Nuclear Engineering and Design 219 (2002) 207 /223

www.elsevier.com/locate/ned

A preliminary design study for the HYPER system

Won S. Park
, Tae Y. Song, Byoung O. Lee, Chang K. Park
Korea Atomic Energy Research Institute, P.O. Box 105, Yuseong, Daejeon 305-600, Republic of Korea
Received 19 October 2001; received in revised form 28 January 2002; accepted 30 April 2002

Abstract
In order to transmute the long-lived radioactive nuclides such as transuranics (TRU), Tc-99, and I-129 in LWR spent
fuel, a preliminary conceptual design study has been performed for an accelerator driven subcritical reactor system,
called HYPER (HYbrid Power Extraction Reactor). The core has a hybrid neutron energy spectrum which includes fast
and thermal neutrons for the transmutation of TRU and fission products, respectively. TRU are loaded into the
HYPER core in a TRU /Zr metal form because a metal type fuel has very good compatibility with the pyro-chemical
process which retains the self-protection of transuranics at all times. On the other hand, Tc-99 and I-129 are loaded as
pure technetium metal and sodium iodide, respectively. Pb /Bi is chosen as a primary coolant because Pb /Bi can
provide a good spallation target and produce a very hard neutron energy spectrum. As results, the HYPER system does
not need any independent spallation target system. 9Cr /2WVTa is used as a window material because this advanced
ferritic/martensitic steel is known to have a good performance in the highly corrosive and radiative environment. The
support ratios of the HYPER system are about 4 /5 for TRU, Tc-99, and I-129. Therefore, a radiologically clean
nuclear power, i.e. zero net production of TRU, Tc-99 and I-129 can be achieved by combining 4 /5 LWRs with one
HYPER system. In addition, the HYPER system, having good proliferation resistance and high nuclear waste
transmutation capability, is believed to provide a breakthrough to the spent fuel problems the nuclear industry is facing
with.
# 2002 Elsevier Science B.V. All rights reserved.

1. Introduction
Nuclear is being emerged as one of the most
promising sustainable energy sources for the 21st
century. Many evolutionary or innovative concepts are being investigated as a future nuclear
system. Not a single system is supposed to meet
the requirement, radiological cleanliness the future nuclear system has to satisfy. The combina-

Corresponding author
E-mail address: [email protected] (W.S. Park).

tion of power reactors with transmutation reactors

for incineration of waste is believed to be the
preferred option.
A LWR (Light Water Reactor) has been the
main reactor type in the Republic of Korea, and it
is expected to continue as this for next a few tens
of years. The operation of a LWR unloads about
23 tons of spent fuels per a year in order to
produce 1 GWe with the discharge burnup of
about 40,000 MWD/MTU. A transmutation technology has been developed to reduce the long-term
radiological toxicity of LWR spent fuel due to the

0029-5493/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 9 - 5 4 9 3 ( 0 2 ) 0 0 2 0 7 - 8

208

W.S. Park et al. / Nuclear Engineering and Design 219 (2002) 207 /223

long-lived radioactive nuclides such as Pu, MA

(Minor Actinide), and some fission products. A
dedicated transmutation reactor is estimated to
reduce the toxicity level of LWR spent fuel by a
factor of 100 within two or three hundred years
(OECD/NEA, 1998). The transmutation technology can provide many advantages: (1) the amount
of nuclear waste volume to be disposed is reduced
so that the repository size can be reduced; (2) the
repository design and construction cost can be
saved dramatically by shortening the management
period from almost infinite to two or three
hundred years; (3) the public acceptance for the
repository site can be enhanced by returning the
site after 2/3 hundred years, etc.
Various types of transmutation technologies
have been studied related with the Pu and MA
recycling scenario (Bowman et al., 1992; Rubbia et
al., 1995; Salvatores et al., 1997; Bowman, 1998;
Sasa et al., 2001; Knebel et al., 2001). A thermal
neutron system is not technically meaningful for
the transmutation of multi-recycled Pu and MA,
because of a low fission probability of most
isotopes of multi-recycled Pu and MA. As a
consequence, both critical and sub-critical fast
neutron systems are being studied as dedicated
transmutation systems in many countries. A
Comparative study of ADS and FR in an
advanced fuel cycles has been performed by the
OECD/NEA P&T expert group since 1999 and it
has been completed by the end of 2001. All subcritical systems being investigated throughout the
world utilize an accelerator to drive the system as
shown in Fig. 1. The accelerator for driving the
sub-critical system has to be able to provide large
proton currents with high energy. Thus, a linear
type accelerator is being considered in many
countries. The sub-criticality enables several advantages: (1) a fuel with a large fraction of MA can
be loaded into the sub-critical system without any
reactivity induced safety problems, which would
not be the case in a critical system; (2) a subcritical system has a higher transmutation capability than a critical system having the same
power capacity, because a critical system cannot
be loaded with pure TRU (Trans-uranium:Pu plus
MA) alone but has to include a certain amount of
uranium for the safety reasons; (3) the material

