Annals of Nuclear Energy 43 (2012) 77–82
Contents lists available at SciVerse ScienceDirect
Annals of Nuclear Energy journal homepage: www.elsevier.com/locate/anucene
Neutronics analysis of Americium-based fuel for long-life core Masanori Nakamura ⇑, Masaki Saito, Hiroshi Sagara Research Laboratory for Nuclear Reactors, Tokyo Institute of Technology, 2-12-1, O-okayama, Meguro-ku, Tokyo 1528550, Japan
a r t i c l e
i n f o
Article history: Received 30 June 2011 Received in revised form 31 December 2011 Accepted 2 January 2012 Available online 30 January 2012 Keywords: Long-life core Trans-plutonium isotopes Americium 238 Pu Space reactor
a b s t r a c t Feasibility study on the ultra long-life lithium (Li) cooled fast reactor loaded with conventional LWRgrade Am has been performed. Americium-241 has a potential to reduce the initial excess reactivity because it has relatively high capture cross section to be effectively converted to 238Pu, a fissile nuclide in fast neutron spectrum. For the better neutron economy, 7Li enrichment in coolant, nitride fuel with 15N enrichment, and lead–bismuth (Pb–Bi) reflector were selected and parametrically analyzed to find the optimal condition of criticality achievement with Am-based fuel. In the case of single region homogeneous core with only Am nitride fuel, it was found the condition of criticality sustained more than 100 years operation though the core has a large gradient of flux level distribution. The flattening of geometrical neutron flux distribution were also studied by adjustment of the fuel composition of Am and fissile material in dual region core. With these mechanisms, the change of burn-up reactivity was within 3% and ultra long-life core with over 100-year-life and less than 1.5 radial peaking factor could be achieved simultaneously throughout the operation. Safety parameters such as Doppler and void coefficient are also improved by dual region core. This mechanism of ultra long-life core is expected to be applied to future nuclear reactor concepts such as a space nuclear reactor. Crown Copyright Ó 2012 Published by Elsevier Ltd. All rights reserved.
1. Introduction In order to support the human activities in space, stable power resources for decades of long period will be required in the future. Because of the low power density and very severe mass restriction by the ability of rocket performance, there are difficulties to satisfy these demands by solar energy and chemical battery such as fuel cell. Space nuclear reactor is one of the answers to solve these problems. Space nuclear reactor is very attractive in its utilization as a long-life stable electric power resource for human activities in the space in, for example, Moon and Mars. Many concepts have been suggested to satisfy these contradicted conditions, and most of the concepts were based on the highly enriched uranium fuel. Highly enriched uranium, 235U enriched over 20%, however, is defined as a ‘‘direct use nuclear material’’ for military purpose by IAEA (2001), there are many institutional restrictions to utilize such a fuel in non-weapon possessing countries. Minor Actinides (MAs) has a very attractive potential to enables long life core both thermal and fast neutron reactor (Nikitin et al., 1999; Saito, 2002) because MAs of spent fuel in LWR have large neutron capture cross section. Mainly there are three elements in MAs, Neptunium (Np), Americium (Am) and Curium (Cm). Neptunium-237 and 241Am have a comparable fission and capture cross section in fast neutron, the loading of these nuclides in the ⇑ Corresponding author. Tel.: +81 3 57343060; fax: +81 3 57342959. E-mail address:
[email protected] (M. Nakamura).
core does not affect the eigenvalue (k-eff) in unity, though Cm isotopes affects k-eff largely because of its large m value compared to the values in Np and Am isotopes. In the transmutation chain of 237Np and 241 Am by neutron capture shown in Fig. 1, these nuclides are mainly transmuted to 238Pu with different paths, in the case of 237 Np, it takes the path of 237Np(n, g) ? 238Np(beta, 2.1 d) ? 238Pu, and in the case of 241Am, it takes the path of 241Am(n, g) ? 242Am(beta, 16 h) ? 242Cm(alpha, 163 d) ? 238Pu. Because the half life of the compound nuclide in 237Np neutron capture reaction, 238Np, is 2.1 days shorter than that in 241Am neutron capture reaction, 242Am(beta, 16 h) and 242Cm(alpha, 163 d), the production rate of 238Pu, working as fissile nuclide in fast neutron, from 241Am is more gradual than that from 237Np. It is expected that the quantity of 241Am would be increased very much in the future by short half-life of 241Pu(beta, 14.3 y) included in Pu accumulating in the Pu fuel cycle, and 242mAm is a very attractive fissile nuclide for designing a small space reactor because it has very large fission cross section (Ronen and Leibson, 1998). Furthermore, no usage of highly enriched uranium and Pu can make flexibility of technology development in non-weapon state countries. The present paper focuses on the possibility of long-life core loading Am. As a coolant material, lithium (Li) is utilized in the present paper since it enables high efficiency of electricity conversion because of high boiling point and specific heat (Kambe and Uotani, 1997; Nikitin et al., 2003). Because gravity and air utilization in space is negligibly small, removal heat of convection and conductivity is not effective but radiation, so high operating temperature
0306-4549/$ - see front matter Crown Copyright Ó 2012 Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.anucene.2012.01.001
78
M. Nakamura et al. / Annals of Nuclear Energy 43 (2012) 77–82
Fig. 1. Transmutation chain of trans-uranium.
