Hybrid fission-fusion nuclear reactors

Transcription

1 Hybrid fission-fusion nuclear reactors Massimo Zucchetti 1,2 1 Massachusetts Institute of Technology, MIT, Cambridge (MA), USA 2 Politecnico di Torino, Italy zucchett@mit.edu Abstract A fusion-fission hybrid could contribute to all components of nuclear power fuel supply, electricity production, and waste management. The idea of the fusion-fission hybrid is many decades old. Several ideas, both new and revisited, have been investigated by hybrid proponents. These ideas appear to have attractive features, but they require various levels of advances in plasma science and fusion and nuclear technology. As a first step towards the development of hybrid reactors, fusion neutron sources can be considered as an option. Compact high-field tokamaks can be a candidate for being the neutron source in a fission-fusion hybrid, essentially due to their design characteristics, such as compact dimensions, high magnetic field, flexibility of operation. This study presents the development of a tokamak neutron source for a material testing facility using an Ignitor-based concept. The computed values show the potential of this neutron-rich device for fusion materials testing. Some full-power months of operation are sufficient to obtain relevant radiation damage values in terms of dpa. Index Terms: hybrid reactors, fission-fusion, neutron source, Ignitor 1. Introduction Largely in anticipation of a possible nuclear renaissance, there has been an enthusiastic renewal of interest in the fusion-fission hybrid concept, driven primarily by some members of the fusion community. A fusion-fission hybrid consists of a neutron-producing fusion core surrounded by a fission blanket. Hybrids are of interest because of their potential to address the main long-term sustainability issues related to nuclear power: fuel supply, energy production, and waste management. A fusion-fission hybrid could contribute to all components of nuclear power fuel supply, electricity production, and waste management. The idea of the fusion-fission hybrid is many decades old. Several ideas, both new and revisited, have been investigated by hybrid proponents. These ideas appear to have attractive features, but they require various levels of advances in plasma science and fusion and nuclear technology: One tokamak-based proposal combines ITER physics and technology (the leading magnetic fusion technology) with sodium-cooled fast burner reactor technology, plus the associated fuel reprocessing/refabrication technologies (the leading related burner reactor technologies). By building on the most advanced systems in both fusion and fission, this hybrid concept would require the least amount of advanced technology development. ITER is designed to achieve a duty factor of 25% for burn periods greater than 10 minutes, and to operate continuously for periods of 12 consecutive days. However, it is designed to operate only about 4% of the cumulative time over its 14 year DT operation period. This performance level is well below the 50-75% required availability for a hybrid system, so significant fusion technology reliability advances would still be required (as for any fusion concept), and the technology to integrate the two systems (such as dealing with a liquid metal in a magnetic field) would need to be developed. A reprocessing fuel cycle was proposed in which the actinides from LWR spent fuel were burned to greater than 90% in the hybrid. Any waste management strategy using either pure fission technology or fusion-fission hybrid technology will still require a long-term geological repository for the final remaining long-lived waste. Although technologically deployable long-term solutions for fuel and waste management may not be needed for half a century, there is a short-term political problem facing the nation. With work on Yucca Mountain halted, there is no perceived progress on addressing the waste management problem on any time scale. Concerning economics, there is general consensus that a hybrid capable of producing a certain amount of electric power would be noticeably more expensive than an LWR producing the same amount of power. Economic comparisons thus have to be made on an overall systems basis. For example, we must ask what is the overall cost of a group of LWRs plus necessary hybrids versus a combination of LWRs plus perhaps a larger number of fast reactors, with each system producing the same amount of power and reducing the waste to the same level. Advocates of hybrid reactors suggest that a fusionfission hybrid can be developed on a shorter time scale than for pure fusion electricity because the required plasma physics and some technology requirements are substantially reduced. Some of the panels and also the skeptics argued that some technology may be more complicated in a hybrid because of the integration of fusion and fission technologies. Perhaps more important, the pace of development will be dominated by engineering and technology and not by plasma physics. As far as proliferation is concerned, hybrids produce significant quantities of fissile materials, generally not retained in individually accountable fuel rods, and hence raise significant proliferation concerns. As a first step towards the development of hybrid reactors, fusion neutron sources can be considered as an option. The foreseen future expansion of nuclear power would involve a solution to burn the long half-life transuranics (TRU) in the spent nuclear fuel discharged from LWRs. Moreover, high-energy neutrons can be