diversion difficulty can be maximized to enhance

the proliferation resistance because the sub-criticality results in a high fuel composition flexibility.
An accelerator driven sub-critical transmutation
system named HYPER (HYbrid Power Extraction
Reactor) is being developed within the framework
of the national mid- and long-term nuclear research plan. The basic system concept and the key
technical issues were developed in Phase I (1997 /
2000). Fig. 2 shows the fuel cycle scenario being
employed in the HYPER system design. There is
no pure plutonium stream. Plutonium is always
together with MA and plenty of lanthanides
(fission product). The pyro-chemical processes
are designed for both front- and back-end cycles
because the pyro-process retains the self-protection of transuranics at all times (ATW, 1999).
TRU /Zr was selected as a fuel type for the
HYPER system because of its good compatibility
with the pyro-chemical process and with the fast
neutron spectrum. Many types of core primary
coolants were investigated. The comparative study
decided Pb /Bi as a primary candidate for the
coolant because it can be also used as a spallation
target and has no fire problem (it is chemically
inactive). A single beam window of 9Cr /2WVTa
has been designed to transport the proton beam
from the accelerator to the sub-critical reactor
core. This advanced ferritic/martensitic steel, 9Cr /
2WVTa, is known to have a good performance in
the highly corrosive and radiative environment.
Some experiments will be performed to develop
Pb /Bi coolant/target technologies and to produce
out-pile data for TRU /Zr fuel in Phase II (2001 /
2003). The experiments for Pb /Bi will be performed by joining the multi-lateral collaboration
program, MEGAPIE (Salvatores and Heusener,
1999). In Phase III (2004 /2006), a conceptual
design for HYPER system will be done.
The results of the preliminary design study
performed during Phase I are introduced and
discussed in this paper.

2. Core concept
The basic requirements for the core design are to
maximize the proton beam current efficiency, the

W.S. Park et al. / Nuclear Engineering and Design 219 (2002) 207 /223

Fig. 1. The concept of the HYPER system.

Fig. 2. Cycle scenario of the HYPER system.

209

210

W.S. Park et al. / Nuclear Engineering and Design 219 (2002) 207 /223

cycle length, the support ratio, and the safety

features. Most of core analyses were done using
Monte Carlo transport code, MCNAP based on
ENDF/B-VI library (Chung et al., 2000).

2.1. Basic neutronic characteristics

Fast and thermal neutron spectra have been
compared in terms of their efficiency for the
transmutation of TRU and Tc-99, I-129 (Park et
al., 1999). The results showed that a fast neutron
spectrum is much more efficient for the transmutation of TRU in terms of accelerator beam power
economy, pin power peaking, and equally good
transmutation capability for all transuranics. On
the other hand, thermal neutrons are much more
effective for the transmutation of Tc-99 and I-129.
As results, the HYPER core has been designed to
have a hybrid neutron energy spectrum: a fast
neutron spectrum for the TRU burning zone and a
highly localized thermal neutron spectrum for the
Tc-99 and I-129 burning zone. A hexagonal type
array was employed to design a compact core
pitch-to-diameter and to achieve a hard neutron
energy spectrum for TRU burning. A study to
optimize the pitch-to-diameter (P/D ) ratio has
been performed by Park et al. (1999). The HYPER
core does not have an optimal value for P/D in
terms of neutronics. Instead, the P /D ratio has
been decided to be 1.48 for a thermal hydraulic
optimization. The sub-criticality level of the core
has been determined by considering the reactivity
changes due to the power defect, beam tube
rupture. The analysis showed that the sub-criticality of 0.98 is enough to keep the core subcritical under any anticipated conditions.
Table 1 represents the design parameters of the
HYPER core. The design analysis shows a Doppler coefficient of about /0.36 pcm K 1. The
coolant void and temperature coefficients are also
found to be negative though they are very small.
The homogeneous void coefficient for the beginning of cycle (BOC) is about /140 pcm per %
void. However, the local void coefficient in the
central region of the core is evaluated to be slightly
positive. The coolant temperature coefficient is
about /2.1 pcm 8C1.

Table 1
Design parameters of the core
Parameters [unit]

Values

System
Core thermal power [MW]
Core inlet/outlet temperature (8C)
Active core height [m]
Effective core diameter [m]
System multiplication factor
Accelerator beam power [MW]
Average discharge burnup [%at.]
TRU transmutation capability [kg per year]
Number of fuel assemblies
Average linear power density [kW m 1]
No. of burnable absorber rods per assembly

1000
340/510
1.6
2.6
0.98
/10 /20
/25
258
225
12.48
7

Assembly
Assembly pitch (cm)
Pitch-to-diameter
Tube thickness (cm)
Tube material
Rods per assembly

16.13
1.48
0.3556
HT-9
217

2.2. Core configuration

Part of the system output power (about 10 /
15%) is used to drive the accelerator for the
production of high energy protons in the HYPER
system. The system economy can be improved by
increasing the beam current efficiency, which is a
function of core geometry and of the TRU fuel
loading pattern.
A critical liquid metal core employing sodium
coolant would have very short core height (/1.0
m) in order to keep the void coefficient negative by
increasing the neutron leakage (Waltar and Reynolds, 1981). Keeping the void coefficient negative
is not a big concern because the HYPER system is
a sub-critical system utilizing Pb /Bi as a coolant.
Fig. 3 shows the variations of K -eff due to the
change of height-to-diameter ratio. The K -eff of
the HYPER system shows the maximum when H/
D is 0.62 (160 m). Therefore, the active core height
of the HYPER system is increased up to 1.6 m in
order to maximize the multiplication of spallation
neutrons while keeping the coolant flow velocity
below the design limit of less than 2 m s 1 (Kim et
al., 2001a). The axial position of the beam tube
also effects the spallation neutron economy and

W.S. Park et al. / Nuclear Engineering and Design 219 (2002) 207 /223

211

peaking. Table 2 shows the results of five different

radial configuration studies. The optimum radial
configuration of the HYPER core is determined
based on the Case 3 of Table 2 that has three
different TRU zones and fission product assemblies in the middle of the core. Case 3 produces an
acceptable power peaking and a reasonable transmutation capability for Tc-99 and I-129. A fuel
with a low TRU fraction is designed to be loaded
in the innermost zone and a high TRU fraction
fuel is loaded in the outermost region. Refueling
shall be performed based on a scattered loading
with multiple batches (7 /8) for each zone. Fig. 5
shows the designed core configuration.
Fig. 3. K -eff (ADS) variation due to the change of H /D ratio.

