NUCLEAR SCIENCE AND ENGINEERING: 152, 204–218 ~2006!

A Physics Study of a 600-MW(thermal) Gas-Cooled Fast Reactor

Hangbok Choi,*† Gérald Rimpault, and Jean C. Bosq

Commissariat à l’Energie Atomique Cadarache Center
13108 Saint-Paul-lez-Durance Cedex, France

Received March 19, 2004

Accepted May 3, 2005

Abstract – A neutronic feasibility study was performed for a 600-MW(thermal) gas-cooled fast reactor
fuel cycle through recycling simulations. Sensitivity calculations were also performed for various physics
design parameters such as the plutonium volume fraction of the fuel, fuel burnup, core material volume
fraction, and the power density. The results showed that the initial breeding gain of ⫺0.04755 is sufficient
to sustain the recycling of the actinides with a reasonable amount of natural uranium and plutonium feed
material. The comparative calculation on the core power density has shown that it is feasible to reduce
the amount of minor actinides and spent fuel in the high power density core (98.4 MW/m 3 ) compared to the
reference core (58.2 MW/m 3 ). It was also found that the fuel cycle cost is saved by 0.4 mills/kW{h for
the high power density core compared to the reference core.

I. INTRODUCTION 4. resource saving by optimizing the utilization of

the available resources of fissile and fertile
Considerable interest has been shown recently in the material
gas-cooled fast reactor ~GFR! core design, which is due 5. enhanced resistance to proliferation risks
to the improved capability of the GFR for plutonium
and minor actinide management as a result of the hard 6. potential for applications other than electricity.
neutron spectrum as well as the small voiding effect. Since the 1960s the GFR has been studied in both
Currently, the advanced designs of the high-temperature the United States 3,4 and Europe.5 As the sodium coolant
gas-cooled fast reactors are being studied as candidates has been shown to have several operational difficulties
for the next-generation reactor concept to be introduced in a fast reactor such as the need for an intermediate
to the reactor park by 2040 ~Ref. 1!. The design concept coolant loop, its chemical reactivity with air and water,
of this reactor, including the fuel and fuel cycle, copes the lack of optical transmission, and the need to manage
with the fuel cycle strategies set by the technology road- and ultimately dispose of the sodium, the GFR has re-
map of the Generation IV nuclear energy systems as ceived renewed interest as a feasible option that can be
follows 2 : used as the next-generation reactor. In principle, it is
1. economic competitiveness against other electric- possible in the GFR to fulfill the strategic requirements
ity generation means, with a special emphasis on mentioned previously. For example, the high-temperature
low capital cost gas coolant system replaces the steam cycle with a closed-
loop gas turbine ~Brayton! cycle. With the gas turbine, a
2. enhanced safety through an increased resistance net plant efficiency of ;50% can be realized, which will
to core damage in case of a severe accident surely improve the economics.6 The gas coolant also en-
3. cleanliness by minimizing the production of long- hances a low void reactivity, which allows the relaxation
lived radioactive waste of the safety-related constraints in a core heavily loaded
with higher actinides.
*Permanent address: Korea Atomic Energy Research In- Recently, Hejzlar et al. studied a 600-MW~thermal!
stitute, P.O. Box 105, Yusong, Taejon, 305-600 Korea gas-cooled fast reactor based on the modular helium
†E-mail: choih@kaeri.re.kr reactor design, investigating the fuel and matrix type to

204
 A PHYSICS STUDY OF A 600-MW~thermal! GAS-COOLED FAST REACTOR 205

achieve a long-life core with a low power density as well

as the feasibility of decay heat removal and fuel cycle
economics.7 A similar study was performed by Garnier
et al., which recommended a specific core power in the
range of 50 to 100 MW0m 3 and defined other major sys-
tem parameters.8 The fuel concept and fuel cycle model
were also proposed with the results of the preliminary
safety assessment on the decay heat removal.
Based on Garnier et al.’s 600-MW~thermal! GFR con-
cept, a neutronic feasibility study is performed in this
study for a wide range of physics design parameters in-
cluding the high power density. The neutronic analysis
compares major physics performance parameters of the
core for different design variables. Emphasis is given to
the breeding gain ~BG! that determines the recycling
capability of the proposed core as well as other physics
performance parameters such as the reactivity swing,
void reactivity, Doppler reactivity, etc. Then the re-
cycling calculation is performed to assess the feasibility Fig. 1. Schematic view of a fuel block.
of the fissile self-sufficient fuel cycle. Finally, the fuel
cycle cost of the GFR is estimated for comparison be-
tween the reference and high power density cores even
though there are large uncertainties in the unit costs of
average volume ratio of the fuel and matrix was deter-
the fuel cycle components such as the fabrication of the
mined to be 50050. For the depletion of the fuel, the
particle fuel and reprocessing of the spent fuel.
maximum fuel burnup is limited to 5 wt% of the initial
heavy metal. The average fuel temperature was set at
12278C, which is much less than the limiting tempera-
II. REFERENCE CORE MODEL ture ~14958C! of the high-temperature gas-cooled
reactor.11
The thermal power of the reference core is 600 MW. The hexagonal fuel block shown in Fig. 1 is the base
The core volume is 10.3 m 3, which results in a power unit that forms the reactor core. The fuel block is made
density of 58.2 MW0m 3. The core contains 28 tons of of SiC, which is consistent with the matrix material used
heavy metal ~HM!, which gives a specific power of 21.4 for the fuel particle. For the cooling of the fuel block,
kW0kg HM. The fuel used for the initial core loading is there are 61 vertical holes ~flow tube! that are arranged
reprocessed plutonium mixed with natural uranium. The in a hexagonal form with a pitch distance of 2.18 cm.
plutonium is from the spent fuels of current French pres- The flat-to-flat diameter of the fuel block is 21.5 cm.
surized water reactors, which are cooled down until the When these blocks are deployed in the core, they are
year 2016. In the initial core, the volume fraction of the arranged with a lattice pitch of 22.2 cm to allow the
plutonium fuel in the plutonium-uranium mixture is 17%. coolant to flow between the fuel blocks. For the fuel
The chemical form of the fuel will be uranium-plutonium assembly lattice under the operating conditions, the vol-
carbide to create more fissile breeding compared to the ume ratio of the fuel, matrix, structure, and coolant is
oxide fuel.9 The reference core uses helium for the reac- 2502509041, respectively.
tor coolant, which allows the reactor to be operated at a
much higher temperature compared with a water-cooled
system, leading to improvements in the electricity gen- II.B. Control0Shutoff Assembly
erating efficiency.
The control and shutoff rod designs have not yet
II.A. Fuel Block been determined. Therefore, it was assumed that one of the
existing control rod designs can be used for the reference
In the reference core, a dispersed fuel is employed.10 core. For example, the control rods can be arranged in a
The dispersed fuel is a CERCER type with very small hexagonal array 12 as shown in Fig. 2. The reference ma-
spherical uranium-plutonium carbide fuel particles terial used for the absorber is enriched boron carbide
;1 mm in diameter embedded in thin carbon ~pyrocar- ~B4C!, which is enriched by 90% of 10 B. Each absorber
bon! layers that can accommodate the fission products, rod is covered by SiC cladding. The hexagonal cluster is
surrounded by a silicon carbide ~SiC! layer of a hexag- embedded in a hexagonal wrapper made of SiC. The
onal shape being an extremely important diffusion bar- volume ratio of the absorber, structure, and coolant is
rier. In the neutronic modeling of the fuel particles, the 25028047. The bottom of the absorber is connected to

