No nuclear weapon has ever been made using U-233, because of inevitable U-232 contamination. [Correction: one tiny experimental bomb, see comments.] Separating out U-232 is even more complex than U-235 enrichment or plutonium breeding.

Thorium absorbs a neutron, then decays to Protactinium-233, which sends out a bright gamma cascade, before it decays to Uranium-233 that would be used in the reactor. Pa-233 has a one-month half-life, so for about 3 months that fuel is gamma hot.

Gamma rays from Pa-233 and U-232 destroy electronics needed for any bomb, harm technicians, and are easily detected on land or by satellite, impossible to disguise.

“The uranium-233 produced from thorium-232 is necessarily accompanied by uranium-232,… [which] has a relatively short half-life of 73.6 years, burning itself out by producing decay products that include strong emitters of high-energy gamma radiation. The gamma emissions are easily detectable and highly destructive to ordnance components, circuitry and especially personnel. Uranium-232 is chemically identical to and essentially inseparable from uranium-233.” Hargraves, American Scientist Vol 98, July 2010

“Only a determined, well-funded effort on the scale of a national program could overcome the obstacles to illicit use of uranium-232/233 produced in a LFTR reactor. Such an effort would certainly find that it was less problematic to pursue the enrichment of natural uranium or the generation of plutonium. In a world where widespread adoption of LFTR technology undermines the entire, hugely expensive enterprise of uranium enrichment — the necessary first step on the way to plutonium production — bad actors could find their choices narrowing down to unusable uranium and unobtainable plutonium.” Hargraves, American Scientist Vol 98, July 2010

“In the context of proliferation resistance, … The local fuel processing of the breeder and burner configurations eliminates the possibility of diversion during transport. The fission-product-saturated fuel salt of the minimal fuel processing converter reactor is highly self-guarding during transportation. Further, the transport casks are massive because of the required amounts of shielding. In general, diversion of molten salt materials is difficult. The reactor operates as a sealed system with an integrated salt processing system that is technically difficult to modify once contaminated. The hot salt freezes at relatively high temperatures (450-500°C), so it requires heated removal systems. FS-MSRs operate with very low excess reactivity. Loss of a significant amount of fuel salt would change the core reactivity, which could be measured by a well-instrumented reactivity monitoring system. During operation (with the exception of deliberate fissile material removal for a breeder or addition for waste burner), the fissile materials always remain in the hot, radioactive salt. However, FS-MSRs, with integrated fuel separation, may be unsuitable for deployment in nonfuel-cycle states to minimize dispersal of separation technologies.” Fast Spectrum Molten Salt Reactor Options, Oak Ridge National Laboratory, July 2011

Less Uranium Shipped

Terrorists steal uranium, virtually always, during shipment.

With LWRs uranium is shipped several times after mining: for enriching, making into pellets and rods, delivery to the reactor, and long-term storage.

A 1GW LWR needs 35 tons of enriched uranium per year. Since LWR has a lifespan of 30-50 years, there would be 1050-1750 tons shipped.

With LFTRs, uranium is only used to start the reaction (uranium is produced within the reactor from thorium). Molten Salt Reactors generally have online processing, so any transuranic elements generated in the reactor would get circulated back to the reactor core to be fissioned. A 1 GW LFTR would use 1/4 ton mined uranium in 50 years (and about 50 tons of thorium).

Breakeven operation [make as much uranium as consumed] only requires approximately 800 kg of thorium per GW(e) year added simply as ThF4. Start-up fissile requirements can be as low as 200 kg/GW(e) D. LeBlanc / Nuclear Engineering and Design 240 (2010) 1644-1656

(A waste-burning fast-spectrum MSR would likely be located at the LWR waste storage facility or in the LWR steam containment building, so there would be no shipping uranium. A thorium-burning LFTR would require shipping thorium, which is barely radioactive, and is common worldwide.)

Exportable Technology

“The no-heavy-metal separation converter cycle FS-MSR reactor presents a distinctive capability for a highly proliferation-resistant resource-sustaining fast-spectrum reactor. The potential lack of fissile material separation technology within a converter cycle FS-MSR has the potential to enable a fast-spectrum reactor that is exportable to nonfuel-cycle states without requiring a fuel return. Because of the ability of a fast-spectrum reactor to tolerate the accumulation of significant amounts of fission products, the only fuel processing that appears necessary for many years of FS-MSR converter cycle operation is capture of the fission gases (possibly extracted via helium sparging) and mechanical filtering of the noble metal fission products particles as they accumulate in the fuel salt.” Fast Spectrum Molten Salt Reactor Options, Oak Ridge National Laboratory

Don’t waste resources pushing for greater protection for nuclear reactors as sources of nuclear weapons material. They are easy to protect (internet-enabled sensors of many types, surveillance, etc.) and have chemicals that are extremely difficult for terrorists to work with. Those resources would be better used detecting the easy methods of making nuclear weapons materials!

Remember the way we made plutonium for the first bomb. The first production reactor that made plutonium-239 was the X-10 Graphite Reactor, the main production reactor was the Hanford B Reactor, both specialized for weapons-grade plutonium production. A “graphite pile reactor” is a very simple design, of un-enriched uranium and graphite — that’s what we should actually be watching for, in some cave near a uranium deposit. They’ll die while building it, but that might be okay for them.

Wikipedia Plutonium-239 says “reactor-bred plutonium will invariably contain a certain amount of Pu-240 due to the tendency of Pu-239 to absorb an additional neutron during production. Pu-240 has a high rate of spontaneous fission events (415,000 fission/s-kg), making it an undesirable contaminant. As a result, plutonium containing a significant fraction of Pu-240 is not well-suited to use in nuclear weapons; it emits neutron radiation, making handling more difficult, and its presence can lead to a “fizzle” in which a small explosion occurs, destroying the weapon but not causing fission of a significant fraction of the fuel… Moreover, Pu-239 and Pu-240 cannot be chemically distinguished, so expensive and difficult isotope separation would be necessary to separate them. Weapons-grade plutonium is defined as containing no more than 7% Pu-240; this is achieved by only exposing U-238 to neutron sources for short periods of time to minimize the Pu-240 produced.” [That “fizzle” is normally called “pre-detonation”, goes off by itself, oops you just blew up your own building. Separating U-235 from U-238 (3 weights apart) is difficult; Pu-239 from Pu240 (1 weight apart) is much harder.]

Both LWR and MSR are much more difficult sources of material for making a nuclear bomb or dirty bomb than just starting with natural (unenriched) uranium to make plutonium.