2.3. Support ratio

Fig. 4. Source efficiency as a function of beam tube axial

location.

the axial power shape. Fig. 4 shows the variation

of the neutron source efficiency as a function of
the beam tube axial position. The axial power
becomes a symmetrical shape and the neutron
source efficiency reaches at the maximum value
when the end of beam tube is located at about 10/
20 cm above the core mid-plane. Therefore, the
end of beam tube was determined to be at 15 cm
above the mid-plane in the HYPER core.
The core radial configuration (or fuel loading
pattern) affects the neutron source efficiency as
well. A uniform fuel loading would result in the
best neutron source economy because of a low
neutron leakage probability. However, the uniform loading produces an unacceptable power

The support ratio means how many LWRs can

be covered by one dedicated burner for a zero net
TRU production when each reactor has the same
power level. The HYPER system, the dedicated
burner in this case, produces power by transmutation or fission of TRU. In general, the maximum
achievable support ratio of a dedicated burner is
reached when all fissions are TRU transmutations.
As result, the maximum support ratio can not be
higher than /6.0 theoretically.
Transmutation inevitably needs recycling. The
HYPER system adopts the pyro-chemical process
because it has a poor nuclide separation capability.
Based on literature data, the pyro-chemical process is assumed to have 99.9% uranium recovery
rate and 90% lanthanide recovery rate (OECD/
NEA, 2001). As a consequence, the ratio of TRU
to uranium is 9 /1 when TRU are recovered from
LWR spent fuel for the transmutation in the
HYPER system. The HYPER core reaches a
pseudo-equilibrium condition of the nuclide composition approximately after the 20th cycle. The
fraction of uranium in heavy metal is about 20%.
The transmutation capability of the system is
about 250 kg of TRU a year when the capacity
factor is assumed to be 80%. About 90% of fissions
are from TRU transmutation in the HYPER
system. The other 10% are from the uranium.
Thus, the support ratio of the HYPER system is
about /5 if the same capacity factor can be

W.S. Park et al. / Nuclear Engineering and Design 219 (2002) 207 /223

212

Table 2
External neutron source multiplication for various core configurations
Case no.

Multiplicitya

Required beam (mA)b

Power peaking

Core configuration

1
2
3
4
5

39.4
35.7
25.8
16.9
31.4

12.05
13.51
17.55
25.71
14.22

1.859
1.273
1.223
1.327
1.289

Uniform loading
3 Zones without FP Ass.
3 Zones with FP Assemblies in the middle core
3 Zones with FP Assemblies in the inner core
3 Zones with FP Assemblies in the outer core

a
b

K -eff is 0.97 for all cases.

The beam current required to produce 1000 MWth.

assumed. Table 3 shows the variation of actinide

concentration at BOC and EOC of the 20th cycle.
2.4. Cycle length and burnable absorber
The HYPER core has a relatively small amount
of fertile nuclides compared to a typical liquid
metal reactor (LMR) core that has an uranium
fraction of more than 60%. In addition, there is no
excess reactivity. A burnable absorber concept
utilizing B-10 mixed with ZrH2 is adopted to
increase the cycle length by reducing the reactivity
drop rate. The burnable absorbers are installed
within the fuel assembly as a rod. Each fuel

assembly has seven burnable rods as shown in

Fig. 6. The burnable absorber was estimated to
reduce the reactivity swing by /40% for the same
cycle length (Kim and Park, 2000).
The reactivity drops still very rapidly, even
though the burnable absorbers are introduced in
the HYPER system. Therefore, the cycle length is
designed to be about 140 days only by considering
the limited beam current and power peaking
variation between BOC and EOC. The required
beam current for the pseudo-equilibrium is evaluated to be 1 GeV 11 and 22 mA of protons at
BOC and EOC, respectively, with the sub-critical
level of 0.98 at BOC.

Fig. 5. Core radial and axial configurations.

W.S. Park et al. / Nuclear Engineering and Design 219 (2002) 207 /223

Table 3
Variation of actinide concentration at the 20th cycle
Nuclide

Inventory (kg)

U-233
U-234
U-235
U-236
U-238
Np-237
Pu-238
Pu-239
Pu-240
Pu-241
Pu-242
Am-241
Am-242
Am-242m
Am-243
Cm-242
Cm-243
Cm-244
Cm-245
Cm-246

BOC

EOC

Variation

0.2144E/02
0.9166E/01
0.3203E/01
0.3949E/01
0.9675E/03
0.7585E/02
0.1456E/03
0.8774E/03
0.1299E/04
0.2797E/03
0.3898E/03
0.1049E/03
0.7905E/01
0.9818E/01
0.1191E/03
0.8502E/01
0.1090E/01
0.1142E/03
0.3564E/02
0.3084E/02

0.1902E/02
0.9215E/01
0.3121E/01
0.3986E/01
0.9516E/03
0.6977E/02
0.1432E/03
0.8082E/03
0.1270E/04
0.2674E/03
0.3845E/03
0.1010E/03
0.7898E/01
0.9773E/01
0.1182E/03
0.8498E/01
0.1081E/01
0.1142E/03
0.3571E/02
0.3072E/02

/0.0002
/0.039
/0.082
/0.037
/15.9
/6.08
/2.40
/69.2
/29.0
/12.3
/5.3
/3.9
/0.0001
/0.045
/0.9
/0.004
/0.009
0.0
/0.07
/0.12

213

loaded into different sites because of their different

neutronic characteristics. TRU is to be loaded as a
fuel to drive the system. Fission products, on the
other hand, are only a target to be hit by neutrons
for transmutation.
3.1. Basic form of TRU fuel
A transmutation system needs a recycling process with sufficient proliferation resistance. One of
the basic requirements for the selection of the
HYPER system fuel type is a good compatibility
with a proliferation resistant process such as the

Fig. 6. Burnable absorber loading concept.

