study on fast spectrum bwr core for actinide recycle

12 Pages·2005·0.26 MB·English

FR0108162
STUDY ON FAST SPECTRUM BWR CORE FOR ACTINIDE RECYCLE
M. YAMA0KA(1), Y. YAMAM0T0(1), N. ABE(1), K. HIRAIWA(2) and T.YOKOYAMA(2)
(1) Toshiba Corporation, Power and Industrial Systems Research and Development Center,
4-1,Ukishima-cho, Kawasaki-ku, Kawasaki 210-0862-Japan
e-mail; [email protected]
(2) Toshiba Corporation, Isogo Nuclear Engineering Center,
8, Shinsugita-Cho, Isogo-Ku, Yokohama 235-8523 - Japan
Key words: BWR, fuel, hydraulics
A study has been performed on neutronic and thermal characteristics of a fast
spectrum BWR core with tight fuel lattice for an innovative fuel cycle system named
BARS (BWR with an Advanced Recycle System) aiming at Pu multi-recycling and
MA's (Minor Actinides) burning. The coolant void reactivity tends to shift to positive
direction due to the fast neutron spectrum, so the neutron streaming channel and axial
enrichment distribution was introduced in the core design to overcome the tendency. It
was found that negative void reactivity is attained for whole burnup cycle by using
Monte Carlo code. As for the thermal characteristics, the critical power measurement
test was performed in order to get the database and the result was compared with the
Arai's critical power correlation. As the result, Arai's correlation can predict our critical
power test data considerably. Through our core design study and benchmark tests, we
will make the conceptual design of BARS core successfully.
Introduction
The LWR will play an important role in nuclear power generation for a longer time than
expected before owing to the delay of commercialization of the LMFBR. It is one of the
promising candidates for recycling various actinides such as U,Pu,and Minor
actinides(MAs) because it is well demonstrated. In particular, BWR with tight fuel
lattice has an advantage in breeding or high conversion of fissile plutonium as boiling
in water decreases the amount of moderator to induce a fast neutron spectrum. The
fast spectrum is also advantageous in multi-recycling of Pu because the production of
higher isotopes are suppressed. Moreover, it helps to burn fuels with recovered MAs
and low decontaminated FPs such as rare earth's (REs), which endows the recycling
system with nuclear proliferation resistance.
Taking into account the advantage of the fast spectrum BWR, an innovative fuel cycle
system named BARS (BWR with an Advanced Recycle System) is proposed as a
oooe
future fuel cycle option aiming at enhanced utilization of uranium resources and
reduction of radioactive wastes[1,2]. As shown in Fig.1, in the BARS, the spent fuel
from conventional LWRs is recycled as a MOX fuel for a BWR core with fast neutron
spectrum through the oxide dry-processing and the vibro-packing fuel fabrication.
Spent fuels from conventional BWRs
Fast spectrum BWR Vibro-packing fuel fabrication
•4
Fig. 1 Concept of BARS (BWR with Advanced Recycle System)
The fast neutron spectrum is obtained through triangular tight fuel lattice, which tends
to shift coolant void reactivity more positive, so that major concern in core design is
how to keep it negative under the condition of target core characteristics. Yokoyama et
al reported the specifications and characteristics of plutonium breeding option named
BBWR (Breeding BWR) of the BARS core with negative void reactivity[1,2].
In this paper, a neutronics design is presented that assures negative void reactivity
during whole burnup period for plutonium breeding option. Also, the void reactivity
coefficient is analyzed physically to show the neutronic features of the BARS core and
the advantage of the streaming channel. A comparison is performed between the
BARS core and the LMFBR core that has similar fast neutron spectrum.
Moreover, we have made an experimental study on thermal characteristics of the tight
fuel lattice. In the tight lattice core, the gap between the rods must be narrow, about
1.3 millimeter, for achieving high conversion ratio. The gap between the rods becomes
narrower, the critical power performance may be worse due to smaller flow area.
Because there are few critical power data for such tight lattice, the critical power
measurement test was performed in order to get the database. Also, Arai's critical
power correlation that will be used for BARS thermal core design was compared with
our critical power test data.