NUCLEAR SCIENCE AND ENGINEERING VOL. 152 FEB. 2006

206 CHOI, RIMPAULT, and BOSQ

Fig. 2. A conceptual control assembly model.

the follower, which is made of the structural material. In

the follower region, the volume ratio of the structure and
coolant is 28072.

II.C. Reactor Core

The core is divided into two radial regions with dif-
ferent plutonium charge fractions. The number of assem-
blies is 64 and 78 for the inner and outer regions,
respectively. Six positions are reserved for control as-
semblies, while three positions are reserved for shutoff Fig. 3. Layout of the reference GFR core.
assemblies. The equivalent radii of the inner and outer
core are 95.4 and 143.2 cm, respectively. The radius of
the whole system including the reflector is 212.1 cm. At
the upper and lower ends of the active core, axial reflec- The ERANOS code has been benchmarked to verify
tors are positioned, which are 100 cm thick. The system its adequacy as a design tool of the GFR. For example,
pressure is 70 bar under normal operating conditions. the computational benchmark calculation was performed
The layout of the reference core is shown in Fig. 3. The against the Japan Nuclear Cycle Development Institute
design parameters of the core and fuel are summarized ~JNC! neutronics tool that adopted the Monte Carlo
in Table I. method for the heterogeneity treatment of the fuel sub-
assembly. The result showed that the criticality is esti-
mated within 0.5%Dk, which is considered to be
III. REFERENCE CORE CALCULATION satisfactory at the conceptual design state.14 The neutron
streaming algorithm was also validated by the analytic
The neutronic properties of the reference core were and Monte Carlo methods especially for the core depres-
calculated by the ERANOS code 13 system, which has surizing effect.15 For the fast reactor loaded with higher
various computing modules for reactor physics and fuel actinides, it is important to consider the nuclear data
cycle analyses. In the ERANOS code, the lattice param- uncertainty in the neutronic design. Currently, experi-
eters are generated by the ECCO module, using the JEF2.2 mental and numerical studies on the nuclear data uncer-
nuclear data library. To save the computing time and tainty are conducted for the development of the GFR.16,17
working space, the diffusion option was used for most of
the parametric calculations with a 33-group working li- III.A. Reference Core Characteristics
brary that was generated from the 1968-group master
library. Regarding the geometry modeling, the heteroge- The neutronic calculations provide core perfor-
neity of the fuel assembly is well described by the ECCO mance parameters such as the BG and reactivity con-
calculation. For the geometry model of the core, a cylin- stants. The BG is defined as the difference of the weighted
drical ~RZ! model is used along with the diffusion option. isotopic concentration at the beginning of life ~BOL!

NUCLEAR SCIENCE AND ENGINEERING VOL. 152 FEB. 2006

 A PHYSICS STUDY OF A 600-MW~thermal! GAS-COOLED FAST REACTOR 207

TABLE I
Design Parameters of the Reference GFR Core

System parameters
Reactor thermal power 600 MW
Fuel temperature 12278C
Structure temperature 6658C
Coolant temperature 6658C
Absorber temperature 6278C
Coolant inlet temperature 3958C
Core pressure 70 bar
Dispersed fuel
Fuel compound ~U, Pu!C
Fuel particle diameter 1 mm
Inert matrix SiC
Fuel0matrix volume ratio 50050 Fig. 4. Reactivity change of the reference core.

Fuel assembly
Flat-to-flat diameter 21.5 cm
Block thickness 0.3 cm 6658C. The voiding condition is simulated using a cool-
Number of coolant holes 61 ant density of 5.010 ⫻ 10⫺5 g0cm 3 at 1 bar and 6658C.
Coolant hole diameter 1.45 cm For the cold condition of the Doppler reactivity calcula-
Coolant hole thickness 0.04 cm
Coolant hole pitch ~triangular! 2.18 cm
tion, it is assumed that the fuel temperature is reduced to
Active fuel height 34 cm 1808C.
To perform the core calculation, the axial position of
Core design the control0shutoff assemblies should be specified, and
Number of fuel assemblies 142 therefore a sensitivity calculation was performed by
Number of control assemblies 6
changing the parking position of the control rods. As
Number of shutoff assemblies 3
Number of radial reflector blocks 180 the bottom of the control rod moves from the core top
Active core height 170 cm ~parking position 0 cm! to the upper boundary of the
Lattice pitch 22.2 cm reflector ~parking position 100 cm!, the core reactivity
Interassembly gap 0.7 cm increases by 736 ⫻ 10⫺5, which is 15% of the maximum
core reactivity achievable. To keep the reactivity loss as
Control0shutoff assembly
Absorber B4C ~90% enriched small as possible ~e.g., ,1%!, the axial position of the
10
B! control rod was determined to be 80 cm above the active
Structure SiC core, as was the parking position of the shutoff rod. It
Number of absorbers 37 should be noted, however, that the parking position needs
Follower SiC to be adjusted in the future when considering the me-
Reflector
chanical property of the control rod driving system such
Structure compound Zr3Si2 as the delay time.
Number of layers ~radial! 3 to 4
Axial reflector thickness 100 cm III.A.1. Fuel Management
Structure0coolant volume ratio 80020
For the initial core loaded with uranium-plutonium
~radial!
Structure0coolant volume ratio 60040 fuel, the depletion calculation was performed under a
~axial! three-batch mode. The criticality was evaluated at the
end of the second cycle ~burnup step! and by constrain-
ing the discharge burnup to 5 wt% as shown in Fig. 4.
The cycle length is 782 full power days ~FPDs! and the
residence time ~lifetime! of the fuel will be 2346 FPDs
and the end of life ~EOL! divided by the weighted iso-
~6.4 yr!. The plutonium volume fractions are 14.9 and
topic concentration at the BOL. The weighting factor
18.6% for the inner and outer cores, respectively. The
is estimated as an equivalent 239 Pu concentration. In
reactivity swing of one cycle is estimated to be 1516 ⫻
the ERANOS code, the weights of 238 U and 239 Pu are 0
10⫺5 based on the linear reactivity model.18
and 1, respectively. Other key performance parameters
such as the reactivity swing, void reactivity, and the Dopp- III.A.2. Neutronic Parameters
ler reactivity also have the conventional definitions. In
this study, 4 He is used as the coolant material in which The neutronic parameters and peak power of the core
the nominal density is 3.562 ⫻ 10⫺3 g0cm 3 at 70 bar and are summarized in Table II, in which the EOL indicates

