«E.A. Hoffman, T.A. Taiwo, W.S. Yang and M. Fatone Reactor Analysis and Engineering Division Argonne National Laboratory, Argonne, IL 60439, USA ...»
A PARTICLE-BED GAS-COOLED FAST REACTOR CORE DESIGN FOR WASTE
E.A. Hoffman, T.A. Taiwo, W.S. Yang and M. Fatone
Reactor Analysis and Engineering Division
Argonne National Laboratory, Argonne, IL 60439, USA
The issue of waste minimisation in advanced reactor systems has been investigated using the
Particle-Bed Gas-cooled Fast Reactor (PB-GCFR) design being developed and funded under the U.S.
Department of Energy Nuclear Energy Research Initiative (USDOE NERI) Programme. Results indicate that for the given core power density and constraint on the maximum TRU enrichment allowable, the lowest amount of radiotoxic transuranics to be processed and hence sent to the repository is obtained for long-life core designs. Calculations were additionally done to investigate long-life core designs using LWR spent fuel TRU and recycle TRU, and different feed, matrix and reflector materials. The recycled TRU and LWR spent TRU fuels give similar core behaviours, because of the fast spectrum environment which does not significantly degrade the TRU composition.
Using light elements as reflector material was found to be unattractive because of power peaking problems and large reactivity swings. The application of a lead reflector gave the longest cycle length and lowest TRU processing requirement. Materials compatibility and performance issues require additional investigation.
1. Introduction A solution must be found for the LWR spent nuclear fuel (SNF) that is accumulating in spent-fuel ponds and dry-storage sites, if nuclear is to play a prominent role in future energy generation in the U.S. It is currently planned that the spent fuel would ultimately be stored in a waste repository (Yucca Mountain in the U.S. case). The capacity of this repository has been set by law and there are requirements to determine the need for an additional repository by 2007. Clearly, if nuclear maintains its share of the electric generation capacity or continues to grow, the limit set for Yucca Mountain will be exceeded. This would necessitate additional repositories, unless approaches can be found to increase the effective repository capacity. Given the current difficulties encountered in setting up the first repository, the waste to be sent to the repository needs to be minimised. An approach for this is to recycle and incinerate the radiotoxic transuranics (TRU) in nuclear power reactor systems. Nuclear waste minimisation is one of the key attributes that have been defined for Generation IV reactors. In addition to waste minimisation, the burning of the legacy TRU in advanced systems should contribute to increasing the repository load capacity.
The issue of nuclear waste minimisation is being investigated under a USDOE-sponsored NERI project on the Particle-Bed Gas-cooled Fast Reactor (PB-GCFR) design that is evaluating the impact of different fuel forms and types on core performance. In addition, the other Generation-IV-systems goals of improved safety, enhanced proliferation resistance, and reduced system cost are being pursued in the study. The key requirement for the project is the application of a fast neutron spectrum environment to enhance both the passive safety and transmutation characteristics of advanced pebble-bed reactor designs.
A healthy future nuclear enterprise in a sustainable environment requires that fuel material be efficiently utilised. In this regard, the PB-GCFR study has focused on reactor designs with a high TRU conversion ratio. The intent is to design a TRU self-sustaining core in which for a given initial core load, the continuous recycle and application of the fuel would be obtained without need for external TRU make-up material. As a basis for trade-off studies, a reference compact fast-spectrum core based on the pebble-bed system has been developed. This core is designed for a power rating of 300 MWth (about 50 W/cc) and has no blanket zone, to reduce the production of high-purity fissile material. It was additionally imposed on this design that the TRU component of SNF extracted by a separation process be used as fuel, in order to minimise the TRU sent to the repository. This fuel type would also enhance proliferation resistance because the TRU would be unavailable in a repository that could become a plutonium mine.