is desirable from the viewpoint of much removal heat not converted to electricity. Though carbide and nitride fuel combine high thermal conductivity comparable to metallic fuel and high melting point comparable to oxide fuel, Am nitride is selected in the present paper because Am is very stable in nitride form chemically. Isomeric ratio of 242gAm in total 241Am capture cross section is used as 0.84 in fast spectrum based on recent experimental results (Sagara et al., 2010). 2. Goal of core specification A small ultra long life core which enables launch by rocket and easily control in a remote area such as Moon and Mars is designed as a design goal. In the near future, core enables human activities in space has ultra long nuclear life time over 100 years as a first goal. It has a little burn-up reactivity change (under 10%) for reduction of load of equipment for control during burn-up as a second goal. It is decided to keep radial peaking factor (z = 35–40 cm) under 1.5 during burn-up using calculation result of JOYO MK-2 as a reference because it is desirable that radial neutron flux distribution is flat for control of burn-up reactivity by neutron absorber with control drum in outside of core is as a third goal. Finally, it is decided to keep negative Doppler coefficient to maintain minimum safety as a fourth goal. It is defined as Doppler coefficient that difference of reactivity between reference core temperature and increased up with 500 K and it is defines as void coefficient that difference between normal operating Li density and that changed to 0%. Goal of core specification is the followings: 1. 2. 3. 4.
Core life time is over 100 years. Burn-up reactivity change is under 10%. Radial peaking factor is under 1.5 during burn-up. Negative Doppler coefficient.
3. Calculation model Computer codes, SLAROM (Nakagawa and Tsuchibashi, 1984), JOINT and CITATION (Fowler et al., 1971) and cross section library, JFS-3-J-3.2R (Chiba et al., 2002), which was based on Japanese Evaluated Nuclear Data Library JENDL3.2 (Nakagawa et al., 1995), were used in the present calculation. The SLAROM input consisted only of the PREP block to obtain 70-group effective cross sections of each material region by homogeneous cell calculation. JOINT was used to convert 70-group effective cross section data sets from the SLAROM output to the CITATION input. The nuclear characteristics were investigated using a calculation of two-dimensional RZ diffusion theory with depletion chain by CITATION. In this calculation, each zone had uniform nuclide number densities, neutron flux, and neutron spectrum averaged effective cross-sections with
80 zones. Main transmutation chains shown in Fig. 1 were used in the present analysis. After each burn-up calculation, the average number densities were obtained in each zone. The geometry of the core was shown in Fig. 2 composed of a cylindrical core with 40 cm radius and 50 cm height and a reflector with 9 cm thickness radially and 15 cm thickness axially. Number densities of fuel and coolant were used with volume ratio obtained by MONJU. As Fig. 2 shows, nitride fuel was used as a fuel which has high melting point, thermal conductivity, heavy metal density, and smear density (85%), SS316 was used as a cladding in fast reactor and Li was used as a coolant which has high specific heat and melting point from the viewpoint of high conversion efficiency from heat to electricity. Furthermore, Beryllium (Be) reflector was used because it has a large scattering cross section and can be utilized to decrease neutron leakage in the small core. Calculation was done with mean temperature of inner core and outer core are 1273, 973 K, respectively. Cylinder for same mass has minimum S/V value (S and V are defined as surface and volume of core) as a function of core radius. S/V value is close to minimum, i.e. leakage of neutron is small compared to core volume, core radius = 40 cm, core height = 50 cm is chosen as a reference model. It is a two dimensional cylinder core of an equivalent size with JOYO MK-3.