NUCLEAR SCIENCE AND ENGINEERING VOL. 152 FEB. 2006

208 CHOI, RIMPAULT, and BOSQ

TABLE II code. The results showed that the void reactivity is dom-
Neutronic Characteristics of the Reference GFR Core inated by the elastic removal reaction in the core region.
In case of the coolant voiding, the number of scattering
Beginning- reactions ~dominant at ;1 MeV! decreases in the cool-
of-Life End-of-Life ant region, which results in spectrum hardening and pos-
itive void reactivity. The Doppler reactivity is dominated
Reactivity swing per cycle ~pcm! 1516 by the neutron capture of the fertile isotope, which is
Void reactivity ~pcm! 168.8 205.0 dominant at ;1 keV. Though there is a positive contri-
Doppler reactivity ~pcm! ⫺2260.4 ⫺1836.7 bution to the Doppler reactivity by the resonance broad-
Delayed neutron fraction ~pcm! 380.7 355.1 ening of the fission cross sections, it contributes to the
Neutron generation time ~sec! 9.04 ⫻ 10⫺7 8.43 ⫻ 10⫺7 total reactivity only by 13%. The isotope-wise contribu-
Peak power density ~W0cm 3 !
tion to the delayed neutron fraction was also estimated
Inner core 78.0 86.7 by the perturbation calculation. The result showed that
Outer core 81.2 76.9 the 238 U contributes most ~44%! to the delayed neutron
Peak linear power ~W0cm! fraction.
Inner core 102.2 113.7
Outer core 106.5 100.8 III.A.3. Breeding Gain
Breeding gain of the fuel cycle
Inner core ⫺0.00032
The characteristics of the fuel cycle BG can be in-
Outer core ⫺0.04723 vestigated by decomposing the BG into the isotope-wise
Total ⫺0.04755 and reaction-wise terms. The EOL number density ~N3 !
can be written in terms of the BOL number density ~N0 !
and the exponential matrix as follows:
step
the end of the third cycle ~2346 FPDs!. The BG of the
whole core is slightly negative. It can be seen that the
N3 ⫽ (i exp~Midt !N0 ,
negative component of the BG comes from the outer
core, which has a higher fissile content compared to the where M is the transmutation matrix, which includes the
inner core. In the initial core, the plutonium volume frac- cross section, decay constant, and transmutation flux.
tion ratio of the inner and outer core is 0.80 ~or 0.81 in The i and dt are the burnup step number and step length,
terms of fissile content ratio!. As the fissile material in respectively. For the reference core, the positive term of
the outer core burns, however, the fissile content ratio the BG is mostly from the radioactive decay of 239 Np
increases to 0.89, and the peak power of the inner core ~0.52987! and the neutron capture of 240 Pu ~0.07819!,
exceeds that of the outer core at the EOL state. while the negative term is mostly due to the loss ~cap-
The variations of the reactivity terms are generally ture ⫹ fissions! of 239Pu ~⫺0.35764! and 241Pu ~⫺0.09094!
consistent with the isotopic content changes. For the re- as well as the neutron capture of 238 U ~⫺0.20076!. How-
activity swing, the core reactivity decreases because of ever, because the neutron capture of 238 U produces 239 Np
the fission products buildup and fissile material burning. that becomes 239 Pu, the neutron capture of 238 U can even-
The void reactivity increases by 21% at the EOL state tually be treated as a positive term ~0.32909!.
because of the neutron spectrum shift that enhances more The BG ~or ratio! is determined by the initial iso-
importance to the neutrons at ;1 MeV where the elastic topic number density and transmutation flux. Because
scattering is dominant. The Doppler reactivity increases the initial number density is determined by the pluto-
by 19% because of the 238 U loss. The kinetic parameters nium content, the effect of the initial number density can
such as the delayed neutron fraction ~ beff ! and prompt be estimated by the sensitivity of the breeding ratio to
neutron lifetime ~L! decrease by 7% because of the 235 U the plutonium volume fraction, which is ⫺27.7%. There-
burning and 238 U capture. Especially, the magnitude of fore, to have a high breeding core, the plutonium volume
L is smaller than that of the thermal reactors by a factor fraction should be reduced, which will be limited, how-
of 100. It is true that L of the GFR is similar to that of the ever, because of the criticality requirement of the core.
liquid metal fast breeder reactor loaded with mixed ox- As an alternative, the fuel volume fraction in the core
ide fuel. However, because the thermal conductivity of can be increased so that the plutonium fraction is re-
the gas coolant is very low ~e.g., k ⫽ 0.0188 W0m{K for duced without losing core reactivity to a certain extent.
He at 600 K! when compared to that of the liquid metal 19 For the effect of the transmutation flux on the breeding
~e.g., k ⫽ 11.2 W0m{K for sodium at 293 K!, it will be ratio, the sensitivity can be divided into two groups.
necessary to confirm the integrity of the GFR core in The high-energy neutrons above 67 keV have negative
case of severe accidents. effects on the breeding ratio, while the neutrons below
The void and Doppler reactivity of the BOL state were 67 keV have positive effects. However, the sensitivity
investigated by the perturbation option of the ERANOS itself is much smaller ~⫺0.56%! when compared to the

NUCLEAR SCIENCE AND ENGINEERING VOL. 152 FEB. 2006

 A PHYSICS STUDY OF A 600-MW~thermal! GAS-COOLED FAST REACTOR 209

is ,60 cm, the largest variation of these neutronic pa-

rameters is 7.4% compared to the reference condition
~control rods at ⫺80 cm!. Therefore, it is expected that
the effects of the control rod insertion on the neutronic
parameters are not significant for the reference core de-
sign for which the critical insertion depth is 10 cm.

