is nearly 100 g/kg for both thorium and depleted uranium. Loading of unseeded fer-tile material is normally practiced in fast reactors outside the active core in radial and axial blanket regions.The disadvantage of such loading is that the absolute flux level seen by the blanket region would be 3–10 timeslower than that in the active core region. Hence fissile depletion would occur at a much faster rate in compar-ison to fissile production rate in blanket regions. It is advantageous to accommodate the seedless fertile mate-rial inside the core to achieve comparable flux values in the fissile seed and fertile zones. In order that theseseedless fertile zones see uniformly high flux level, it is necessary to distribute them uniformly in the entire corewithout clustering. A typical arrangement of accommodating seedless fertile zones in between the fissile zonesis schematically indicated in Fig. 1.

Fig. 2 shows a typical arrangement of the seedless thoria cluster. Loading of large volume/mass of fertilematerial without any seed is mandatory for enhancing the fissile production to be matched with fissile deple-tion. This would require much higher feed enrichment in seed fuel rods since the seed fuel rods not only pro-duce the fission power but they also provide surplus neutron flux ambience for the neighboring fertile (incoreblanket) regions. In order that this process lasts for longest possible duration it is necessary to maximize the

Fig. 2. Typical arrangement of seedless thoria cluster.

Fig. 3. Typical arrangement of seed fuel cluster.


initial feed in seed zone and somehow decrease the burning rate of the seed fuel rods. This can be achieved in athermal reactor system by two means.

The seed fuel rods are deliberately chosen to be of thinner dimension with sufficiently large fissile content.With higher seed content, the flux level would decrease for a given power density. The effective absorptioncross section also would decrease due to self-shielding of fissile atoms. In addition, the seed fuel rods are sur-rounded by thoria rods so that they would receive a much diminished thermal flux from the moderator regionoutside. Typical arrangement of seed fuel cluster is shown in Fig. 3. The power mismatch between the seed fuelrods and the fertile thoria rods is minimized by three means. The thoria rods are chosen to be not fresh but onecycle-irradiated ones. They are supposed to be exposed in the same reactor in previous cycle(s). A batch fuelloading is preferred. A typical optimized five batch fueling scheme with 72 assemblies per batch is illustrated inFig. 4, where the color scheme indicates the number of fuel cycles seen by a batch of fuel assemblies. In thelater discussions we would give a three batch fueling scheme with 120 assemblies per batch, which could oper-ate for two years. Depending on the fuel cycle duration and the location of these thoria clusters in the coreduring their first cycle, they would contain nearly 60–70% of the asymptotic in situ bred fissile content. Thethoria rod diameter is also chosen to be 26% larger.

Since they face the thermal flux directly from moderator, the relative power factor in the thoria rods isfound to be at least 0.5–0.6, when the seed fuel rods are fresh and contain nearly 15–20 times the fissile con-tent. With burnup the power share of thoria rods steadily increases due to further fissile accumulation andnear discharge burnup, it may even cross unity, i.e., the thoria rods which were initially receiving neutrons

Moveable ThO2 (19)

Fixed ThO2 (72)

I Cycle (Fresh)

II Cycle

(72)

III Cycle

IV Cycle

V Cycle (Last)

(72)

(72)

(72)

(72)

Fig. 4. ATBR core 360 seeded + 91 thoria fuel clusters.


from the inner seed fuel would become the supplier of neutrons and also take the major share of power withinthe fuel cluster. It is necessary to reach such high burnup of the order of 50,000 MWD/T when the cross overof power share-shift from seed to fertile thoria rods takes place, for the best physical characteristics like two-year cycle length with minimum external control maneuvers. The Pu seed content was therefore chosen to beas high as 20% and 14% respectively in the inner and middle ring of seed fuel rods shown in Fig. 3. The diam-eter of the inner and middle fuel rods was chosen respectively as 10 mm and 9 mm while the diameter of outerthoria rods was chosen as 12.6 mm.