3. TRU fuel and fission product target

Both, TRU and Fission Products (FP) are to be
loaded into the HYPER system for transmutation.
As they have different loading types, they will be

Fig. 7. Cross-sectional view of alloy fuel and dispersion fuel.

W.S. Park et al. / Nuclear Engineering and Design 219 (2002) 207 /223

214

pyro-chemical process. As the loss of radioactive

nuclides can not be avoided at each recycling
process, a high burnup capability is required as an
selection criteria in order to minimize the loss as
low as possible. A metallic fuel, among other fuel
types such as oxides and nitrides has been selected
because of its good compatibility with the pyrochemical process, with high burnup requirement,
and with fast neutron spectra. Either a TRU /Zr
metal alloy or a (TRU /Zr)/Zr dispersion fuel is
considered as fuel for the HYPER system. In the
case of the dispersion fuel, particles of 90 wt.%
TRU /10 wt.% Zr metal alloy are dispersed in a Zr
matrix. Fig. 7 shows a typical cross sectional view
of alloy and dispersion fuel rods.
If the fuel particles are separated sufficiently, the
areas damaged by fission fragments will not overlap and the matrix remains a continuous metal
phase that has negligible damages from fission
fragments. This relatively undamaged metal matrix can withstand higher burnups without significant swelling. Therefore, it is expected that the
dispersion fuel will generally withstand significantly higher burnup than alloy fuel. In addition,
the dispersion fuel requires only a much smaller
gas plenum than the alloy fuel, because the
dispersion type fuel exhibits less swelling.
Both, alloy and dispersion type fuels, still have
several problems in the fabrication. If the Zr
content exceeds more than 10%, the fabrication
process for the alloy type needs much higher
operating temperatures so that there may not be

an acceptable crucible. In addition, the high

operating temperature causes americium to be
evaporated in the fabrication process of the alloy
type. On the other hand, the maximum allowable
TRU /Zr particle content is limited in the dispersion fuel because of fabrication and swelling
problems. From a fabrication point of view, the
maximum allowable content is 60 wt.% (TRU /
Zr) /40 wt.% Zr based on the technical experience
of silicide dispersion fuel of 38 vol.% U3Si /62
vol.% Al (Wood et al., 1984). This means that
(TRU /Zr)/Zr dispersion fuel cannot be fabricated by the fuel core extrusion method if the Zr
content is less than 40 wt.%. Table 4 describes the
fuel design parameters determined for the HYPER
system.
3.2. TRU fuel performance
A couple of fuel performance analyses have
been done using the MACSIS-H (alloy type fuel
analysis) and DIMAC (dispersion type) codes
developed by KAERI. There is no experience on
TRU /Zr type metal fuel. Most experimental data
and experience implemented into the codes are
with U/Zr or U/Pu /Zr types that have a Zr
fraction of no more than 10%. Fig. 8 shows the
radial temperature distribution at the BOL condition. In the case of the dispersion fuel, the
temperature difference between the center and
the cladding surface is 150 K at a linear power
of 21.6 kW m1. On the other hand, the alloy fuel

Table 4
Design specifications of fuel rod
Parameters

Fuel type

Alloy fuel

Dispersion fuel

Fuel slug

Fuel diameter (mm)

Composition
Density (g cm 3)
TRU density (g cm 3)

4.58
50 wt.% TRU-50 wt.% Zr

12.36
6.18

5.18
45 wt.% (TRU /10 Zr) /55 wt.% Zr

9.16 (TRU /10Zr: 18.37)
3.7

Integrated gap between fuel slug and cladding (mm)

0.7 (75% SD)

0.1 (engineering gap)

Cladding (mm)

5.28
6.70
0.71

5.28
6.70
0.71

12.48

12.48

Inside dia
Outside dia
Thickness

Linear heat rate (kW m 1)

W.S. Park et al. / Nuclear Engineering and Design 219 (2002) 207 /223

215

Fig. 8. Temperature distribution in fuel rod.

Fig. 9. Fission gas release rate in the alloy type fuel.

shows the difference of 330 K. The thermal

conductivity is a function of Zr fraction and the
way of fabrication. In this case, the thermal
conductivity of the dispersion type is better than
that of alloy type although the fraction of Zr or
TRU is the same (Meyer and Lee, 2000; Hwang et
al., 2001).
Fig. 9 shows the fraction of the fission gas
released as a function of fuel burnup for four
different Zr fractions in the case of alloy fuel.
More than 90% of the fission gas is released at 5

at% burnup. The simulations predict that the

fission gas release rate is almost independent of
Zr fractions when the Zr weight fraction varies
from 45 to 55%. The total swelling of the dispersion fuel is the sum of three components; a volume
change due to transformation to a higher Zirconium phase as a result of TRU burnup, a volume
increase due to the accumulation of non-gaseous
fission products, and a volume increase due to
fission gas accumulation. Fig. 10 shows the
volume change due to each component at 10 at%
burnup. The fission gas accumulation adds the
largest contributions to the swelling in the dispersion type fuel. However, the dispersion fuel is
expected to allow higher discharge burnup because
the area of the accumulated fission gas is not
overlapped and the matrix has a continuous metal
phase.
Fig. 11 shows the results of the rod deformation
analysis for the alloy and dispersion type fuels.
The deformation rate of the alloy type starts to
increase abruptly after 7 at.% burnup and exceeds
3% at 20 at.% burnup, which is believed to be the
design limit. The maximum achievable discharge
burnups are expected to be 30 and 20 at.% for the
dispersion type fuel and the alloy type fuel,
respectively (Lee et al., 2001). As results, the alloy
type fuel may not meet the discharge burnup target
of 25 at%.