III.C. Sensitivity to the Design Parameters

For the reference core discussed so far, the paramet-
ric calculations have been performed to determine if there
are any improvements in the neutronic performances.
For all the cases, the reactor power @600 MW~thermal!#,
core height ~170 cm!, and the fuel assembly size are the
Fig. 5. Insertion characteristics of the control rods. same as those of the reference core. These calculations
focus on the change of the BG, which is still negative for
the reference core. The variables used for the parametric
calculations and their variations are listed in Table III.
The reflector has gas coolant to remove the heat,
sensitivity to the isotopic number density because it is which is mostly generated by gamma radiation. The gas
mostly determined by the actinide decay not by the flux fraction of the reflector has only minor effects on the
level. Therefore, it is expected that the breeding ratio core performance parameters. If the gas fraction in-
will decrease if the power density increases even though creases, the BG slightly increases because of the induced
the magnitude of the change is small. fast neutron leakage from the core, which in turn slightly
reduces the reactivity swing.
III.B. Control Rod Effect The effects of the plutonium volume fraction, fuel
burnup, plutonium volume fraction ratio, and the core
The parking position of the control rods was deter- volume on the physics performance parameters of the
mined to be 80 cm above the active core. The insertion core are summarized in Table IV for the BOL state. The
characteristics of the control rods were estimated for the plutonium volume fraction has a direct effect on the BG.
standard absorber ~enriched boron carbide! and com- As the plutonium volume fraction is reduced, the breed-
pared to that of the natural boron carbide absorber, as ing capability of the core increases owing to the induced
shown in Fig. 5. It can be seen that the effect of the neutron capture by the fertile isotopes. However, there
strong absorber is pronounced as the control rods are will be a practical limit in reducing the plutonium vol-
deep in the core. For the reference core of which the ume fraction due to the core criticality requirements.
reactivity swing is 1516 ⫻ 10⫺5 ~under a three-batch For example, if the plutonium volume fraction changes
mode!, the control rods, on average, shall be inserted from 17 to 14%, the initial core reactivity reduces from
into the core to suppress one half of the reactivity swing. 3129 ⫻ 10⫺5 to ⫺6727 ⫻ 10⫺5.
The estimated insertion depth of the control rod is 10 cm. The effect of the fuel burnup was estimated for three
Even though the control rod is inserted, the shape of the burnup values such as 3, 4, and 5 wt%, which corre-
burnup reactivity does not change, and only the magni- spond to 1404, 1875, and 2346 FPDs in terms of the
tude is reduced by ;800 ⫻ 10⫺5 throughout the fuel cycle length ~three-batch mode!. If the discharge burnup
cycle. The flux level, in principle, does not change be- of the fuel decreases, the criticality is maintained with
cause of the control rod insertion because the reactor less amount of fissile material and the BG increases
power stays the same. However, there are local pertur- accordingly.
bations in the group fluxes. For example, the peak neu- The plutonium volume fraction ratio of the inner
tron fluxes at 0.1 MeV increase by 1.5 and 0.3% in the and outer core is adjusted to reduce the peak power of
inner and outer core regions, respectively. the core. The peak power density is the lowest when
The core performance parameter such as the BG is the ratio is 0.8 for the reference core. If the inner and
also affected by the control rod insertion. If the control outer cores have the same plutonium volume fraction
rod is inserted by 10 cm ~criticality condition!, the dif- ~ratio ⫽ 1!, the neutron leakage is reduced when com-
ference in the BG is 0.00009 when compared to the core pared to the reference core and the BG increases by 0.008.
with the control rods at their parking position. The change However, the peak power density increases by 25% when
of the BG is less than 0.0001 if the control rod insertion compared to the reference core. The void and Doppler
is less than 60 cm. The neutronic parameters such as the reactivity are not sensitive to the plutonium volume frac-
reactivity coefficient and delayed neutron fraction are tion ratio. However, the reactivity swing is the smallest
also affected by the control rod. When the insertion depth for the reference core condition.

NUCLEAR SCIENCE AND ENGINEERING VOL. 152 FEB. 2006

210 CHOI, RIMPAULT, and BOSQ

TABLE III
Variables Used for the Parametric Calculations

Range

Gas coolant volume fraction in radial reflector 0 to 50%

Plutonium volume fraction of the fuel 14 to 17%
Fuel burnup 3 to 5 wt%
Plutonium volume fraction ratio ~inner0outer core! 0.5 to 1.0
Material volume fraction of a fuel assembly
Fuel0matrix 40060 to 70030
Gas coolant 31 to 50%
Matrix 20 to 30%
Specific power 51.7 to 98.4 W0cm 3

TABLE IV
Effects of the Plutonium Volume Fraction, Fuel Burnup, Plutonium Volume Fraction Ratio,
and the Core Volume on the Core Neutronic Parameters

Peak Power
Breeding DrBU rVoid rDoppler Density
Gain ~pcm! ~pcm! ~pcm! ~W0cm 3 !

Plutonium volume fraction ~%!

14.0 0.02963 7.4 187.1 ⫺2987.8 80.9
15.0 ⫺0.00109 1787.2 180.2 ⫺2711.0 81.0
16.0 ⫺0.02673 3224.1 174.1 ⫺2468.0 81.1
17.0 a ⫺0.04755 4344.4 168.8 ⫺2260.4 81.2
Fuel burnup ~wt%!
3.0 ⫺0.02624 2423.6 171.3 ⫺2356.7 81.2
4.0 ⫺0.03633 3343.6 170.1 ⫺2310.4 81.2
5.0 a ⫺0.04755 4344.4 168.8 ⫺2260.4 81.2
Plutonium volume fraction ratio
0.6 ⫺0.06074 5283.3 155.9 ⫺2050.2 91.5
0.7 ⫺0.05351 4571.3 164.1 ⫺2184.4 86.5
0.8 a ⫺0.04755 4344.4 168.8 ⫺2260.4 81.2
0.9 ⫺0.04297 4505.0 169.9 ⫺2279.5 93.1
1.0 ⫺0.03953 4904.8 168.7 ⫺2259.8 108.1
Core volume ~m 3 !
6.1 ⫺0.05981 4811.9 155.4 ⫺2046.8 140.1
8.6 ⫺0.05097 4607.3 163.3 ⫺2207.0 96.9
10.3 a ⫺0.04755 4344.4 168.8 ⫺2260.4 81.2
11.6 ⫺0.04568 4173.7 173.0 ⫺2297.6 72.0
a
Reference core condition.

The effect of the core volume ~or power density! for SA-084, SA-118, and SA-160, respectively. In gen-
was examined for three different configurations as shown eral, if the core size is reduced, the neutron leakage is
in Fig. 6. The core configuration SA-084 has the small- pronounced, which reduces the void reactivity of the
est core volume and the highest power density, while the whole core. However, because the neutron economy is
SA-160 is the largest. In between those, the SA-118 con- getting worse, the core requires more fissile material ini-
figuration was assessed for comparison. In all the cases, tially and the reactivity swing increases. In the high power
the fuel discharge burnup was fixed at 5 wt%. The cor- density core, the BG is reduced by 0.012 when com-
responding cycle lengths are 1386, 1953, and 2640 FPDs pared to the reference core.