To summarize, we may say that the new type of reactor should consist of thin seed fuel rods of very highfeed enrichment serving as torch or fire to ignite the surrounding fertile breeding zones. The seed fuel wouldgenerate surplus neutrons in addition to power. These neutrons move about freely in certain flux traps createdby vacating the seed fuel zones at regular intervals. D2O moderator is ideal for creating large thermal neutronflux traps. Unseeded thoria clusters should be placed in these flux traps to achieve high fissile conversion ratesduring the first cycle of these thoria clusters. The high fissile conversion is continued also in subsequent fuelcycles by placing the irradiated thoria clusters in between fresh seed fuel rods and the D2O moderator. Sincethoria rods occupying the outermost ring receive the major share of thermal neutron flux incident from D2Omoderator region, they also shield the seed fuel rods so that seed would not rapidly deplete. The ratio of vol-umes of the fissile breeding zones and fissile depleting zones, rather their masses, is typically 50:50. A batchfuel loading scheme with optimized loading pattern ensures that the thermal flux faced by every thoria clusteris fairly high and uniformly the same. When this is accomplished, the core keff is maintained just above unityfor as long a time as two years. In fact there will be an increase of keff up to middle of the fuel cycle when theconversion rate in fresh thoria clusters is maximum and thereafter it tapers down slowly. The results of thefresh analysis for equilibrium fuel cycle of the conceptual reactor ATBR with PuO2 seed are presented in thispaper. The reactor grade Pu contains 24% of 240Pu, which also helps in initial suppression of reactivity, and itsconversion to 241Pu also helps in maintaining flat reactivity with burnup. Reactivity change (fall) is higher forother seeds like 235U or 233U.

Table 1 gives the core design parameters of the ATBR core. In ATBR we adopt the steam generating heavywater reactor (SGHWR) model. A power level of 600 MWe or 1875 MWt is considered. The reactor considersvertical pressure tubes (PT) of Zr–Nb (2.5%) with ID/OD of 176/187 mm at a hexagonal lattice spacing of300 mm. Boiling light water coolant at a pressure of 70 bar is considered. The ID/OD of Zr-2 calandria tube(CT) is 204/207 mm with an air gap separating the PT and CT. D2O moderator is filled in the calandria tank of8400 mm diameter and 4800 mm height at normal pressure and temperature of 80 �C at hot operating state.

Table 1ATBR core design parameters considered for Pu seeded core

Reactor power MWe 600MWt 1875

Total core flow (tonnes/h) 27 · 103

Average heat rating (w/cm) 172 (neglecting thoria clusters)No. of (ThO2 + PuO2) fuel clusters in the core 360No. of natural ThO2 clusters in the core 120 (fixed) + 25 (moveable) = 145No. of fresh fuel assemblies per batch 120 (seeded) + 120 ThO2 (no seed)Active core height (mm) 3600Average fuel temperature 600 �CAverage coolant temp. (boiling H2O–70 kg/cm2) 286 �CCoolant inlet subcooling (kcal/kg) 7–20D2O—moderator temperature �C 80Radial D2O reflector thickness (mm) 600–700Axial D2O reflector thickness (mm) 600Calandria tank size (m) �8.4 dia · 4.8 heightControl system fast shutdown/233Pa to233U after S/D shutdown hold

Injection of liquid poison in dry tubes. Stainlesssteel rods in inter-channel space Moderator dumpduring refueling outages

Xenon over-ride Movable ThO2 clusters (25 nos.)

Table 2Description of fuel clusters (seeded + unseeded)

Parameter Fuel type

Seeded fuel cluster (Pu reprocessed from power reactors) Unseeded ThO2 cluster

Inner Middle Outer Single ring

Fuel clad ID/OD (mm) 10/11.4 9/10.4 12.6/14 12.6/14Pitch circle dia (mm) 104 130 158 158No. of fuel rods 24 30 30 30Seed content (wt. %) 20% PuO2 in ThO2 14% PuO2 in ThO2 One cycle irradiated ThO2 Fresh unseeded ThO2

rodsFissile fraction in seed 0.745 >0.94 –Composition of seed (239Pu: 240Pu:241Pu:242Pu);(69.3:24.1:5.2:1.4) 0.5 to 0.7% in situ bred 233U –ID/OD of central

BeO block (mm)10/90 (inclusive of Zr-liner) 10/137

Pressure tube Zr–Nb(2.5%) ID/OD (mm)

176/187 176/187

Air gap ID/OD (mm) 187/204 187/204Calandria tube

Zr-2 ID/OD (mm)204/207 204/207

Hexagonal assemblypitch (mm)

300 300


This would provide a radial reflector of 600–700 mm thickness and axial reflector of 600 mm thickness at bot-tom and top for a core height of 3600 mm.

