Comparison of SFRs and LFRs as Waste Burners

In this paper, two 600 MWe reactors are compared regarding safety relevant reactivity coefficients, waste-burning capabilities and reactivity swings during burn-up. Furthermore, comparisons of unprotected Loss-of-Flow and Loss-of-Heat Sink calculations are presented. In the first part of this paper, oxide fuels with an inert 92Mo matrix (occupying a 50% volume fraction) are investigated. This CERMET fuel is considered to be significantly more stable than an MgO matrix-based CERCER fuel and the isotopic tailoring of molybdenum appears affordable. Manufacturing and reprocessing aspects of these new fuels are discussed in some detail. The LFR core under consideration is considerably larger than the SFR core and has a slightly higher average TRU enrichment (50.0 to 43.0%). The LFR core is larger due to the larger pitch-to-diameter of 1.5 compared to 1.2 for SFR. By introducing additional pins with BeO moderator both cores show a negative Doppler that is larger than the positive coolant reactivity coefficient (LFR: Doppler –50 pcm and 38 pcm coolant reactivity increase, SFR: –54 and 36 pcm); reactivity coefficients refer to a 100 K heat-up. The burn-up swings for the BeO moderated core were –12.8$ per year for the LFR and –23.8$ for the SFR. The LFR burner can annually transmute over 300 kg of plutonium and MAs corresponding roughly to the annual production of the transuranics of a 1.1 GWe LWR. Annual TRU consumption in the SFR burner is slightly less and equal to 263 kg. However, due to lower actinide inventory in the SFR, the actinide burn-up rate is larger than in the LFR. Another significant difference is in the safety behavior. The relatively large LFR overcomes the unprotected Loss-of-Flow due to its superior natural circulation whereas the SFR gets into sodium boiling. The latter may be avoided if fast negative structural feebacks could be proven to be large enough. In the unprotected Loss-of-Heat Sink accident the grace time of the LFR is considerably larger but for these longer times the lower grid expansion, which was not considered, would terminate the accidents. The use of thorium instead of the inert matrix appears to be quite attractive since 233U is generated without producing new minor actinides. Thus, potentially, a good new LWR fuel would become available. The reactivity coefficients of such a core look also quite good: for the LFR core and ThZrH1.6 the Doppler is -113 pcm and the coolant reactivity is 45 pcm. The burn-up swing is only about 0.4$ per year. The problem is the remote handling of the 233U that is due to the presence of a hard gamma emitter from a descendent of 232Th and also the 232U alpha emission. However, since remote handling is also used for MOX fuel, there should be no principal problem for reprocessing and fabricating a 233U / Th fuel. More studies are needed for the homogeneous burning of minor actinides in self-breeders, which in any case looks very promising. Finally, more detailed studies are required to further optimize core designs with respect to their fuel cycle performance, particularly for the SFR.

TUCEK Kamil;  CARLSSON Johan;  VIDOVIC Dragan;  WIDER Hartmut;  SOMERS Joseph;  GLATZ Jean-Paul; 

2007-02-16

American Nuclear Society

JRC33020

https://publications.jrc.ec.europa.eu/repository/handle/JRC33020,