NUCLEAR SCIENCE AND ENGINEERING VOL. 152 FEB. 2006

 A PHYSICS STUDY OF A 600-MW~thermal! GAS-COOLED FAST REACTOR 211

Fig. 6. Core configurations of SA-084, SA-118, and SA-160.

For the fuel assembly, the sensitivity of the BG to fraction decreases, which requires higher plutonium con-
the material volume was estimated as shown in Fig. 7. In tent. For example, if the gas fraction increases from 31.5
the case of the fuel volume change, the matrix volume to 41.0%, the plutonium volume fraction increases by
changes accordingly, but the coolant volume remains the 1.5%, which is in fact a combined effect of both the
same. If the fuel volume increases, the amount of fissile coolant and fuel volume changes. For the effect of the
plutonium required for the criticality of the core is re- matrix volume, the coolant volume changes along with
duced, which increases the BG and decreases the reac- the matrix volume change but the fuel volume remains
tivity swing appreciably. If the fuel fraction increases the same. If the matrix volume fraction changes from
from 19.9 to 29.9%, the plutonium volume fraction de- 20.2 to 30.6% ~this corresponds to the coolant volume
creases by 5.6%. In the case of the coolant volume change, change from 45 to 35%!, the plutonium volume fraction
the fuel0matrix volume is adjusted, but the volume ratio increases by 0.9%, which is the smallest variation among
of the fuel to the matrix is fixed at the reference value of the three cases.
50050. Therefore, if the gas fraction increases, the fuel The sensitivity calculations have shown that the phys-
ics performance of the proposed GFR is well maintained
for the perturbation ranges considered in this study. The
BG of the reference core is negative but close to zero.
The BG can be increased by adjusting the plutonium
content; however, the reactivity swing is also very sen-
sitive to the plutonium content. In order to confirm the
potential of the fissile self-sufficient GFR fuel cycle, the
recycling simulation was performed for the reference
core.

IV. RECYCLING SIMULATION

In the GFR fuel cycle proposed in this study, all the

actinides are recycled homogeneously after being mixed
with feed material to make up for the mass loss due to
depletion. In this fuel cycle, because uranium isotopes
are continuously transmuted to higher actinides, the breed-
Fig. 7. Sensitivity of the breeding gain to the material ing capability of the core will increase as the fuel is
volume fraction of a fuel assembly. recycled.

NUCLEAR SCIENCE AND ENGINEERING VOL. 152 FEB. 2006

212 CHOI, RIMPAULT, and BOSQ

IV.A. Reference Core Simulation

The recycling simulation can be performed based on
the mass balance obtained from the ERANOS depletion
calculation. The main effort in the recycling simulation
is to determine the fuel composition for the subsequent
fuel cycle, which was done outside the ERANOS code in
this study. The strategy used for the recycling simulation
is as follows:
1. All the actinides are recovered and fission prod-
ucts are removed from the spent fuel through a dry re-
process after a 5-yr cooling period.
2. Because the fuel is preferentially burned in the
outer core while it is bred reasonably in the inner core, Fig. 8. Variation of the reference core reactivity.
material transfer between the inner and outer core spent
fuel is allowed to maximize the utilization of the spent
fuel.
3. Natural uranium is used as fertile feed and repro- is plotted in Fig. 9. It can be seen that the core requires
cessed plutonium is used as the fissile feed to control the plutonium feed up to the third recycle and discharges the
plutonium volume fraction. surplus actinides from the fourth recycle, which is con-
sistent with the BG change.
4. If the fissile breeding is too much from the core, The characteristics of the recycled core are deter-
surplus fuel material is sold ~discharged! after the dry mined by the fuel composition change from the initial
reprocess. core. As the breeding capability of the core gradually
increases through recycling, the fissile content of the
5. The fissile contents of the inner and outer cores fuel decreases as shown in Table V. However, the amount
are adjusted to maintain the initial plutonium volume of minor actinides will increase as the fuel is continu-
fraction ratio. ously recycled. The fraction of uranium isotopes de-
6. The target fissile content is determined to sustain creases while that of plutonium ~or higher actinides!
the criticality of the core in the subsequent fuel cycle. increases as the fuel is recycled. For the fissile isotope,
the 235 U fraction is negligible after seven recycles, and
The amount of feed material can be obtained from 239 Pu remains as the dominant fissile isotope.
the mass balance of the heavy metal and fissile isotopes.
In the mass balance calculation, the initial total heavy IV.B. Simulation with Different
metal inventory is balanced with the residual heavy metal, Design Parameters
fission products, fuel transfer, natural uranium feed, and
plutonium feed inventories. The initial fissile content is IV.B.1. Effect of the Fuel Volume
also balanced with the residual fissile and feed fissile
contents. The amount of feed material is then deter- From the parametric analyses of the design param-
mined by minimizing the total amount of the plutonium eters, it was found that the fuel fraction has the dominant
feed for the appropriate amount of the fuel transfer. effect on the BG. The advantage ~or disadvantage! of the
The recycling simulation was successfully per- different fuel fraction was assessed by the recycling sim-
formed for the reference core up to seven recycles ~51.4 ulation. The fuel fractions used for the simulation are 40
full power years including the initial core fuel cycle!. and 60%, and the results are shown in Fig. 10 for the
The results showed that the breeding gain becomes variation of the core reactivity. If the fuel fraction changes,
positive from the second recycle. The initial reactivity the cycle length also changes to meet the discharge burnup
and reactivity swing drop appreciably as the fuel is re- of 5 wt%. For the fuel fractions of 40 and 60%, the fuel
cycled ~Fig. 8!. If the fuel is continuously recycled, the cycle lengths are 1887 and 2796 FPDs, respectively.
higher actinide content increases and the safety-related If the fuel fraction is high ~60%!, the BG becomes
parameters ~void reactivity, Doppler reactivity, delayed positive ~0.00396! from the initial fuel cycle and, there-
neutron fraction, and prompt neutron generation time! fore, the core has to discharge the surplus fuel material
deteriorate. However the power distribution, represented continuously. Compared to the reference core, the amount
by the peak power densities of the BOL and EOL cores, of surplus fuel is 40 times higher as shown in Table VI
is well controlled because the reference plutonium vol- for the accumulated mass flow. If the fuel fraction is low
ume fraction ratio is maintained throughout the re- ~40%!, the BG is more negative and the core requires
cycling simulation. The mass flow of the reference core more fissile material. After seven recycles, the amount