BARS Core Description
Table 1 shows typical BARS core and fuel specifications, which are almost the same
as those described in the previous papers. To achieve a fast neutron spectrum, a tight
lattice fuel assembly was adopted where a water to fuel volume (W/F) ratio is about
0.5. It is well known that the void reactivity coefficient in LWRs has the tendency to be
positive as the neutron spectrum harder. Then, a new core concept was introduced in
order to improve the void reactivity coefficient under the restriction of core diameter by
adopting a neutron streaming channel described below.
OQO©
Normal
Streaming
ruel Assembly
Channel
Axial Blanket
LJ Normal Fuel Assembly
0 Partial Fuel Assembly
+ Control Rod
Number of Pins: 658
02 *SÜE»«»**#
oooe
Table 2 Neutronics Analysis Method
Bumup calculation Criticality and
Item
Void reactivity calculation*
Cross section library JENDL-3.2 JENDL-3.2
Fuel lattice bumup code
Code Monte Carlo code ; MVP
(Transport theory)
Infinite fuel lattice Three-dimensional full core
Geometry model
(fully heteroqeoues)
Spectrum calc. 95 groups
Energy groups Continuous energy
Burnup calc. 3 groups
*)Fuel composition was given from burnup calculation at each burnup step.
Characteristics of BARS core
Table 3 shows burnup characteristics of the core loaded with MOX fuel accompanied
with no MAs and REs. The average void fraction is designed to be about 60%. The
operation cycle length is 1year and the number of refueling batch is four. The Pu
enrichments of load fuel were determined so that criticality is attained at the end of the
equilibrium cycle (EOEC). The result is shown in Table 4 and Fig.6. The enrichment of
outermost assemblies has been set higher than the other fuels aiming at radial power
flattening. The enrichment is varied in the axial direction : The enrichment located in
the lower region of fuel assemblies near the axial core center is about 20% to 30%
lower than those above and below the region. The breeding ratio is 1.04 with average
discharge bumup of 44GWd/t.
Table 3 Burnup Characteristics of BARS
Item unit BARS
Refueling scheme - 1 year - 4 batches
(1)Burn-up reactivity swing % k/k 1.6
(2)Breeding ratio 1.04
(3)Core average discharge fuel burn-up GWd/t 44
(4)Maximum linear heat rate W/cm about 400
{5)Heavy metal inventory t 132
- Load fuel : MOX without rare earths
- Pu fissile / Pu total = 0.575, U enrichment = 0.02
Table 4 Pu Enrichment Distribution
Low enrichment
(weight %) zone
Low Other 160 /
enrichment zones / *k \ 20
zone •V1
80 30
Inner 11 14 30
f
Assembly
Normal Partial (cm)
Outermost 11 17
Assembly Assembly
Assembly
Fig. 6 Axial Pu Enrichment Distribution
Table 5 shows the void reactivity. It is negative through the equilibrium cycle although
it has uncertainty due to Monte Carlo calculation. It can be seen that the void reactivity
oooe
tends to be positive as fuel burnup. The figure in parentheses indicates the void
reactivity for the core without axial enrichment distribution. The axial enrichment
distribution leads to more negative void reactivity. Therefore, it is one of the means to
ensure negative void reactivity at EOEC.
Table 5 Void reactivity coefficient of BARS core
Void reactivity coefficient*
Core condition
( k/k/%void)
at the initial core
-0.00049 (-0.00020)*
(no fuel bumup)
at the end of
-0.00001 (notcalc.)
the equilibrium cycle
*) Monte Carlo calculation with standard deviation of 0.00002 k/k / %void
**) In case of axially uniform enrichment distribution
Since there is no fuel in the upper half of the partial fuel assembly and the neutrons
tend to leak from the streaming channel, the neutron flux in the upper-half of the
normal assembly is lower than that in the lower-half of that. By placing a low
enrichment zone on lower half of fuel assembly, it is possible to flatten the axial
neutron flux distribution. Therefore, the neutron streaming effect by void fraction
change remarkably increases due to the increase of the neutron flux in the upper-half
of the normal assembly adjacent to the streaming channel. The axial enrichment
distribution is also advantageous from the viewpoint of power flattening. However, its
consistency with the vibro-packing fuel fabrication should be investigated.