2 very useful for the following processes, such as testing of candidate nuclear materials, production of radioisotopes (for medical applications and research), detection of specific elements or isotopes in complex environments, radiotherapy, and alteration of the electrical, optical, or mechanical properties of solids D-T fusion neutron sources sufficient to drive sub-critical advanced reactors can be an answer for these needs. A tokamak neutron source could be designed and built soon, extrapolating present designs of fusion tokamaks, paying attention to some additional R&D, such as emphasize quasi-steady state operation, disruption avoidance, component reliability, materials, etc. as well as selected tokamak physics and technology advances. The development of a radiation damage resistant structural material is a major challenge for both the core and the neutron source of advanced burner reactors. A sub-critical advanced burner reactor with a fusion neutron source (a fusion-fission hybrid ) will be more complex and expensive than a critical version of the same reactor. A principal advantage of a sub-critical reactor with a variable strength neutron source is that it can achieve deeper TRU fuel burnup and thus require significantly fewer complex and expensive fuel reprocessing/refabrication steps. Compact high-field tokamaks can be a candidate for being the neutron source in a fission-fusion hybrid, essentially due to their design characteristics, such as compact dimensions, high magnetic field, flexibility of operation. 2. Materials and Methods This study presents the development of a tokamak neutron source for a material testing facility using an Ignitor-based concept. Ignitor is a proposed compact high magnetic field tokamak, aimed at reaching ignition in DT plasmas and at studying them for periods of a few seconds. In order to act as a suitable neutron source for materials testing, Ignitor operating parameters have been revised, as discussed below, to achieve a longer plasma discharge length, which produces neutron fluences that are shown to be appropriate for studying fusion-relevant radiation damage to materials. We have assumed the neutron energy spectrum in the Ignitor first wall as reported in [1]. The total neutron flux on the first wall, computed per source neutron produced in the plasma, is 3, n/cm 2 s [1]. At maximum performance, with DT 50/50 discharges, the neutron production in Ignitor is 3, n/s (see figure 1). Figure 1: Neutron spectrum in the Ignitor first wall In a simple macroscopic model the number of displaced atoms depends on the total available energy Ea and the energy required to displace an atom from its lattice position E d E a DPA = κ (1) 2 E d The factor K is a normalisation constant (displacement efficiency) with the value 0.8. The value of the displacement energy E d is in practice chosen to represent empirical correlations and is in principle dependent on the chemical composition of the material. If the neutron flux and spectrum is known, and the material composition too, the radiation damage rate, measured in dpa/s (Displacement per Atom)/(second) may be evaluated with the formula: r N dv deφ(, E) ρσ i R, DPA, i( E) i= 1 DPA = N (2) ρi i= 1

3 Where the summation runs over all N isotopes in the material mix, is the DPA cross section for the isotope i, and the atomic densities. To computer material damage, a recent, multi-group dpa cross section data base has been obtained by the NEA Data Bank [2]. It is an ENDF/B-VII Damage Library, processed with NJOY in 211 energy groups, with a VITAMIN-J+ structure. Passing from the continuous energy to the multigroup structure, the above formula can be rewritten as: DPA rate (dpa/s) = kniφgσ g 2Ed Where: K = 0.8 (displacement efficiency) N i (atoms) is the concentration of isotope i ф g is the total neutron flux (n/cm 2 /s) in the group g σ g (ev*barn) is the damage cross section in group g E d (ev) is the displacement energy. This value is different for elements with different Z number. The values of the dpa have been obtained using the ACAB activation code [3]. 