Table 2 gives the description of the two types of ATBR fuel clusters. The seed fuel cluster consists of tworings of 24 and 30 seed fuel rods at pitch circle diameter (PCD) of 104 and 130 mm respectively. The thoriarods are at PCD of 158 mm. The pure thoria cluster contains only one ring of seedless thoria rods at PCD of158 mm. There is a BeO block inside the above two types of clusters with OD of 90 mm and 137 mm respec-tively. This BeO serves to eliminate 37 inner fuel rods. It was done to decrease the local peaking within the seedfuel cluster. The seed fuel considers reactor grade Pu with PuO2 content of 20 and 14 w/o in the inner andmiddle rings. The Zr–Nb (1%) clad is 0.7 mm thick for all the rods.

2. Calculation model and results

The lattice calculations of ATBR fuel clusters are done by the CLUB module [9] of the PHANTOM codesystem [10]. The recent calculations have been performed by the IAEAGX library in 172 energy groups and inWIMS/D format. This nuclear data library was generated as part of IAEA CRP on WIMS library updateproject [7,8]. CLUB code solves the transport problem by collision probability method. First flight collisionprobability method is used within each macro ring region and three term expansion is used for angular cur-rents at macro interfaces. The outer hexagonal boundary is circularized.

The fertile thoria clusters are simulated first. Since it does not contain any external seed, the thoria clusterburnup calculations are done at three typical one group flux levels of (0.5, 1.5 and 2.5) · 1014 n/cm2/s main-tained constant and up to 1500 days of irradiation. The seed fuel clusters are then simulated with the thoriarods isotopic densities picked up at fluence levels of 300, 500, 700 and 900 days at the above three flux levels.These are the typical expected cycle length duration. Figs. 5a and 5b show the variation of k1 with burnup forthe thoria cluster and the seed fuel cluster for typical flux value of 1.5 · 1014 n/cm2/s. It is seen that k1 of thethoria cluster rises sharply up to 800 days and then stays nearly flat. For the seeded fuel cluster the initial k1 is�1.2 and is higher with higher starting fluence levels in thoria rods. The difference however narrows down withburnup and it is noted that k1 remains above unity even up to an assembly average burnup of 50 GWD/T. Atthis burnup the unseeded thoria rods also achieve these same burnup like seed fuel rods. For the purpose ofcomparison the k1 plot is given for a cluster with average Pu content of 8.55% in all the rods. It is seen that thek1 variation of the seed fuel with thoria rods of in situ fissile content is less by about 100 mk. For 235U or 233U

0 400 800 1200 16000.0

0.2

0.4

0.6

0.8

1.0

Kin

f

ThO2 Cluster - Flux = 0.5X1.0E14 n/cm2 /sec



Burnup (Days)

Fig. 5a. k1 of thoria cluster after irradiation at different flux levels.

0 10 20 30 40 500.9

1.0

1.1

1.2

1.3

Burnup (GWD/T)

Kin

f

Average Pu Content 8.55% in all

20%/ 14%/ in situ (300X1.5X1.0E14 n/cm2/sec)



Fig. 5b. k1 of seed fuel cluster with different initial fluence levels for thoria rods.

0 10000 20000 30000 40000 50000

Burnup (MWD/T)

0.6

0.8

1.0

1.2

1.4

Rel

ativ

e P

ow

er

Inner Ring - 20% Pu

Middle Ring - 14% Pu

Outer Ring - in situ U-233

Fig. 6. Relative power in each fuel ring for seed fuel cluster.


seed the k1 variation will be much more. The relative power share in the three rings is plotted in Fig. 6. It isseen that the thoria ring power is 0.7 initially and crosses unity at 50 GWD/T.