The relatively low power and power density of the PB-GFR design are due to the requirements that the system be passively safe and be optionally exportable to international markets having little or no nuclear infrastructure. On this basis, the TRU content of the heavy-metal in the fuel has been limited to 20%, in support of non-proliferation goals of the U.S. government. Additionally, intrinsic proliferation protection is provided by the pebble fuel form. Because the system is for foreign export, it is currently planned that the core would be loaded off-shore and following long-life irradiation (15 to 30 years), the core would be removed and discharged off-shore. If fuel reprocessing is required, this too would be done off-shore.
When TRU is recycled, reprocessing losses only would be sent to the repository. In this case, the amount of TRU lost to the repository is determined by the amount of TRU to be processed and the recovery factor in recycle. Therefore, in order to minimise the TRU sent to the repository, it is desirable to design a reactor fuelled with TRU in such a way that the total amount of TRU fuel required to operate the reactor for its lifetime is minimised. This implies that the initial TRU inventory 964 and TRU mass flow for the core lifetime should be minimised. Various trade-studies have been performed to investigate the impact of typical fuel management schemes on fuel waste minimisation.
Based on the finding of this work, additional studies have been performed to optimise the long-life core, by investigating the impact of fuel materials on the core physics performance.
In Section 2, the analysis approach used for the study is discussed. The reference core design is presented in Section 3. This design is for a long-life core, which intuitively should result in the minimum amount of TRU being sent to the repository, for a given core size. The results of the sensitivity study on various fuel management schemes are presented in Section 4. Typical schemes such as 3-batch schemes with various cycle lengths (1, 3, 10 years), and semi-continuous schemes were evaluated. In Section 5, the additional optimisation studies employing different fuel, matrix and reflector materials are reported. The conclusions from the work are presented in Section 6.
2. Analysis method
Full-core equilibrium cycle calculations have been performed using the REBUS-3 fuel cycle analysis code.  Region-dependent 33-group cross-sections were generated with the MC2-2 code  based on ENDF/B-V nuclear data. Beginning of cycle material compositions and temperatures were used in the MC2-2 calculations. An R-Z computational model with homogenised pebbles has been developed for the PB-GCFR core. The flux distributions were obtained using the finite-difference diffusion theory option of the DIF3D code,  based on the observation that it gives similar results as Sn transport calculations for the homogenised core model.
From physics considerations, it is expected that such simple homogenisation models should be able to adequately treat the double heterogeneity effect of fuel particles in pebbles, in the fast system that is of interest in this study. This is because the resonance region’s (4 to 9 000 eV) contribution to the multiplication factor is small in such systems. Preliminary evaluations of this effect, with higher fidelity (deterministic transport and Monte Carlo) codes confirm this trend. However it is additionally planned to perform a more rigorous investigation of this issue in the future using an MCNP model as reference.
3. Reference design
A reference compact fast-spectrum core based on the pebble-bed system has been developed.
This core is designed for a power rating of ~300 MWth (about 50 W/cc) and uses the pebble fuel type.
Each pebble consists of a spherical container (pebble) made of matrix material that contains a central zone of coated or particulate fuel dispersed in the matrix material. Since the graphite used as matrix material in the Pebble-Bed Modular Reactor (PBMR) design might not be appropriate for the fast concept, because of its strong neutron moderating properties, alternative matrix materials were used in this study. The requirements for high operating temperatures that give a high thermal efficiency (45-50%) necessitated the use of refractory metals and high melting point materials having favourable fast neutron properties (e.g., low absorption cross-section).
The pebble employed in this study has an outer radius of 6 cm. A dispersion fuel zone diameter of
5.5 cm was used in the study (compare to 5 cm for PBMR), in order to increase the fuel loading in the pebble. The pebbles are packed into a cylindrical core and are assumed to occupy 61% of the volume.
Helium gas flows through the pebble-bed to remove the heat generated by the fission process. The helium coolant, being neutronically benign (i.e., low macroscopic absorption cross-section), does not adversely affect the fraction of neutrons available for converting fertile nuclides to fissile nuclides.