NUCLEAR SCIENCE AND ENGINEERING VOL. 152 FEB. 2006

 A PHYSICS STUDY OF A 600-MW~thermal! GAS-COOLED FAST REACTOR 213

Fig. 9. Mass flow of the reference core. Fig. 10. Effect of the fuel fraction on the core reactivity.

of plutonium feed is five times higher in the low fuel estimate their impacts on the recycling of the reference
fraction core compared to the reference core. Therefore, core fuels. In this study, the process loss was varied from
it can be seen that the mass flow is significantly influ- 0.1 to 0.5%, and the fission products contamination fac-
enced by the fuel fraction, even though the comparison tor was changed from 1 to 5%.
was made after the same number of recycles not at the The result of the recycling simulation is shown in
same fuel burnup. Fig. 11 for the two cases: case one with the process loss
of 0.1% and fission products contamination by 1% and
IV.B.2. Effect of the Reprocess Parameters case two with the process loss of 1% and fission prod-
ucts contamination by 5%. It can be seen that there is
In the recycling simulation of the reference core, it almost no difference in the core reactivity variations even
was assumed that all the actinides are recovered for the if the process loss and fission product contamination are
subsequent use and the fission products are completely applied. However, these factors definitely have effects
removed. However, in reality, it is inevitable that there on the mass flow. Table VII shows the accumulated
is a process loss during the reprocessing of the spent amounts of the feed and surplus material, which are com-
fuel and the fabrication of the new fuel. Also it is possi- pared to the results of the reference core simulation. If
ble that some of the fission products may go with the there is a process loss, more natural uranium and pluto-
actinides because of their chemical property, though this nium feeds should be provided to compensate for the
requires proven experimental data. Therefore, at the stage loss of the fuel material. If the fission products are re-
of a feasibility study, it is reasonable to assume the pro- cycled together, the fuel will require more fissile mate-
cess loss and fission product contamination factors and rial and less fertile material to maintain the core reactivity.

TABLE V
Fuel Mass of the Reference Core

Beginning-of-Life End-of-Life

Fissile Minor Total Fissile Minor Total

Recycle Isotopes Actinides Actinides Isotopes Actinides Actinides

0 3176.6 33.3 28076.8 3091.8 141.1 26672.5

1 3177.8 191.6 28076.8 3126.0 236.3 26668.1
2 3144.9 273.5 28076.9 3119.9 290.4 26666.1
3 3109.4 323.3 28077.0 3104.5 325.7 26665.1
4 3078.5 357.3 28076.8 3088.8 350.8 26664.3
5 3053.8 382.1 28076.9 3075.5 369.3 26663.9
6 3034.4 400.3 28076.9 3064.7 382.9 26663.7
7 3019.4 413.8 28076.9 3056.3 393.0 26663.4

NUCLEAR SCIENCE AND ENGINEERING VOL. 152 FEB. 2006

214 CHOI, RIMPAULT, and BOSQ

the feed and surplus material were estimated after the

third and seventh recycles for which the integral power
generation is 5.63 and 11.26 TWd, respectively, for the
reference core. For the high power density core, the ac-
cumulated mass was estimated after the 6th and 13th
recycles, of which the electricity generation is 5.82 and
11.64 TWd, respectively. As given in Table VIII, an ini-
tial fuel loading is saved by 41% for the high power
density core. The use of natural uranium is slightly lower
in the high power density core, but the amount of pluto-
nium feed is three times higher. Because the fuel is con-
tinuously recycled in the core, only the surplus material
is discharged during the operation. In the high power
density core, the total amount of surplus material is three
times higher than that of the reference core. But, after
Fig. 11. Effect of the process loss and fission products
the final operation, the amount of spent fuel is lower in
contamination.
the high power density core by 41%. Therefore, the
amount of minor actinide generation is also lower in the
high power density core by 27%, which is shown in
Table IX.
IV.B.3. Effect of the Power Density
IV.C. Comparison of the Fuel Cycle Cost
The recycling simulation has been performed for the
high power density ~Pvol ⫽ 98.4 W0cm 3 ! core up to 13 Based on the mass flow, including the initial loading
recycles ~53.2 full power years including the initial core and final disposal, the fuel cycle cost ~FCC! can be es-
fuel cycle!. In this core, the BG becomes positive after timated. The basic method adopted for the calculation of
the fifth recycle. The reactivity swing drops appreciably the FCC is the constant money ~levelized lifetime cost!
in the high power density core as the fuel is recycled. method that is explained in the Organization for Eco-
However, as shown in Fig. 12, the reactivity changes nomic Cooperation and Development0Nuclear Energy
more frequently because of the shorter cycle length, and Agency ~OECD0NEA! report published in 1993 ~Ref. 20!.
the reactivity drop curve is more linear because of the To be consistent with recent studies on the future nuclear
low breeding capability when compared to the reference fuel cycle,21,22 the unit costs of the fuel cycle compo-
core. As for the safety-related parameters such as the nents were taken from the Generation IV roadmap report
void reactivity, Doppler reactivity, delayed neutron frac- of the fuel cycle crosscut group, which has used the con-
tion, and the prompt neutron generation time, they de- sensus values developed by the OECD0NEA expert
teriorate as the fuel is continuously recycled, which was group 23 :
also found from the reference core calculation.
1. The natural uranium purchase and conversion costs
The high power density core requires plutonium feed
are 50 and 8 $0kg, respectively.
continuously and discharges surplus actinides from the
third recycle as shown in Fig. 13, which is due to the 2. The reprocessed plutonium cost is 800 $0kg, which
unbalance of fissile breeding in the inner core and fissile is the separation cost of the plutonium from the
burning in the outer core. The accumulated amounts of commercial-grade spent fuel of the light water reactors.

TABLE VI
Effect of the Fuel Fraction on the Accumulated Mass Flow

Accumulated Mass Flow After 7 Recycles ~kg!

Fuel Matrix
Fraction Fraction Natural Reprocessed
~%! ~%! Uranium Plutonium Surplus

40 40 7 007.5 1791.9 0.0

50a 50 11 205.4 334.4 249.9
60 60 23 446.7 1.9 9917.6
a
Reference core

NUCLEAR SCIENCE AND ENGINEERING VOL. 152 FEB. 2006

 A PHYSICS STUDY OF A 600-MW~thermal! GAS-COOLED FAST REACTOR 215

TABLE VII
Effect of the Process Parameters on the Accumulated Mass Flow

Accumulated Mass Flow After 7 Recycles ~kg!