3. Results and discussion An initial evaluation has been made for a target of pure iron located in the Ignitor first wall. The dpa rate, (3) expressed in terms of displacements per atom per neutron produced in the plasma, is: D1 (Fe) = 3, dpa/n (4) In a full power year of operation, this translates into a yearly dpa rate of: D2 (Fe) = 33,84 dpa/y (5) These data are consistent with evaluations found in literature for Iron in other fusion devices, like IFMIF, ITER, DEMO, etc. [4]. The IFMIF neutron source and irradiation device [5] foresees a high-irradiation volume of about 0,5 liters, with a damage from 20 dpa/y up to 50 dpa/y for pure iron [6]. To obtain at least the minimum IFMIF irradiation performance, the Ignitor-based concept should produce approximately 6, neutrons per year. If the neutron source of 3, n/s is taken as a reference, this requires 1, s of yearly operation, that is, 7 months of full power operation per year, or a duty cycle of about 59%. However, if the Ignitor-based concept could increase the neutron production rate to n/s, a yearly damage rate of 20 dpa could be achieved with just 6, s of operation per calendar year, that is, less than 2,5 months per year, or a duty cycle of about 20%. The same evaluations done for pure Iron have been performed for some fusion-relevant materials, like ASI316L, EUROFER, SiC/SiC, Mo, Graphite, V-15Cr- 5Ti. Results are available in Figure 2. Figure 2: Radiation damage simulation in an Ignitor-like materials testing device Radiation damage (dpa/n) 4,00E-26 3,50E-26 3,00E-26 2,50E-26 dpa/n 2,00E-26 1,50E-26 1,00E-26 5,00E-27 0,00E+00 Iron AISI 316L EUROFER SiC/SiC Mo Graphite D1 (dpa/n) 3,22E-26 3,34E-26 3,54E-26 3,57E-26 1,94E-26 1,58E-26 Material

4 4. Shielding calculations One of the question arisen by such a high neutron production as that required for materials testing is the radiation damage to the machine magnets, and in particular to the insulators. In previous calculations, the radiation damage to the Ignitor components was evaluated. They referred to a quite old version of the machine, called IGNITOR-ULT, that is a bit different from the present version. In particular, the neutron flux on the external TFC (Toroidal Field Coils) computed for that version of the design was about n/cm 2 s (per neutron generated in the plasma), while the most recent results for the present Ignitor machine, as computed by Ansaldo, evaluate the same flux as n/cm 2 s (per neutron generated in the plasma). The results for Ignitor-ULT computed a dose on the TFC insulator of about 6.14 MGy (MegaGrays) for a production of n in the plasma, deriving from a total neutron fluence of n/cm 2 of the insulator. That dose was due to neutrons for about 60% of the total, and the rest was due to gamma rays. Those values are consistent, as far as the order of magnitude is concerned, with results obtained in literature, for instance those in [7]. In that investigation, a fluence of n/cm 2 caused an absorbed dose in G10CR (the material used for Ignitor TFC insulators) of 2.2 MGy. Then, if the fluence was equal to the one in Ignitor-ULT, we would have a dose of about 8 MGy, while we obtained as mentioned above 6.14 MGy. The relatively higher dose can be easily explained considering that in [7] the neutron energy was that of pure fast fission neutrons, around 2 MeV, where the highest dose conversion factor is found. Dealing now with the most recent version of Ignitor, we may then evaluate the total dose on the TFC insulator as follows: DTFC = 6.14 / * / MGy/(neutron generated in the plasma), and therefore: DTFC = MGy/(neutron generated in the plasma) If we now evaluate the dose on the TFC insulator necessary to produce 10 dpa of irradiation on a tested material, for instance Iron, we have that for such a radiation damage (10 dpa) about n are necessary. That irradiation would then produce a dose on the TFC insulator of: DTFC (10dpa) = 28,000 MGy (6) This value is quite beyond any possible radiation resistance of the insulator material. Then, some modifications, either to the design or to the choice of the material, or both, must be applied in order to overcome the problem. Shielding solutions that do not imply modifications to the machine assembly and design are obviously preferable and should be considered first. The shielding capability of the vessel could be improved by the addition of B, which has the well-known capability of absorbing neutrons. The technique of using borated shields is a usual one in the nuclear industry. Previous evaluations [8] found that adding 1 wt% of Boron to the vessel material (INCONEL-625), if B is totally enriched in B-10 (the isotope with the high neutron capture crosssection) would reduce the DTFC of about 10%. Better results are obtained combining composition adjustment to slight modifications of the vessel zone. If the thickness of the vessel is increased by 1 cm and 1% of B-10 is added to INCONEL-625, then the DTFC is reduced of 23%. About 60% of the DTFC is due to neutrons. Neutron absorption in the vessel could be increased by a moderator, which, softening the spectrum, would increase thermal neutron capture. This improvement would be particularly beneficial if B-10 is added to the INCONEL-625 composition. In our case, this could be done for instance - by adding a thin layer of graphite behind the first wall. Graphite tiles used to be the Ignitor first wall material, before being substituted by molybdenum. If we put, between the Mo first wall and the vessel a thin layer of 1 cm of graphite, then, including the other modifications specified above, the DTFC is reduced of 30% overall. Other similar solutions, with a soft impact on the design, can be envisaged and can