216

W.S. Park et al. / Nuclear Engineering and Design 219 (2002) 207 /223

Fig. 10. Volume changes of the dispersion fuel at 10 at.% burnup.

3.3. Fission products for the transmutation

The fission products that deserve most attention
are Tc-99, Cs-135, and I-129. Tc-99 and Cs-135 are
the dominant isotopes in risk analyses of spent fuel
disposals. I-129, which is not incorporated in the
vitrified HLW, is the dominant isotope in the
radiological effects of reprocessing effluents or
even from spent fuel in certain geological formation. Unlike Tc-99 and I-129, cesium separated
from spent fuel is not a single isotope but a
mixture of long-lived Cs-135, the short-lived Cs-

137 and the stable isotope Cs-133. All Cs isotopes

are present in about equal quantities. As a consequence, a parasitic capture in especially Cs-133
will occur during irradiation. Taking also into
account the relatively low neutron absorption
cross sections of Cs-135, the transmutation of
cesium is not considered feasible in most cases.
Therefore, only Tc-99 and I-129 are the fission
products to be transmuted in the HYPER system.
The preliminary results of the basic material
studies have shown that a pure metallic form is the
most desirable one for Tc-99: a fabrication route

Fig. 11. Rod deformation rate for the dispersion and alloy type.

W.S. Park et al. / Nuclear Engineering and Design 219 (2002) 207 /223

217

for casting the Technetium metal has been developed and irradiation experiments did not show
any evidence of the swelling or disintegration of
the metal (OECD/NEA, 1998). On the other hand,
an elemental form is found to be unacceptable for
Iodine because of its volatility and chemical
reactivity. Thus, metal iodides are being considered. Sodium iodide (NaI) and calcium iodide
(CaI2) are the desirable forms. Sodium iodide is
expected to have melting problems when the
sodium is liberated from iodide due to the
transmutation.
3.4. Fission product target design configuration
(Park et al., 2000)
Thermal neutrons are much more efficient for
the transmutation of Tc-99 and I-129 than fast
ones. The evaluation shows that the transmutation
rate can be improved considerably then. In order
to have a thermal neutron spectrum in the
HYPER system, some moderation has to be
included. Graphite and calcium hydride have
been studied in terms of their effectiveness for
moderation. The comparative study shows that a
calcium hydride is much better for the production
of localized thermal neutrons.
The introduction of a moderator causes power
peaking problems in the TRU assemblies that
surround the FP target. Many types of target
configurations were investigated in terms of power

Fig. 12. Configuration of the fission product assembly.

Fig. 13. Neutron energy spectrum nearby FP assembly.

peaking and transmutation rate. The configuration where Tc-99 is loaded as a plate type in the
outer-most region and I-129 is loaded as NaI rods
mixed with CaH2 rods in the inner region was
concluded to be the best. Fig. 12 shows the
configuration of FP target. Monte Carlo transport
code, MCNAP has been used with ENDF/B-VI
data library for the design of fission product target
as it is in the core design. Fission product target
and the TRU assemblies surrounding fission
product target were described by pin-by-pin
model. The designed FP target configuration is
estimated to have the transmutation capability of
57.8 and 13 kg per year for Tc-99 and I-129,
respectively when 6 FP targets are loaded. The
support ratios of the HYPER system for Tc-99
and I-129 are estimated to be 5.7 and 4.0,
respectively. The support ratio for Tc-99 is much
larger while that for I-129 is slightly less compared
with the support ratio for TRU.
Fig. 13 shows the neutron energy spectra in the
region of fission product target. The spectrum in
the fuel assemblies surrounding the FP target is
relatively softer than that of normal fuel assemblies. This softened spectrum causes an increment
to the local power. The calculation predicted a
local power peak of 1.232, which is within the
acceptable range. In addition, the loading of FP
targets makes the core coolant void coefficients
more negative, but the Doppler coefficient less
negative. It can be concluded that the FP target

218

W.S. Park et al. / Nuclear Engineering and Design 219 (2002) 207 /223

loading into the HYPER core does not cause any

severe safety problems.

4. Cooling system
4.1. Basic configuration
A Lead /Bismuth (Pb /Bi) eutectic alloy was
selected as a coolant material for the HYPER
system. The advantages expected from Pb /Bi
cooling are:
1) Pb /Bi has good nuclear properties, i.e. low
absorption cross-sections and low moderating
power, which is quite desirable for the incineration of TRU.
2) Pb /Bi can be used as an excellent spallation
target itself. Using Pb /Bi as a coolant removes the necessity of an independent target
and its cooling system.
3) Pb /Bi has good cooling capabilities and heat
transfer characteristics, i.e. high boiling temperature and high heat transfer rate.
4) Pb /Bi coolant system has enhanced safety
characteristics. It has no violent reaction with
air or water and it is an effective gamma
shielding material.
On the other hand, some problems have to be
solved for the use of Pb /Bi as a coolant. Pb /Bi is
highly corrosive to structural materials. The corrosion problems are major issues, which limit wide
use of Pb /Bi as a coolant material. In addition,
Pb /Bi requires higher pumping power than alkali
metals or water to operate a cooling system
because of its higher density.
The thermal efficiency of the power cycle
strongly depends on the temperature at which
heat is supplied by the primary to the secondary
coolant. It is obvious that the coolant temperature
should be set as high as possible for an optimum.
However, mechanical and corrosion characteristics of structural materials set the upper limit.
According to Russian results (Adamov et al., 1997;
Yachmenyov et al., 1998), the maximum allowable
temperature of Pb /Bi coolant is approximately
650 8C. The lower limit of the coolant tempera-

ture can start from the Pb /Bi melting point, i.e.