965 Thus, reactivity losses can be effectively compensated by the inclusion of fertile material, which tends to reduce the enrichment and excess reactivity requirements of long-life systems. For the reference design, depleted uranium is employed as the fertile material.
The fuel pebble design used for this study utilises the same matrix and coating material (or uncoated). This design permits a core fuel volume fraction of up to ~30% which ensures a sustained critical mode operation for a 15-30 years fuel irradiation cycle. Currently we have used potentially compatible fuel and matrix forms for this fuel type, mixed uranium-transuranics carbide [(U,TRU)C] fuel in a zirconium carbide (ZrC) matrix. The core volume fractions are 30.5% fuel, 30.5% matrix, and 39% helium (He) coolant. A 50 cm thick SS-316 reflector (20% He coolant) is assumed. Additional fuels material study is however required to ensure that these are feasible fuel-matrix forms in the irradiation and temperature fields of the PB-GCFR.
4. Impact of fuel management scheme on TRU losses
The Generation IV goals of waste minimisation and effective fuel utilisation (sustainability) can be met by designing a TRU breakeven core. For this core, the amount of TRU fissioned is equal to the amount created from uranium conversion. In this case, the TRU charged per cycle is equivalent to the TRU discharged. Based on the constraint of a breakeven core, the total amount of TRU required to operate a reactor for its lifetime was estimated for various fuel management schemes, in order to determine the favourable approach for waste minimisation. A thirty-year core lifetime was used as basis for comparison of the various cases. This implies that the core would be initially loaded, periodically reloaded (depending on the fuel management scheme), and ultimately offloaded at the end of the thirty years. Considerations were therefore given to all the fuel separations and losses resulting from this assumed scenario. The material from the final core is processed and used as fuel for a newer core in the sustainable nuclear enterprise. Because of the fast spectrum and the TRU breakevendesign, the discharge isotopic vector would be similar to the charge vector, and the irradiated fuel should be readily applicable to the new core. If necessary, the discharge fuel will be blended with the legacy LWR transuranics.
The lifetime TRU processed is the figure of merit in this study. It is assumed that the (U,TRU)C fuel can be separated to (1) obtain high purity uranium, (2) recover the TRU, and (3) partition the fission products. The uranium would be stored for future use (preferred) or buried in a low-level disposal site. Fission products and TRU losses from the separation stage would be packaged into waste forms and buried at the repository after cooling. The amount of TRU ultimately reaching the repository has an effect on the amount of nuclear waste that can be contained in the repository, as this material constitutes the primary long-term hazard. Waste minimisation and the ability to effectively package the material in the repository is dependent on the lifetime TRU processed. With multi-batch management, additional partial reloads of transuranics are required to start-up the initial core, while later batches utilise the reprocessed TRU. In this regard, it is assumed that there is an interval of five years between fuel discharge and recycle back into the core following fuel separation and fabrication.
The fuel management schemes considered in this study include a single long-life cycle of 30 years, 3-batch schemes with cycle lengths of 1, 3 and 10 years, and semi-continuous schemes (½and 1-year cycles representing continuous loading of fuel). The results for these cases are summarised in Table 1.
In order to limit the solution search space for this study, the core geometry was fixed with a height-to-diameter ratio of one. The fuel volume fraction and enrichment were searched to satisfy the requirements of the TRU breakeven design and the criticality condition during the cycle. The core
From Table 1, it is seen that the TRU discharge rate (in MT/GWth-yr) is dependent on the burnup. Maximising the burn-up reduces the TRU discharge rate and hence the required processing. For a fixed fuel residence time (30 years), the initial TRU inventory increases slightly with the cycle length and hence the TRU discharge burn-up decreases slightly. As a result, longer cycle length increases the TRU processing per cycle. However, because of less frequent recycling, the lifetime TRU processed decreases considerably with the cycle length.
For a fixed number of batches, the discharge burn-up can be increased by maximising the cycle length or the specific power under the material irradiation and safety constraints. As discussed above, in the PB-GCFR design, the passive safety requirements limit the specific power significantly.