0 10000 20000 30000 40000 50000

0

20000

40000

60000

Fu

el

Rin

g B

urn

up

(M

WD

/T) Inner Ring - 20% Pu

Initial Fluence in Outer ThO2 Rods - 1.5X1.0E14 n/cm2/sec X 700 days

Middle Ring - 14% Pu

OuterRing - in situ U-233

Assembly Burnup (MWD/T)

Fig. 7. Burnup accumulation in the three rings of seed fuel cluster.


The local peaking is initially 1.32 and occurs at middle ring. It monotonically decreases and falls belowunity at 37 GWD/T. Thinnest fuel rods (9 mm diameter) are used in this ring. The burnup accumulation inthe three fuel rings for starting fluence level of (1.5 · 1.0E14 n/cm2/s · 700 days) in thoria rods is shown inFig. 7. It is important to note that the cumulative burnup in unseeded thoria rods is 54 GWD/T when theassembly burnup is 50 GWD/T. It is also interesting to note that the inner 20% Pu seeded rods just reach51 GWD/T at this burnup while the middle ring with 14% Pu reach 55 GWD/T. The thoria rods started ata burnup of 10 GWD/T due to prior irradiation up to the above fluence level. Thus at the time of dischargeboth the seeded and unseeded fuel rods are seen to reach the same high discharge burnup of �50 GWD/T,averaged over fuel height.

The core calculations are done by the code TRISUL developed for the purpose [11]. TRISUL is a hexag-onal 3D finite difference diffusion code with one mesh per assembly radially. The lattice parameter interpola-tion for obtaining the two group cross sections of seed and thoria fuel clusters as a function of burnup, void(steam fraction), flux history and fluence thereof, and the xenon and Doppler perturbations applied and thethermal hydraulic model used, are explained in detail in Refs. [11,12]. Power-void iterations are done to obtainsteam formation profile consistent with the power profile.

Fig. 8a gives the burnup distribution at BOC and at EOC of optimized three batch loading in equilibriumcore. 1/6th core with rotational symmetry is simulated. The average flux seen by the thoria clusters in the pre-vious cycle is also a key parameter in deciding the cycle length of a reload pattern. We choose the 120 thoria

BOCEOC

FTYPE

1676128835

2

016846

1

03463

4

05604

3656251528

3

017110

1

2828239798

3

019484

1

07502

4

2881043863

3

04477

4

01065

4

3782453289

3

1963437769

2

020242

1

019652

1

017362

1

1661228627

2

018833

1

07746

4

1978937870

2

08099

4

020102

1

06300

5

1669328667

2

01055

4

1954936962

2

3771453131

3

019527

1

3713852452

3

3740952956

3

020034

1

017255

1

3662851507

3

06903

4

1938136909

2

07518

5

3736952593

3

07981

4

019437

1

04348

4

05464

1785234177

2

1901935615

2

1912836260

2

1936336855

2

019121

1

1991237754

2

019902

1

3165546208

3

2836039652

3

3602650053

3

06161

5

1898535554

2

06882

4

3761553042

3

07646

4

1989737733

2

07323

4

016875

1

03304

4

017374

1

3560849657

3

1889135040

2

3711151913

3

1962437008

2

018954

1

3782653200

3

019379

1

3613151057

3

016664

1

2754738721

3

05289

5

3423048095

3

06117

4

1678232969

2

06655

4

018976

1

07836

4

3375149854

2

07523

4

1905736452

2

06433

5

1677231512

2

1648827680

2

Fig. 8a. ATBR 1/6th equilibrium core 120 FAs/Batch BOC/EOC burnup distribution MWD/T.