125 8C. For safe operation, Pb /Bi temperature
must be sufficiently above 125 8C. Therefore 125
and 650 8C can be the general temperature limits
of Pb /Bi coolant. The core inlet and outlet
temperatures of Pb /Bi coolant are decided to be
340 and 510 8C, respectively by considering the
factors mentioned above.
The coolant velocity of the primary cooling
system can also cause a design constraint. Coolant
velocity affects the integrity of structural materials
as well as the pump work. The primary cooling
system of HYPER should be designed with low
coolant velocity as long as it can satisfy the other
design requirements. Since Pb /Bi does not significantly absorb or moderate neutrons, it allows
the use of a wide lattice that favors lower coolant
velocity. The P /D ratio of the HYPER core is
chosen to be 1.48 and the corresponding Pb /Bi
velocity is 1.5 m s 1, which is a relatively low
coolant velocity compared to that of typical LMR
reactors using sodium as coolant. For instance, the
velocity of coolant in the Korea Advanced Liquid
Metal reactor (KALIMER) is about 5 m s 1 (Kim
et al., 2001b). Instead of wire spacers commonly
used for tight lattices, grid spacers are more
suitable to ensure proper separation of the fuel
rod here.
A loop type configuration has been selected for
the preliminary design of the HYPER system and
a study has been performed to optimize the
number of loops. The number of loops is determined by considering the coolant velocity and
pressure drop across the loop. Fig. 14 shows the
results. A three-loop concept is estimated to be the
optimum. The mass flow rate for each loop is
15,523 kg s 1. Each loop has one super-heater and
two evaporators. It is possible to eliminate the
intermediate heat transport system with Pb /Bi
coolant. A steam cycle is adopted for the HYPER
system because it has shown a long and successful
experience. Fig. 15 shows the overall view of the
HYPER cooling system.
4.2. Preliminary performance analysis
Although, the Pb /Bi velocity within the fuel
channel is sufficiently reduced, the roughly esti-

W.S. Park et al. / Nuclear Engineering and Design 219 (2002) 207 /223

219

Fig. 14. Velocity and pressure variation vs. no. of loops.

mated pressure loss across the reactor vessel is

about 5 atm similar to that of typical LMR
reactors (Tang et al., 1978). This is mainly due to
the high density of Pb /Bi. The complex geometry
of the inlet and outlet components of the core also
contributes to a large amount of the pressure loss.
Therefore, it is expected that natural circulation
does not play an important role in cooling the
HYPER core under normal operating conditions.
In order to increase the natural convection force,
the height difference between core and steam
generator has to be increased. Fig. 16 shows the

fraction of natural convection force as a function

of the height difference in the HYPER system. The
pressure loss of the primary side in the HYPER
system is expected to be more than 1.0 MPa.
Therefore, the contribution of natural convection
can not be more than 10% even in the case of a 30
m height difference.
A sub-channel analysis was performed using
SLTHEN (Steady-state LMR core Thermal Hydraulics code based on Energy model) for a typical
single fuel (TRU) assembly of the HYPER core
(Yang, 1997; Kim et al., 2001c). The maximum

Fig. 15. Cooling system of the HYPER.

220

W.S. Park et al. / Nuclear Engineering and Design 219 (2002) 207 /223

Table 5
Results of the sub-channel analysis

Fig. 16. The fraction of natural convection for various height

differences.

coolant outlet temperatures of the average and the

hot assembly cases were estimated to be higher
than the average coolant outlet temperature by
14.2 and 22.88, respectively, as shown in Fig. 17,
when flow split and heat transfer between subchannels were considered. Table 5 shows the
results of the sub-channel analysis for the average
and hot assemblies. Additional sensitivity calculations were performed for the various inter assembly gap flow rates and turbulent flow mixing. The
maximum coolant and cladding temperatures,
which are major parameters in the conceptual

Parameter

Average assembly Hot assembly

(Fz /1.2, Fr /1.0) (Fz/1.2, Fr /
1.6)

Average exit coolant

temperature (8C)
Peak coolant temperature (8C)
Peak clad temperature (8C)
Pressure drop (kPa)

510.0

612.0

524.2

634.8

536.7

654.7

32.6

32.8

Inlet velocity (m s 1 )
Average
Interior
Edge
Corner
Inter gap

1.433
1.449
1.382
1.097
0.000

1.433
1.449
1.382
1.097
0.000

design stage, were not largely affected by the

turbulent mixing in HYPER design conditions.

5. Beam window and spallation target

5.1. Basic design requirements and configuration
In order to supply an additional neutron source
for the operation of the sub-critical system, a
proton beam is produced in the accelerator and

Fig. 17. Hot subchannel temperatures in average and hot assemblies.