Process Fission Product
Loss Contamination Natural Reprocessed
~%! ~%! Uranium Plutonium Surplus

0.0 0.0 11 205.4 334.4 249.9

0.1 0.0 11 290.5 356.5 144.9
0.5 0.0 11 870.7 484.5 0.0
0.1 1.0 11 246.9 371.6 126.7
0.5 5.0 11 699.8 599.7 0.0

Fig. 12. Reactivity change for the high power density core. Fig. 13. Mass flow of the high power density core.

3. The dry reprocess and fabrication costs are 2000 where

and 2600 $0kg, respectively, which are used for the fast
reactor transuranic fuels. T1 ⫽ maximum value of the lead time in the front
end of the fuel cycle
4. The interim storage ~including transportation! and
T2 ⫽ maximum value of the lag time in the back end
disposal costs are 200 and 400 $0kg, respectively.
of the fuel cycle
The levelized FCC is derived in terms of mills0 r ⫽ discount rate.
kW{h by dividing the net present value ~NPV! of the
entire fuel cycle cost by the NPV of the total electricity To estimate the FCC, it is necessary to define several
output over the plant lifetime ~L!. The NPV is obtained parameters related to the fuel cycle model as follows:
based on the cash flow of all the component costs ~F! or
electric power ~E! discounted to the base year ~t0 !. There- 1. The lead time for the fuel purchase is 12 months
fore, the levelized FCC can be written as follows: before the start of each cycle.
t⫺t0⫹L⫹T2 2. The lead time for the fuel fabrication is 6 months
Fi ~t !
before the start of each cycle.
(i t⫽t(⫺T ~1 ⫹ r! ~t⫺t0 !
0 1
FCC ⫽ , 3. The lag time for the interim storage is zero.
t⫽t0⫹L
E~t !
( ~1 ⫹ r! ~t⫺t0 !
4. The lag time for the disposal is 360 months after
t⫽t0 the shutdown of the plant.

NUCLEAR SCIENCE AND ENGINEERING VOL. 152 FEB. 2006

216 CHOI, RIMPAULT, and BOSQ

TABLE VIII
Comparison of the Accumulated Mass After Recycling

Pvol ⫽ 58.2 W0cm 3 Pvol ⫽ 98.4 W0cm 3

Accumulated Feed ~kg! Accumulated Feed ~kg!

Recycle Natural Reprocessed Recycle Natural Reprocessed

~TWd! Uranium Plutonium Surplus ~TWd! Uranium Plutonium Surplus

0 23 320.7 4756.1 0.0 0 13 520.7 3087.7 0.0

~0.00! ~0.00!
3 5 301.9 334.4 0.0 3 2 750.6 576.7 0.0
~5.63! ~3.33!
7 11 205.4 334.4 249.9 6 5 062.0 844.7 78.0
~11.26! ~5.82!
13 10 821.8 1434.1 585.5
~11.64!

TABLE IX
Comparison of the Isotopic Mass of the Spent Fuel After Recycling

Recycle ~Pvol ⫽ 58.2 W0cm 3 ! Recycle ~Pvol ⫽ 98.4 W0cm 3 !

3 7 3 6 13

U 21 098.4 20 863.5 12 208.3 12 039.6 11 890.6

Np 25.2 34.7 14.3 18.9 22.8
Pu 5 241.0 5 406.9 3 369.1 3 488.7 3 594.7
Am 259.1 299.7 155.4 184.9 213.3
Cm 41.4 58.6 28.4 42.5 52.5
Fissile 3 104.5 3 056.3 1 956.0 1 924.7 1 894.2
Minor actinides 325.7 393.0 198.0 246.3 288.7
Total 26 665.1 26 663.4 15 775.4 15 774.6 15 774.0

5. The escalation of unit cost is 2%. the fraction of reprocessed plutonium feed cost is 21%.
Because the core is charged mostly with recycled fuel,
6. The discount rate is 5%.
the fraction of the natural uranium purchase cost is very
The thermal efficiency of the reactor system is 48%, low compared to the typical once-through fuel cycle.
which is used for the gas turbine modular helium reac- The FCC is sensitive to some of the reactor system
tor.24 The capacity factor was assumed to be 90%. The and fuel cycle parameters. For example, the sensitivities
reference value of the plant lifetime was set at 30 yr, of the FCC to the thermal efficiency and capacity factor
which is typical for conventional power reactors. The are ⫺120 and ⫺12%, respectively, while those to the
number of recycles that corresponds to this plant life- discount rate and plant lifetime are 8.8 and ⫺1.5%, re-
time is three and six for the reference and high power spectively. The FCC of the high power density core is
density cores, respectively. Then the FCCs are expected lower than that of the reference core by ;0.4 mills0
to be 9.40 and 8.99 mills0kW{h for the reference and kW{h, which is primarily due to the fuel fabrication for
high power density core, respectively. For the reference the initial core loading. For the initial fuel loading of the
core, the fractions of the transuranic fuel fabrication and high power density core, the amount of fuel fabrication
reprocessing cost are 58 and 19%, respectively, while is lower by 41% when compared to the reference core.