4. Results 4.1. Parametric survey for criticality of single region core with Am nitride fuel A Li-cooled small reactor core with only LWR-grade Am nitride fuel is studied parametrically. Atomic density ratio of Am was used as shown in Table 1. For the better neutron economy, 7Li enrichment in coolant, 15N enrichment in nitride fuel and a material for the reflector are considered as parameters shown in Table 2. Single region core means that fuel Am nitride fuel is loaded in inner core and outer core homogeneously. Thermal power of the whole reference core is designed as 5MWt and specific power is 4.3 MWt/t. Mass loading, decay heat and radioactivity of initial Am nitride fuel in total core are estimated as about 1200 kg, 0.1 MWt and 4.4 MCi (3.7 Ci/g). In case A, natural Li as a coolant, Be as a reflector. Am nitride as a fuel with natural nitrogen isotope ratio is used. As a result of burn-up reactivity change in the case A, it takes subcritical during burn-up as shown in Fig. 3 with bold line. In the case B, pure 7Li is used to reduce an excess neutron capture of 6Li(n, t), and a result is shown in Fig. 3 with dashed line. In the case C, Be of reflector is replaced to Pb–Bi eutectic (Pb–Bi, Pb:Bi = 45:55) for hardening the neutron spectrum beside the reflector function and enhancing fission of Am and a result is shown in Fig. 3 with dotted line. Though the neutron reflection function of Be is much larger than that of Pb–Bi, burn-up reactivity
79
M. Nakamura et al. / Annals of Nuclear Energy 43 (2012) 77–82
Height[cm] 80.0
15~30cm Axial Reflector
75.0
Inner Core
Zone Number
70.0 65.0 61.0 57.1 51.4 45.7 40.0
Radial Reflector
1
B4C
34.3 28.6
9~29cm
22.9 19.0
2
15.0 10.0
Outer Core
Axial Reflector
5.0 5.7
11.4
17.1
22.8
28.5
34.2
37.0
40.0
44.5
49.0
50.0 Radius[cm]
Fig. 2. Core layout.
in the case B is slightly larger than that in the case A because contribution of fission in the outer core to reactivity increases much. In the case D, pure 15N is used in Am nitride fuel to reduce an excess neutron capture of 14N(n, g) and a result is shown in Fig. 3 with chain line. In the case D, it is confirmed that criticality sustained over 100 years, and the increase of reactivity during burn-up is caused by fission of 238Pu transmuted from 241Am. For further analysis, 100% enriched 7Li in coolant, 100% enriched 15N in nitride fuel and Pb–Bi as a reflector material is fixed.
Table 1 Pu and MA composition in spent nuclear fuel from PWR (at.%).a
a
(a)Pu 238 Pu 239 Pu 240 Pu 241 Pu 242 Pu
1.9 56.8 22.9 12.0 6.4
(b)MA 237 Np 241 Am 242m Am 243 Am 243 Cm 244 Cm 245 Cm 246 Cm
47.53 35.52 0.07 12.86 0.03 3.74 0.22 0.02
4.2. Burn-up characteristic survey changing reflector thickness
3.5%EU 40 GWD/t 5 years cooling.
Table 2 Calculation parameters in single region core with Am fuel.
a
Case
Coolant
Reflector
Fuela
A B C D
Li-nat Li7 Li7 Li7
Be Be Pb–Bi Pb–Bi
N-nat N-nat N-nat N15
Fuel type: Am nitride.
Parametric study of Pb–Bi reflector thickness was performed in single region core with Am nitride fuel, thickness parameter are shown in Table 3. Axial reflector thickness is fixed and radial reflector thickness is increased in cases D–F. And radial reflector thickness is fixed and axial reflector thickness is increased in cases F–I. Results of burn-up reactivity change during irradiation with parameters of reflector thickness is shown in Fig. 4. Initial reactivity increases slightly within 1% by increase of reflector thickness in spite of dramatically reduction of neutron leakage from 19% to 11% in total neutron loss. This is because it is compensated a reduction of neutron leakage and increase of neutron capture of Am. On the other hand, the core life time increases because more conversion from 241Am to 238Pu is enhanced in the outer core due to increase of both the number of reflected neutron and softening of neutrons spectrum by increase of reflector thickness. In the case I, comparing to case D, initial reactivity increases only 1% but core life time increases from 100 years to 170 years drastically. It is showed in Fig. 5(a) and (b) that initial production of neutron in the inner core (represented by zone 1 in Fig. 2 and the outer core (represented by zone 2 in Fig. 2), respectively. As shown in Fig. 5(a) and (b), zone 1of the inner core is hardly affected by the thickness
Table 3 Pb–Bi reflector thickness as a parameter (cm).
Fig. 3. Burn-up reactivity change during irradiation changing parameters.
