Thorium breeder and burner fuel cycles in reduced-moderation LWRs compared to fast reactors

• We compare thorium (Th) fuel cycles in reduced-moderation LWRs (RMLWRs) and SFRs.
• RMLWRs have higher reprocessing and fuel fabrication requirements than SFRs.
• RMLWRs can achieve comparable radiotoxicity and repository decay heat to SFRs.
• RMLWRs draw on existing LWR knowledge base but unlike SFRs have never been built.
• Lower void coefficient from Th: essential for RMLWRs; improves SFR safety case.

In this paper, a reduced-moderation LWR (RMLWR) system is compared to a sodium-cooled fast reactor (SFR) as a reference fast system for sustained transuranic burning and self-sufficient break-even cycles using thorium (Th) as the fertile material. SFRs can achieve a higher discharge burn-up (and incineration rate – in the case of burners) than RMLWRs, which results in a highly significant reduction of ∼40% and ∼60% in fuel reprocessing and fabrication requirements for breeder and burner reactors respectively. With sufficiently reduced moderation, the repository radiotoxicity and decay heat of actinide process waste from RMLWRs is very similar to that from SFRs, assuming same fraction of process losses per reprocessed fuel inventory, for the breeder and burner cases. In all cases, fuel fabrication is challenging due to high energy gamma and spontaneous neutron sources. The gamma source is comparable in the burner cores but the spontaneous neutron source in the burner options is approximately an order of magnitude lower for the SFR after a large number of recycles. Th fuel is highly preferable to uranium fuel in RMLWRs due to a large improvement in the void coefficient, which must be kept negative. This constraint makes it difficult, if not impossible, to fuel an RMLWR burner or breeder with uranium. Similarly, the void coefficient in SFRs is significantly better with Th fuel, although it is probably not essential for this to be negative. Although the neutron spectrum and core geometry are substantially different, RMLWRs are a variant of conventional LWR technology. RMLWRs have never been built, whereas some SFRs have been built and operated. Despite this, SFRs are anticipated to have a higher capital cost, which appears to outweigh the lower reprocessing and fuel fabrication requirements, although this conclusion is sensitive to the reactor size, cost and availability factor assumptions.