FLUXHISTORY

1302

1320 720 176

1237 1302 1320

1154 1240 1302 830 297

828 1052 1218 1241 1296 1320

720 1317 830 1296 1238 1076 1317 297

1137 830 1076 1156 1076 1237 1295

1218 1238 1156 1291 1137 1302 1240 828 174

1291 1238 1218 1154 830 720 1137 1296 1317

1240 1154 1240 1237 1052 1320 828 1241 1317 705

1291 1241 1241 1154 1076 828 720 1156 1238 1373 1373

1052 1291 1156 1295 1218 1052 1373 1295 1295 1237 1137 1296 1373

Fig. 8b. ATBR 1/6th equilibrium core 120 FAs/Batch—one group absolute flux history values in units of 1011 n/cm2/s (core averagevalue = 1.118 · 1014 n/cm2/s).


clusters out of the total 145 clusters. These thoria clusters are chosen to be the ones with the highest flux his-tory levels. Fig. 8b shows the absolute one group flux history levels seen by the thoria clusters. The core aver-age flux history is 1.118 · 1014 n/cm2/s. For interpolation of cross sections three dimensional distribution offlux history and fluence thereof is considered. Since the cycle length achievable was seen to be 720 days,the seed fuel clusters consider the thoria rod fluence to be the product of respective flux history level and720 days. The longer cycle length and higher fluence also tacitly increases the fissile inventory in equilibriumcore through the in situ production of 233U in the same core in previous cycles. Fresh seed fuel clusters areavoided at core periphery locations. This is to enable a low leakage loading pattern. Fig. 9 shows a schematicdiagram of the three batch fueling scheme. The three batch fueling scheme is normally more difficult to

Fig. 9. ATBR equilibrium core 120 FAs/Batch-720 EFPD. Schematic diagram of the three batch fuels and thoria clusters.

0 120 240 360 480 600 720Core Burnup in FPD

1.01

1.02

1.03

K-e

ffec

tive

Fig. 10a. ATBR equilibrium core 120 FAs/Batch-720 EFPD keff vs. cycle burnup in FPD.

0 120 240 360 480 600 720Core Burn up in FPD

1.0

1.4

1.8

Pea

kin

g F

acto

r

3D Peak

Radial Assembly Peak Factor

Axial Peak Factor

Fig. 10b. ATBR equilibrium core 120 FAs/Batch-720 EFPD keff vs. cycle burnup in FPD.


optimize in comparison to the five batch fueling scheme shown in Fig. 4. The variation of keff, and peakingfactors as a function of core burnup are shown in Figs. 10a and 10b. The keff varies from 1.0212 at BOCto 1.0265 at 390 FPD and 1.0187 at EOC. The variation of ±4 mk over a period of two years would requireminimal control management and can be partly met by coolant inlet enthalpy variations or small movement ofmoveable thoria clusters. Fig. 11a gives the radial power distribution at BOC, 360 FPD (MOC) and at 720FPD (EOC). Fig. 11b gives the plots of axial power profile at these burnup values. It is seen that 3D powerpeaking monotonically decreases from 1.72 to 1.26. The radial and axial peaking factors also decrease mono-

BOCMOCEOC

799665634

1091937890

1110179

01538

939842800

1083959915

709646638

120111051043

2248355

948846804

1143228

03166

927882842

11171030965

126811451070

125411051035

1096975931

735675668

110110801048

2255373

11041029966

2269379

127811331059

1206304

732673667

03165

1027995956

910881847

117011151061

919873832

954883836

127211291057

1088969928

869848829

1225342

10441000953

2247364

911869830

2265376

123710931028

1139223

01537

951927912

973944919

1013976938

1040998953

113310941052

10831016960

124011271060

912818783

693633628

810795797

1197315

971943918

1224342

910882849

2252370

10921013955

2242349

1065945906

1105172

1009983988

812796798

940918904

863844825

1025993955

111310861051

920877839

119210991040

935840799

1077927881

739615588

0162276

803783795

1195316

939918911

1215335

111310841052

2257381

957920882

2248360

1085982925

2211307

978814766

761612575

Fig. 11a. ATBR 1/6th equilibrium core 120 FAs/Batch—radial power distribution at BOC/MOC/EOC.

00 6600 112200 118800 224400 330000 33660000..4400

00..6600

00..8800

11..0000

11..2200

11..4400

Axial Distance from Core Bottom (cm)

Rela

tive A

xial P

ow

er

BOC (0 FPD)

MOC (360 FPD)

EOC (720 FPD)

Fig. 11b. ATBR-600 equilibrium core with PuO2 seed in ThO2 axial power distribution at BOC/MOC/EOC.

Fig. 12a. ATBR-600 equilibrium core with PuO2 seed in ThO2—epithermal flux distribution at BOC-0 FPD.