Case
Radial
Axial
D E F G H I
9 19 29 29 29 29
15 15 15 20 25 30
80
M. Nakamura et al. / Annals of Nuclear Energy 43 (2012) 77–82 Table 4 Calculation parameters in dual region core with Am based fuel. Case
Inner core/outer core
I J K
Am/Am Am/Pu0.3Am0.7 U0.3Am0.7/Pu0.3Am0.7
4.3. Americium based fuel for the ultra long life core with flattened flux distribution
Fig. 4. Burn-up reactivity change during irradiation changing reflector thickness.
In Sections 4.1 and 4.2, single homogeneous core model was used to felicitate the analysis and evaluate the sensitivity of each parameter. In general, single homogeneous core lead a large gradient of power distribution geometrically and it affect the fissile nuclide conversion and also safety parameters. The current section focuses on the geometrical flattening of flux distribution and also analyzes the effect of it on burn-up reactivity. Dual region core model is considered in this analysis, core is divided to two region, the inner and outer core, shown in Fig. 2 with the same volume. Fuel compositions in the inner/outer core are taken as parameters as shown in Table 4 and cases I–K were compared. As shown in Fig. 6(a), a part of Am nitride fuel of the outer core is replaced by Am–Pu nitride fuel to increase fission reaction in the outer core. As shown in Fig. 6(b), a part of Am nitride fuel of the inner core is replaced by Am-natural U(NU) nitride fuel to decrease neutron flux level in the inner core. Solid line represents initial flux ratio level with the maximum to the minimum value
(a) Zone 1
(b) Zone 2 Fig. 5. (mRf Ra)/ in Am fuel, (a) zone 1, and (b) zone 2.
of the reflector because the distance from axial/radial reflector is enough absorb on the way to the zone. The contribution to reactivity is almost proportional to number density of Am in zone 1. The contribution of 241Am to reactivity in the outer core (zone 2) is very different from that in any case. In the case D, the contribution of 241 Am to reactivity is slightly positive, though it is negative in case F, and it is very negative in case I. This is the effect of softening neutron spectrum because of increase of reflector thickness mentioned previously in this paragraph. Americium-243 works negative to the reactivity in every case but negative contribution to reactivity is relatively larger than that of 241Am in cases F and I because fission cross section of 243Am is smaller than that of 241 Am in fast spectrum. Micro fission cross section of 242mAm is higher than that of 241Am and 243Am but contribution to reactivity is little because number density is very small.
Fig. 6. Controllability parameters doping Pu and U, (a) doping Pu to outer core (inner core fuel: AmN, outer core fuel: (Pux, Am1x)N), and (b) doping U to inner core (inner core fuel: (Uy, Am1y)N, outer core fuel: (Pu0.3Am0.7)N).
81
M. Nakamura et al. / Annals of Nuclear Energy 43 (2012) 77–82
Fig. 7. Geometrical flux distribution of the core, (a) case I at the Beginning Of Irradiation (BOI), (b) case J at BOI, (c) case K at BOI, (d) case I in 100 years, (e) case J in 100 years, and (f) case K in 100 years.
in entire the zones and dashed line represents initial k-eff. With a respect to the doping ratio of Pu into the outer core, neutron flux level of the outer core relatively increases. When Pu composition in Am–Pu nitride fuel is increased to 30 atom%, initial flux ratio decreases from 2.7 to 1.5. On the other hand, initial k-eff increases to 1.17 (case J) because of too much loading of fissile material. As an alternative approach, NU composition in Am–NU nitride fuel is increased in the inner core for to neutron flux flattening. Because neutron flux of the inner core decreases relatively, initial flux ratio is under 1.5 when NU composition in Am–NU nitride fuel is increased to 30%, the same condition of the case K. In this regards neutron flux distribution as flatten to satisfy the design goal in the present study and. initial k-eff is about 1.03 in the case K. Initial flux distribution of core is showed with Fig. 7(a)–(c). As shown in Fig. 7(a), it has a large gradient flax distribution which has a peak in the center. In Fig. 7(b), flux of inner core is kept low level and flattened because of Pu of outer core. In Fig. 7(c) compared to Fig. 7(b), flux of inner core is more kept low level because loading Am is decreased and natural U is loaded to inner core. Flux distribution of core in 100 years is showed with Fig. 7(d)–(f). As shown in Fig. 7(d), flux distribution of the core in 100 years is similar to the initial one because core neutron spectrum is almost same in every zone and it transmuted from 241Am to 238Pu for proportion to neutron flux. In Fig. 7(e) and (f), effect of the outer core is dominant during burn-up so flux distribution in 100 years is kept flat and the design goal of core specification is satisfied because radial peaking factor (z = 35–40 cm) of most flat Fig. 7(f) during burn-up is kept under 1.5. It is defined as Doppler coefficient that difference of reactivity between reference core temperature and increased up