W.S. Park et al. / Nuclear Engineering and Design 219 (2002) 207 /223

221

effectiveness of beam and to minimize the radiation damage and mechanical stress of the beam
window (Song et al., 2000).
5.2. Performance analysis

Fig. 18. Beam target lay out.

injected into the center of the HYPER core

through beam window as shown in Fig. 1. As
mentioned, the Pb /Bi coolant flowing through the
central target channel is used as the spallation
target. The factors affecting the lifetime of the
beam window are corrosion by Pb /Bi and radiation damage. 9Cr /2WVTa is used as a window
material because an advanced ferritic/martensitic
steel is known to have a good performance in the
highly corrosive and radiative environment. A
single beam window is adopted so that there is
no independent window cooling system required.
The following design goals were chosen for a
stable and safe operation of the target and for a
reasonable lifetime of the beam window */the
maximum allowable temperature and stress of
the beam window are set at 600 8C and 180
MPa, respectively. The temperature limit is
decided to avoid corrosion problems. The stress
limit, 180 MPa is 1/3 of the yield strength for 9Cr /
2WVTa. The lifetime of the beam window is set to
be 1 year.
Fig. 18 shows the cross-sectional view of the
beam target. A parabolic beam shape is adopted to
avoid high thermal stresses at the beam boundary.
The radius of the beam is adjusted to maximize the

The designed beam target produces 1.88 /1017

neutrons per 1 mA of 1 GeV protons. The average
energy of the spallation neutrons is estimated to be
14 MeV. When the beam current is 13 mA, the
maximum temperatures of the window and Pb /Bi
become 534 and 499 8C, respectively. The maximum Pb /Bi velocity is 2.1 m s1. That is
supposed to be slightly high and will cause an
erosion problem. The maximum mechanical and
thermal stresses of the window are 155 and 104
MPa, respectively. Preliminary design analyses
using FLUENT (Fluent Tutorial Guide, 1993)
and ANSYS (ANSYS Users Manual Rev. 5.0)
codes show that the current design configuration
meets the requirements in terms of thermal hydraulics.
The effect of radiation damage is dominant in
deciding the window lifetime. LCS (LAHET Code
System) (Prael and Lichtenstein, 1989) is employed
to evaluate dpa (displacements per atom) and He
production rates due to the protons and spallation
neutrons. When the beam current is 13 mA, with a
parabolic distribution and circular shape, the
maximum radiation damage is 76 dpa per year at
the window center. About 80% of dpa are caused

Fig. 19. Radiation damage (dpa) distribution at the window.

222

W.S. Park et al. / Nuclear Engineering and Design 219 (2002) 207 /223

Table 6
Characteristics of the target design
Parameter
Beam window
Material
Structure
Diameter/thickness
(cm)

Characteristics

9Cr /2WVTa
Cylinder with circular shape end (single
window)
40/0.2

Beam
Incoming beam

1 GeV proton with 35 cm diameter

(parabolic shape)
Neutron production 1.88/1017 s 1 mA
Average neutron en- 14 MeV
ergy
Radiation damage
Maximum dpa
Maximum He production

5.85 dpa per year mA

277 appm per year mA

Thermal damage (I/13 mA)

Mechanical stress
155 MPa
Thermal stress
104 MPa
Maximum tempera- 534 8C
ture

by the spallation neutrons. On the other hand,

most of the He production is caused by protons
and is calculated to be 3598 appm per year at the
window center. Fig. 19 shows the radiation
damage (dpa) distribution in the window.
The spallation reaction produces more than 800
different radioactive nuclides. One of the dominant long-lived radioactive nuclides due to the
spallation is Pb-205. About 151.6 g of Pb-205 is
generated when 1 GeV, 13 mA protons are injected
into Pb /Bi target for 1 year. Table 6 summarizes
the design characteristics of the beam target.

mutually exclusive requirements. The core geometry is designed based on the spallation neutron
efficiency rather than the negative void coefficient
concern. The uranium recovery rate from LWR
spent fuel is assumed to be 99.9% for the core
analysis. The uranium fraction becomes about 20
wt.% of heavy metal when the core nuclide
composition reaches at a pseudo equilibrium
status. In order to reduce the reactivity swing
due to the lack of fertile nuclides, a burnable
absorber is considered. Either alloy or dispersion
type metallic fuels are being considered for the
HYPER system. A dispersion type fuel is supposed
to give better irradiation performance compare to
an alloy type fuel. Pb /Bi is used as a coolant and a
loop concept is preferred as a primary cooling
system. The basic analysis shows that three loops
are optimal to the HYPER system. Each loop has
one super heater and two evaporators. A single
window with a thickness of 2 mm 9Cr /2WVTa is
designed to provide the HYPER core with 1 GeV,
10/20 mA protons. The HYPER system is expected to transmute about 258 kg of TRU, 57.8 kg
of Tc-99 and 13 kg of I-129 and produce 1000
MWth through the transmutation process. The
support ratios of the HYPER system are about 4 /
5 for TRU, Tc-99, and I-129. Therefore, a radiologically clean nuclear power, i.e. with zero net
production of TRU, Tc-99 and I-129 can be
achieved by combining 4/5 LWRs with one
HYPER system.
The developed concept is supposed to have
some uncertainties because there are not much
experimental data available on TRU /Zr fuel and
Pb /Bi coolant/target. Some experiments are
scheduled to verify those fuel and coolant/target
design concepts during 2001/2003.

6. Summary
The design goal of the HYPER system is to
transmute TRU, Tc-99, and I-129. Major consideration has been given to the maximization of the
transmutation capability. A fast neutron spectrum
is more preferable for the transmutation of TRU
while a thermal neutron spectrum is better for Tc99 and I-129. The HYPER core is designed to have
hybrid neutron spectra in order to meet both

Acknowledgements
The authors give many thanks to Dr Yong H.
Kim and Nam I. Tak for their invaluable contributions. The authors also appreciate the financial support of MOST (Minister of Science and
Technology) for the HYPER system development.