Fig. 12b. ATBR-600 equilibrium core with PuO2 seed in ThO2—thermal flux distribution at BOC-0 FPD.


tonically. 3D flux plots at core mid plane and at BOC and at EOC are given in Figs. 12a, 12b and 13a, 13b forepithermal and thermal energy groups. It is seen that the epithermal flux shows depression in thoria clusters,which appear as mounds in thermal flux plots. Thus excess neutrons produced by the seed fuel rods slow downand create the thermal flux traps for effective breeding in the thoria cluster locations.

Table 3 gives the seed input and output contents for the equilibrium core. It is seen that the only 40% of thePu feed is consumed (0.88 T out of 2.2 T). Since thoria rods achieve about 50 GWD/T burnup, nearly 640 kgof the 12.78 T of thorium loading (at the rate of 5 kg/T) in the unseeded thoria rods is used for energy

Fig. 13a. ATBR-600 equilibrium core with PuO2 seed in ThO2—epithermal flux distribution at EOC-720 FPD.

Fig. 13b. ATBR-600 equilibrium core with PuO2 seed in ThO2—thermal flux distribution at EOC-720 FPD.


production in its four fuel cycle operation. About 502 kg of uranium is available from the discharged thoriumwith 87% fissile. This shows that 886 kg of reactor grade Pu is utilized for converting 1142 kg of thorium touranium of which 640 kg is used to produce 50% share of power in the same reactor. Thus the ATBR can beused for Pu incineration and it produces the intrinsically proliferation resistant 233U for sustenance of futurereactor programme. Since 50% of the core loading is always unseeded thorium, the load on reprocessing andrefabrication of back end fuel will be proportionately less.

The void perturbation was evaluated by decreasing the flow and decreasing the inlet subcooling at thefull reactor power. The reactivity change when the core average void changes from 0.246 to 0.6 is�0.32 mk/% void at the beginning of cycle and �0.08 at the end of cycle of equilibrium core. These resultsare comparable to the earlier results obtained with 69 group WIMS library versions available at that time[1].

Table 3Summary of the equilibrium core studies Pu seeded core; 3-batch fueling; 120 assemblies per batch; cycle length—720 EFPD

Parameter Results with IAEAGX-172 grouplibrary

Average discharge burnup (MWD/T) 49,050Pu seed input in fresh batch of 120 assemblies (kg) 2212.1Fissile Pu content (kg) 1647.4Fissile fraction (%) 74.5Seed output in discharged batch (kg) 1834.5Output to input ratio 0.829

Break-up of seed output

Weight of U-total in thorium (kg) 502.3Weight of 233U in thorium (kg) 426.3Weight of 233U + 235U in thorium (kg) 436.9Fissile fraction (%) 87.0Weight of total Pu in thorium (kg) 1332.2Weight of fissile Pu (kg) 731.4Fissile content in output (kg) 1168.3Fissile fraction (%) 63.7

Core reactivity at full power

Minimum reactivity (mk) 18.4Maximum reactivity mk) 25.8Reactivity spread (mk) 7.4 or ±4

Void reactivity coefficient (oq/ov) in (mk/%void) �0.32 (BOC)(core average void changed from 24.6% to 60%) �0.08 (EOC)


3. Conclusions

The ATBR core design with Pu seed in thorium is found to exhibit excellent core characteristics like twoyear cycle length with practically no control maneuvers. This is achieved by using the thin seed fuel rods withhigh Pu content to create large volumes of neutron flux trap zones where the fertile breeding in thoria rods cantake place at a rate matched with fissile depletion rate. The fissile seed remains conserved for a long durationof six years (three fuel cycles) due to its location in the fuel cluster. The seed and fertile fuel rods achieve thesame high discharge burnup of �50 GWD/T. The reactivity coefficients are found to be small. The void coef-ficient in the operating range is negative. This ATBR core with Pu seed has an overall safe and economic char-acteristic ideally suited for large scale induction of thorium. Longer cycle length is also desirable in closed fuelcycles to maximize the time gaps needed for reprocessing and refabrication. ATBR with Pu feed can beregarded as Pu incinerator and it produces the intrinsically proliferation resistant 233U for sustenance of futurereactor programme. Since 50% of the core loading is always unseeded thorium, the load on reprocessing andrefabrication of back end fuel will be proportionately less. It can also be designed for burning minor actinidessince it is designed to accommodate large fertile breeding zones.