The void reactivity was analyzed to show the neutronic features of the BARS core and
the advantage of the streaming channel. Table 6 shows the breakdown of the void
reactivity coefficient at the initial core (no fuel burnup) . In this table, the reactivity
coefficient is decomposed to two main components ; one is caused by change in
fission and capture reactions, and the other by change in neutron leakage. The former
component is a combination of positive contribution due to fission and negative one
due to capture (The sign is in case the reaction increases.) The leakage component
is negative and the absolute value is twice larger than that of the contribution due to
fission and capture reactions. This shows the effectiveness of the streaming channel.
There are two major isotopes in the fission and capture reactions. The reactivity
coefficient due to Pu-240 is 6.7 10"4 k/k/%void, which is mainly attributed to
decrease in capture reaction in low energy region(about 1eV). That due to U-238 is -
7.6 k/k/%void, which is mainly attributed to increase in capture reaction in relatively
high energy region(about 104eV-105eV). The contribution of Pu-239 is very small, but
this is caused by cancellation of changes of the reactions in the high and low energy
regions as shown below.
oooo
Table 6 Breakdown of Void Reactivity
0.20
Coefficient I I I I I I
Reaction Reactivity coef .* Discretized from Monte Carlo Calc.
(0.0001 k/k/%Void) co 0.15
I
LMFBR ,-.
A. Fission and Capture
U-238 -7.6 BARS Core :-: "
CD
Pu-239 -2.0 Q. 0.10
S
Pu-240 6.7
Pu-241 -1.1 co
c
Pu-242 1.1 o 0.05
Stainless steel -1.1 'co
CO
Water 1.4
Others 1.0 0.00
Total -1.7 10"1 10° 101 102 103 104 105 106 107
B.Leakage -3.2
Neutron Energy (eV)
C. Total (=A+B) -4.9
i Initial core
Fig. 7 Energy Distribution of Fission Rate
Figure 7 shows energy dependence of the fission rate in the BARS core and the
typical LMFBR core. Both cores have the largest fission rate in the energy region
about 2MeV-3MeV. The second largest peak in the LMFBR is around 105eV, whereas
that in the BARS core is around 100eV and much lower than that in the LMFBR. This
is caused by the difference of neutron moderation capability between water and
sodium. If the void fraction increases in the BARS core, the reactivity from two peaks
almost cancels out each other. This is why the contribution of Pu-239 is small.
Therefore, the spectrum is not much the same as that of the LMFBR, though it is
harder than that of conventional BWR cores. Moreover, it was found that the tendency
of void reactivity to shift positive direction due to burnup is caused by accumulation of
FPs.
Benchmark tests on NCA
Verification of calculation method is required since the BARS core has unique
characteristics:its neutron spectrum is different from conventional BWRs and LMFBRs
and it utilizes neutron streaming effect that is difficult to evaluate accurately.
Therefore, we have a program to benchmark the calculational method. The NCA
(Toshiba Nuclear Critical Assembly) is a swimming-pool type critical assembly fuelled
with slightly enriched uranium, which has been utilized to verificate both LWR design
codes and the specific fuel design. So far, basic characteristics such as conversion
ratio, neutron spectrum,and power distribution were measured on the core with tight
lattice test zone in the NCA[4,5]. We are now planning to perform a new test including
the void reactivity coefficient and streaming effect important for the BARS core.
Thermal-hydraulic Test
In the BARS core, a tight lattice fuel must be adopted because the conversion ratio
needs to be improved to about 1. However, as a fuel rod gap becomes narrower,
thermal-hydraulic performance, especially critical power, becomes worse. Therefore,
ooo
Cool pump w / Preheater ©
__/~]_J @ Flow control valve
The axial power distribution is stepped-cosign type as shown in Fig. 10. The spacers to
keep the rod gap are honeycomb type as shown in Fig.11. Seven spacers were
positioned with axially 512-mm pitch. On this test bundle, the hydraulic equivalent
diameter was 5.1 mm, and the thermal equivalent diameter was 9.7 mm. The
hydraulic equivalent diameter was about 1/3 of that of BWR fuel, because of tight
lattice configuration. Boiling transition is detected by rod surface temperature increase.