with 500 K and it is defines as void coefficient that difference between normal operating Li density and that changed to 0%, its value is shown in Table 5. In case I, Doppler coefficient is very low negative because it has little effect of resonance region because of hard spectrum and only Am fuel. In case J, Doppler coefficient becomes better with 1 digit because Am of outer core is replaced to Pu and flux of inner core decreases. In case K, it becomes better compared to case J because it has 238U which has a large resonance region in inner core. Void coefficient in case J is kept low compared to case I because flux level of inner core is kept low and flux distribution is flattened. Flux distribution in case K is similar to one of case J, but it becomes worse a little because a fission cross section of 238U in fast energy increases very much. Burn-up characteristic is showed with Fig. 8. Every case satisfies first condition of goal. Core life time is over 100 years. However, case I does not satisfy third design goal because flux distribution is large. Case J does not satisfy second design goal because it has a large burn-up reactivity change. Case K satisfies every design goal and the maximum burn-up is 147 GWd/t. During 100 years reactor operation period, totally 20% of 238Pu transmuted from 241Am (n, g) reaction is lost by itself decay (87.7 y) to 234U. Table 5 Initial safety parameters. Case
Doppler coefficient (Dk/kk0 /K]
Void coefficient (Dk/kk0 )
I J K
1.96E09 1.19E08 2.42E08
9.22E02 7.19E02 8.08E02
82
M. Nakamura et al. / Annals of Nuclear Energy 43 (2012) 77–82
flux level distribution. The flattening of geometrical neutron flux distribution were also studied by adjustment of the fuel composition of Am and fissile material in dual region core. With these mechanisms, the change of burn-up reactivity was within 3% and ultra long-life core with over 100-year-life and less than 1.5 radial peaking factor could be achieved simultaneously throughout the operation. Safety parameters such as Doppler and void coefficient are also improved by dual region core. This mechanism of ultra long-life core is expected to be applied to future nuclear reactor concepts such as a space nuclear reactor. References
Fig. 8. Burn-up reactivity change during irradiation changing flux distribution.
5. Conclusion Feasibility study on the ultra long-life Li cooled fast reactor loaded with conventional LWR-grade Am has been performed. Americium-241 has a potential to reduce the initial excess reactivity because it has relatively high capture cross section to be effectively converted to 238Pu, a fissile nuclide in fast neutron spectrum. For the better neutron economy, 7Li enrichment in coolant, nitride fuel and Pb–Bi reflector were selected and carefully analyzed to find the optimal condition of criticality achievement with Am-based fuel. In the case of homogeneous core with only Am nitride fuel, it was found the condition of criticality sustained more than 100 years operation though the core has a large gradient of
Chiba, G., Hazama, T., Ishikawa, M., 2002. The revision of nuclear constant set for fast reactor JFS-3-J3.2. J. Nuc. Sci. Technol. 1 (4), 335–340. Fowler, T.B., Vondy, D.R., Cunningham, G.W., 1971. Nuclear Reactor Core Analysis Code: CITATION. Oak Ridge National Laboratory report, ORNL-TM-2496, Rev2, USA. International Atomic Energy Agency, 2001. IAEA safeguards glossary 2001 edition. Kambe, M., Uotani, M., 1997. Innovative fast breeder reactor concept ‘RAPID’ for improvement of rector performance and proliferation resistance. Nucl. Eng. Design 170, 9–19. Nakagawa, M., Tsuchibashi, K., 1984. SLAROM: A Code for Cell Homogenizations Calculation of Fast Reactor, JAERI 1294. Japan Atomic Energy Research Institute. Nakagawa, T. et al., 1995. Japanese evaluated nuclear data library version revision2:JENDL-3.2. J. Nucl. Sci. Technol. 32 (12), 1259. Nikitin, K. et al., 1999. Advanced U–Np–Pu fuel to achieve long-life core in heavy water reactor. Ann. Nucl. Energy 26, 1319–1329. Nikitin, K., Saito, M., Artisyuk, V.V., et al., 2003. The application of U–Np fuel and 6Li burnable poison for space reactors. Nucl. Technol. 144, 152–157. Ronen, Y., Leibson, M., 1998. Potential applications of 242mAm as a nuclear fuel. Nucl. Sci. Eng. 99, 278–284. Sagara, H. et al., 2010. Numerical analysis of irradiated Am samples in experimental fast reactor JOYO. In: IOP Conference Series Materials Science, vol. 9. p. 01006. Saito, M., 2002. Multi-component self-consistent nuclear energy system for sustainable growth. Progr. Nucl. Energy 40, 365–374.