The delayed neutron fraction ‘b’ is 2.5–3.2 mk in the Pu seeded core and prompt neutron mean life time is�10–15 ls. The kinetics behaviour of ATBR core would be similar to fast reactors using Pu feed. Howeversince all reactivity coefficients are small in magnitude and minimum control maneuvers will be needed it isexpected that the operation of ATBR can be smooth and stable. The initial core and approach to equilibriumcore loading, control design and the stability characteristics are being studied.

In view of the possibility of long fuel cycle length with minimal control maneuvers, small value of reactivitycoefficient and high seed output to input ratio, the ATBR concept with Pu seed in thorium can be regarded asintrinsically safe and economic and holds promise for large scale use of thorium in the coming years.

References

[1] Jagannathan V, Ganesan S, Karthikeyan R. Sensitivity studies for a thorium breeder reactor design with the nuclear data libraries ofWIMS library update project. In: Proceedings of the international conference on emerging nuclear energy systems ICENES-2000,September 25–28, 2000, Petten, The Netherlands.

[2] Jagannathan V, Jain RP, Krishnani PD, Ushadevi G, Karthikeyan R. An optimal U–Th mix reactor to maximize fission nuclearpower. In: Proceedings of the international conference on the physics of nuclear science and technology, October 5–8, 1998, TheRadisson Hotel, Islandia, Long Island, New York, p. 323.

[3] Jagannathan V, Lawande SV. A thorium breeder concept for optimal energy extraction from uranium and thorium. In: Proceedingsof the IAEA TCM on fuel cycle options in LWRs & HWRs, held at Victoria, Canada, April 28–May 1, 1998, IAEA TECDOC-1122,p. 231.

[4] Jagannathan V. ATBR—a thorium breeder reactor concept for early induction of thorium with no feed enrichment, published as afeature article in BARC Newsletter, No. 187, 1999.

[5] Jagannathan V, Pal Usha, Karthikeyan R, Ganesan S, Jain RP, Kamat SU. ATBR—a thorium breeder reactor concept for an earlyinduction of thorium in an enriched uranium reactor. J Nucl Technol 2001;133(January):1–32.

[6] Jagannathan V, Usha Pal, Karthikeyan R. A search into the optimal U–Pu–Th fuel cycle options for the next millennium. In:Proceedings of PHYSOR’2000 international conference—ANS international topical meeting on advances in reactor physics andmathematics and computation into the next millennium, Session XV.C on advanced reactor concepts and design, May 7–11, 2000,Pittsburgh, Pennsylvania, USA.

[7] Leszczynski F, Aldama DL, Trkov A, editors. WIMS-D library update-final report of a co-ordinated research project, IAEA-DOC-DRAFT, October 2001.

[8] IAEA TECDOC draft report available at the WLUP website: http://www-nds.iaea.org/wimsd/Download/TECDOCDRAFT.PDF,May 2002.

[9] Krishnani PD. Ann Nucl Energy 1982;9:255.[10] Jagannathan V, Jain RP, Kumar V, Gupta HC, Krishnani PD. A diffusion iterative model for simulation of reactivity devices in

pressurized heavy water reactors. Nucl Sci Engg 1990;104:222–38.[11] Jagannathan V. TRISUL—a code for thorium reactor investigations with segregated uranium loading. Report B.A.R.C./E/028/2000.[12] Jagannathan V, Ganesan S, Pal U, Karthikeyan R, Jain RP. Computation tools for search of optimal thorium fuel cycle and their fuel

management strategies, Paper Log086 of Proceedings of M&C 2001, ANS International Meeting on Mathematical Methods forNuclear Applications, September 2001, Salt Lake City, UT, USA.

http://www-nds.iaea.org/wimsd/Download/TECDOCDRAFT.PDF

towards an intrinsically safe and economic thorium breeder reactor

Documents