give a first important contribute to the reduction of the dose on the insulator. Another strategy to minimize radiation damage to the TF insulators is to incorporate a radiation shield into the design of Ignitor between the first wall and the TF coils. Most presently proposed D-T tokamaks exploit their relatively large major radii (~6 meters) by incorporating either neutronabsorbing blankets or dedicated shields for radiation protection, relying on their several meter thickness to significantly attenuate the 14.1 MeV neutron flux before it reaches the TF coils. Ignitor, however, will not incorporate a blanket into its design and will not be able to provide the thickness required by traditional shielding materials, such as steel or tungsten carbide, due to the geometry constraints imposed by the relatively small major radius of 1.3 m. Although the redesign that will enable Ignitor to become a neutron source could make accommodations for a radiation shield, engineering constraints will limit the maximum shield thickness to no more than 20 cm at the midplane, eliminating traditional shield materials as candidates for the radiation shield. The identification of alternative but suitable shielding materials is driven by the two primary requirements of the radiation shield. First, the shield must attenuate neutrons in relatively small thickness, requiring high mass density to increase neutron interactions in the material as well as high hydrogen content to quickly moderate the neutrons. The incorporation of isotopes with a high neutron captures cross section, such as 6 Li or 10 B may enhance the neutron attenuation. Second, the shield must attenuate gamma rays in a relatively small thickness, requiring the incorporation of high-z elements into the shielding material since the photoelectric cross section goes approximately as Z 4.5. In addition, the shield materials should be commercially available and chemically inert to facilitate procurement and installation. Three materials that meet the above criterion have been identified. First, lithium hydride (LiH) is relatively stable, salt-like crystalline solid that possesses moderate density (820 kg/m 3 ), high hydrogen content (8.3x10 28 atoms/m 3 ) and 6 Li content (0.6x10 28 atoms/m 3 ). LiH is light weight, has a relatively high melting point of 692 C, and is commercially available in variety of forms. Although finely powdered LiH can react explosively with atmospheric moisture, compressed LiH used in radiation shields reacts much less violently, increasing weight through water absorption. It has been assumed that safe handling techniques can be employed to prevent the degradation of LiH as a shield material for Ignitor. Second, zirconium hydride (ZrH 2 ) is a metallic powder with a very high mass density (5600 kg/m 3 ), high hydrogen content (7.2x10 28 atoms/m 3 ), and zirconium content (3.6x10 28 atoms/m 3 ). ZrH 2 has been employed successfully as a

5 neutron moderator for TRIGA nuclear research reactors and Topaz-II nuclear space reactors. Third, zirconium borohydride (Zr(BH 4 ) 4 ) is another metallic powder with moderate density (1180 kg/m 3 ), high hydrogen content (7.5x10 28 atoms/m 3 ), and zirconium content (0.5x10 28 atoms/m 3 ), and 10B content (0.4x10 28 atoms/m 3 ). Shielding calculations with those materials are the further step required in future investigations. 5. Conclusions The computed values show the potential of this neutron-rich device for fusion materials testing. Some full-power months of operation are sufficient to obtain relevant radiation damage values in terms of dpa. The setup of a duty cycle for the device in order to obtain such operation times is the next required step to proceed with the evaluation. The estimate of the radiation damage on selected machine components has to been carried out too. Solutions to solve the problem of radiation damage to the insulator of the Toroidal Field Coils have to be explored, either with design and materials modifications, or with the adoption of shielding layers with advanced performance materials. 6. References [1] Neutron fluxes computed by Ansaldo have been provided by ENEA in the frame of the ENEA-Politecnico contract Studi inerenti la valutazione di impatto ambientale ed il rapporto di sicurezza di Ignitor, [2] ENDF/B-VII.0 data processed with NJOY for radiation damage calculations, NEA/OECD Data Bank, July [3] J. Sanz, O. Cabellos, N. García-Herranz. Inventory code for nuclear applications: User's Manual V. 2008, December (NEA-1839 ACAB-2008). [4] O. Cabellos et al., Effect of activation cross-section uncertainties in the assessment of primary damage for MFE/IFE structural materials, IFSA2007 Proceedings, Kobe (Japan), September [5] See web site: [6] V.Heinzel, et al., IFMIF High Flux Test cell-design and design validation, Fus. Eng. and Des. 82 (2007) [7] D. S. Tucker, F. W. Clinard, Jr, G. F. Hurley And J. D. Fowler, Properties of Polymers After Cryogenic Neutron Irradiation, Journal ff Nuclear Materials, 133&134 (1985) [8] S.Rollet, M.Zucchetti et al., Radiation damage calculations for Ignitor components, Journ. Nucl. Mater (1994)