The rod surface temperature is measured by thermocouple, which was set on heater
rod surface.
r -*'" T
1 >
O®®©®©©©©©©@@®®©@®©@©©@©
370B iHggtedl
.i*tSJt,
Fig. 10 Axial Power Distribution Fig. 11 Honeycomb-type Spacer for
Seven Rod Tight Lattice Test Bundle
Test procedure and condition
On critical power test, test conditions of pressure, inlet water temperature and flow
rate, were set to programmed level firstly. Then, the bundle power is raised step by
step every a small magnitude. Critical power is defined as a power when the rod
surface temperature jumps by 14 centigrade from the temperature under nucleate
boiling conditions. Test conditions were as follows.
Pressure: 4, 5, 7, 8 MPa
Mass flux: 500 ~ 2000 kg/m2s
Inlet subcool: 100 ~ 300 kJ/kg
Inlet subcooling was quite larger than present BWR conditions, because this condition
intended boiling length to be near to that of BARS core, because the active fuel length
of BARS core is about 1/2 times long as that of preset BWR core.
Critical power test result
First, the typical behavior of the rod surface temperature at a boiling transition
condition is shown in Fig.12. The test conditions of Fig.5 were pressure 7 MPa, inlet
subcooling 100 kJ/kg and flow rate 1400 kg/m2/s. When the boiling transition was
occurred, the rod surface temperature was different from DNB, and it didn't rise rapidly,
and slight fluctuation occurred in it. These behaviors are the same as those of the
present BWR fuel. Therefore, the mechanism of the boiling transition is considered as
a liquid film dryout as the present BWR fuel.
Figures 13,14, and 15 show the measured results of critical power toward the mass
flux, pressure and inlet subcooling respectively.
. . ' -\ .-I".
Sf 648 Subcooling 300kJ/kg
V
| 623 200 kJ/kg
1k 100 kJ/kg
a> 598 o
a.
6
548
5.0 10.0 15.0
Time [s] 1000 2000
Mass Flux [kg/m2 B]
Fig. 12 Behavior of Rod Surface Temperature Fig. 13 Measured Critical Power Result
at Boiling Transition Toward the Mass Flux
1000 1000
Subcooling: 200kJ/kg
Pressure 7 MPa
" 800
| 600
^ 400
• G=1200 kg/m2 fe
5 200 EG=1500kg/m2 5
X G=?nnn kn/m? £
0
100 200 300 400
200
Inlet Subcooling [kJ/kg]
3 4 5 6 7 8 9
Pressure [MPa]
Fig. 14 Measured Critical Power Results Fig. 15 Measured Critical Power Results
Toward Inlet Subcooling Toward Pressure
The critical power increased as mass flux and inlet subcooling increase. The critical
power decreased as the pressure increase. These critical power tendencies of tight
lattice rod bundle toward flow condition are the same as present BWR fuel. However,
the critical power at the boiling transition condition was 60% lower than that of the
present BWR fuel due to narrower gap.
Verification of critical power correlation for tight lattice
BARS operating conditions are almost the same as those of BWR except for the tight
lattice configuration. Those measured critical power data mentioned in the previous
section (called BEST data after that) is that of the tight lattice bundle which has been
hardly seen under BWR operating condition. Therefore, most of critical power
correlation can not be applied to the tight lattice due to outside of application range.
The critical power correlation proposed by Arai et al. was developed based on the
tight lattice rod bundle data.[6] Those data were tested by LeTouneau et al.[7] The
tested bundles were triangular tight lattice rod bundle. However, minimum rod gap
was 1.5 mm. The verification of the Arai's correlation was needed for the application to
the BARS core. Arai's correlation is as shown below.