NUCLEAR SCIENCE AND ENGINEERING VOL. 152 FEB. 2006

 A PHYSICS STUDY OF A 600-MW~thermal! GAS-COOLED FAST REACTOR 217

V. SUMMARY AND CONCLUSION 4. P. FORTESCUE, G. MELESE-d’HOSPITAL, and L.

MEYER, “Design Studies for 1000 MWe Gas-Cooled Fast
Breeder Reactors,” GA-8581, Gulf General Atomic Inc. ~1968!.
The neutronic and recycling characteristics of a 600-
MW~thermal! GFR core have been assessed. The para- 5. “A Design and Assessment Study of Gas Cooled Breeder
metric calculations showed that a high BG is achievable Reactors,” European Association for Gas Breeder Reactor
in the reference core design in which the fuel fraction is ~1972!.
25% in the fuel assembly lattice along with a matrix
fraction of 25%. The parametric calculations also showed 6. E. L. QUINN, “New Nuclear Generation—In Our Life-
that good neutronic performance is achieved for the ref- time,” Nucl. News, 44, 52 ~Oct. 2001!.
erence core with the current design parameters. The re-
cycling simulation has shown that the initial BG of 7. P. HEJZLAR, M. J. DRISCOLL, and N. E. TODREAS,
⫺0.04755 is sufficient enough to breed the fissile mate- “The Long-Life Gas Turbine Fast Reactor Matrix Core Con-
rial necessary for subsequent fuel cycles. For both the cept,” Proc. 2002 Int. Congress Advanced Nuclear Power Plants
(ICAPP’02), Hollywood, Florida, June 9–13, 2002, American
reference ~58.2 MW0m 3 ! and high power density core Nuclear Society ~2002!.
~98.4 MW0m 3 !, the amount of surplus material is very
small. The simulations have also shown that the repro- 8. J. C. GARNIER, C. POETTE, B. MATHIEU, A. CONTI,
cessing parameters, represented by the process loss and and J. P. GAILLARD, “ Preliminary Design of an Advanced
fission products contamination, have only minor effects Gas Cooled Fast Reactor Core, Fuel Forms and Primary System
on the neutronic and recycling performance of the core. Concept,” Proc. 2003 Int. Congress Advances in Nuclear Power
The comparative analysis on the core power density Plants (ICAPP’03), Córdoba, Spain, May 4–7, 2003, Ameri-
showed that the high power density core is advantageous can Nuclear Society ~2003!.
because it produces less minor actinides and spent fuel
though it consumes more plutonium feed through its life- 9. J. E. MARTIN, Physics for Radiation Protection, John
time. The fuel cycle cost of the high power density core Wiley & Sons, Inc. ~2000!.
is lower by 0.4 mills0kW{h when compared to the ref-
10. J. C. GARNIER, N. CHAUVIN, P. ANZIEU, G. FRAN-
erence core design owing to the reduced amount of fuel COIS, T. Y. C. WEI, T. TAIWO, M. MEYER, P. HEJZLAR, H.
fabrication for the initial core loading. LUDEWIG, and A. BAXTER, “Feasibility Study of an Ad-
vanced GFR Design Trends and Safety Options Status of France
& U.S. Studies,” Proc. GLOBAL 2003, New Orleans, Louisi-
ACKNOWLEDGMENTS ana, November 16–20, 2003, American Nuclear Society ~2003!.

This work was done when H. Choi was attached to Com- 11. K. SAWA, S. SUZUKI, and S. SHIOZAWA, “Safety Cri-
missariat à l’Energie Atomique ~CEA! Cadarache Center dur- teria and Quality Control of HTTR Fuel,” Nucl. Eng. Des.,
ing his sabbatical leave from Korea Atomic Energy Research 208, 305 ~2001!.
Institute. The authors are grateful to M. Delpech and M. Sal-
vatores of CEA for the administrative and technical support 12. A. E. WALTAR and A. B. REYNOLDS, Fast Breeder
for this work. Reactors, Pergamon Press, New York ~1981!.

13. G. RIMPAULT, “The ERANOS Code and Data System

for Fast Reactor Neutronic Analyses,” Proc. PHYSOR 2002,
REFERENCES Seoul, Korea, October 7–10, 2002, American Nuclear Society
~2002!.
1. A. VASILE, G. RIMPAULT, J. TOMMASI, C. DE SAINT
JEAN, M. DELPECH, K. HESKETH, H. M. BEAUMONT, 14. G. RIMPAULT, K. SUGINO, and H. HAYASHI, “Gas
R. E. SUNDERLAND, T. NEWTON, P. SMITH, W. Cooled Fast Reactor Benchmarks for JNC and CEA Neutronic
MASCHEK, D. HAAS, C. DE RAEDT, G. VAMBENÈPE, Tools Assessment,” Proc. Int. Topl. Mtg. Mathematics and Com-
and J. C. LEFÈVRE, “Fast Reactors Fuel Cycle Core Physics putation, Super Computing Reactor Physics, and Nuclear Bio-
Results from the CAPRA-CADRA Programme,” Proc. logical Applications, Avignon, France, September 12–15, 2005,
GLOBAL 2001, Paris, France, September 9–13, 2001, Ameri- European Nuclear Society ~2005!.
can Nuclear Society ~2001!.
15. G. RIMPAULT, D. PLISSON-RIEUNIER, C. DE SAINT
JEAN, and J. TOMMASI, “Enhanced Streaming Algorithms
2. “A Technology Roadmap for Generation IV Nuclear for GCFR Core Depressurization Effect,” Proc. Int. Topl. Mtg.
Energy Systems,” GIF-002-00, U.S. Department of Energy Mathematics and Computation, Super Computing Reactor Phys-
~Dec. 2002!. ics, and Nuclear Biological Applications, Avignon, France, Sep-
tember 12–15, 2005, European Nuclear Society ~2005!.
3. P. FORTESCUE and G. M. SCHULTZ, “A 1000 MW~e!
Gas-cooled Fast Breeder Reactor,” GA-7823, General Dynam- 16. G. RIMPAULT, “Nuclear Data Needs for the Assessment
ics ~1967!. of GEN IV System,” Proc. Int. Workshop Nuclear Data Needs

NUCLEAR SCIENCE AND ENGINEERING VOL. 152 FEB. 2006

218 CHOI, RIMPAULT, and BOSQ

for Generation IV Nuclear Energy System, Antwerp, Belgium, 21. M. BUNN, S. FETTER, J. P. HOLDREN, and B. ZWAAN,
April 5–7, 2005, European Nuclear Society ~2005!. “The Economics of Reprocessing vs. Direct Disposal of Spent
Nuclear Fuel,” DE-FG26-99FT4028, Belfer Center for Sci-
17. R. JACQMIN and J. TOMMASI, “Integral Experiments in ence and International Affairs, John F. Kennedy School of Gov-
Support to GFR,” Proc. Int. Workshop Nuclear Data Needs ernment, Harvard University ~2003!.
for Generation IV Nuclear Energy System, Antwerp, Belgium,
April 5–7, 2005, European Nuclear Society ~2005!. 22. M. J. DRISCOLL and P. HEJZLAR, “Reactor Physics Chal-
lenges in GEN-IV Reactor Design,” Nucl. Eng. Technol., 37, 1
18. R. C. COCHRAN and N. TSOULFANIDIS, The Nuclear ~2005!.
Fuel Cycle: Analysis and Management, American Nuclear
Society, La Grange Park, Illinois ~1990!.
23. Trends in the Nuclear Fuel Cycle, Organization for Eco-
19. Nuclear Engineering Handbook, H. ETHERINGTON, Ed., nomic Cooperation and Development0Nuclear Energy Agency
McGraw-Hill Book Co., New York ~1984!. ~2002!.

20. The Economics of the Nuclear Fuel Cycle, Organization 24. M. P. LaBAR, “The Gas Turbine Modular Helium Reac-
for Economic Cooperation and Development0Nuclear Energy tor: A Promising Option for Near Term Deployment,” GA-
Agency ~1993!. A23952, General Atomics ~Apr. 2002!.

NUCLEAR SCIENCE AND ENGINEERING VOL. 152 FEB. 2006