W.S. Park et al. / Nuclear Engineering and Design 219 (2002) 207 /223

References
ATW, 1999. A Roadmap for Developing Accelerator Transmutation of Waste (ATW) Technology, DOE/RW-0519.
Adamov, E., et al., 1997. The next generation of fast reactors.
Nuclear Engineering and Design 173, 143 /150.
ANSYS Users Manual for Revision 5.0.
Bowman, C. et al., 1992. Nuclear Energy Generation and Waste
Transmutation Using an Accelerator Driven Intense Thermal Neutron Sources. LANL Report LA-UR-91-2601,
LANL.
Bowman, C., 1998. Once-Through Thermal Spectrum Accelerator-Driven System for LWR Waste Destruction without
Reprocessing: Tier-1 Description, ADNA Report, ADNA
Corporation, ADNA/98-04, August.
Chung, Jong Sung et al., 2000. Verification of SMART
Neutronics Design Methodology by the MCNAP Monte
Carlo Code. ANS/ENS International Meeting, Washington
DC, USA.
FLUENT Tutorial Guide, 1993. Version 4.2, Fluent Incorporated.
Hwang, W. et al., 2001. Analysis on the Temperature Profile
and the Thermal Conductivities of the Metalic and the
Dispersion Fuel Rods for HYPER. Korean Nuclear Society
Spring Meeting.
Kim, Chang H. et al., 2001. Sub-channel Analysis of HYPER
Single Fuel Assembly Using SLTHEN Code. Korean
Nuclear Society Fall Meeting.
Kim, Y.H., Park, W.S., Song, T.Y., 2001. Optimization of H/D
(Height-to-Diameter) Ratio in ADS. Korean Nuclear
Society Spring Meeting.
Kim, Y.H., Park, W.S., 2000. A Study on Burnable Absorber
for a Fast Subcritical Reactor HYPER. 6th OECD/NEA
P&T Meeting, Madrid, Spain.
Kim, Young G. et. al., 2001. Thermal Hydraulic Conceptual
Design Characteristics of KALIMER Breakeven Core.
Korean Nuclear Society Spring Meeting.
Knebel, J. et. al., 2001. Thermalhydraulic and Material Specific
Investigations into the Realization of an Accelerator Driven
System (ADS) to Transmute Minor Actinides FZK Report,
FZK6618.
Lee, B. et al., 2001. A Mechanistic Deformation Model of TRU
Metal Dispersion Fuel for HYPER. Korean Nuclear
Society Spring Meeting.
Meyer, M.K., Lee, B.O., 2000. Thermal Conductivity of
Zirconium Matrix Dispersion Fuel Inter-Office Memorandum, MKM-00-DRAFT, ANL.
OECD/NEA, 1998. Status and Assessment Report on Actinide
and Fission Product Partitioning and Transmutation System Studies, NEA/PTS/DOC(98)4.

223

OECD/NEA, 2001. NDC Comparative Study of ADS and FR

in Advanced Nuclear Fuel Cycle, to be published at the end
of 2001.
Park, Won S. et al., 1999. TRU Incineration Characteristics of
Thermal and Fast Subcritical Reactors. Annals of Nuclear
Energy, vol. 26.
Park, W. S. et al., 2000. Fission Product Target Design for
HYPER System. 6th OECD/NEA P&T Meeting, Madrid,
Spain.
Prael, Richard E., Lichtenstein Henry, 1989. User Guide to
LCS: The LAHET Code System, LA-UR-89-3014 Los
Alamos National Lab.
Rubbia, C. et al., 1995. Conceptual Design of a Fast Neutron
Operated High Power Energy Amplifier CERN Report,
CERN/AT/95-44(ET), European Organization for Nuclear
Research.
Salvatores, S. et. al., 1997. The Potential of AcceleratorDriven
Systems
for
Transmutation
or
Power
Production
Using
Thorium
or
Uranium
Fuel Cycles, Nuclear Science and Engineering 126,
333-340.
Salvatores, Bauer, Heusener, 1999. The MEGAPIE InitiativeExecutive Outline and Status as per November 1999. MPO1-GB-6/0_GB.
Sasa, T. et al., 2001. Conceptual Design Study of ADS
Experimental Facilities, Accelerator Applications/Accelerator Driven Transmutation Technology and Applications 01
(AccApp/ADTTA 01), Reno, Nevada, USA, November
12 /15.
Song, T.Y. et al., 2000. Thermal and Stress Analysis of HYPER
Target System. 6th OECD/NEA P&T Meeting, Madrid,
Spain.
Tang, Y.S. et al., 1978. Thermal Analysis of Liquid /metal Fast
Breeder Reactors American Nuclear Society.
Waltar, Alan E., Reynolds, Albert B., 1981. Fast Breeder
Reactors Pergamon Press.
Wood, J.C. et al., 1984. Reduced Enrichment Fuels for
Canadian Research Reactors /Fabrication and Performance
Proc. of the 1984 International meeting in RERTR,
Argonne, IL, ANL/RERTR/TM-6, October 15 /18, pp.
167 /183.
Yachmenyov, G.S. et al., 1998. Problems of Corrosion
of Constructral Materials in Lead-Bismuth Coolant.
Proc.
Int.
Conf.
on
Heavy
Liquid
Metal
Coolants in Nuclear Technologies, Obninsk, vol. 1, pp.
133 /140.
Yang, W.S., 1997. An LMR core thermal-hydraulics code
based on the ENERGY model. Journal of Korean Nuclear
Society 29 (5), 406 /416.