Startup seeks to commercialize IFR

As Dan Yurman reports, several nuclear engineers from US national labs have turned into entrepreneurs, seeking to turn a theoretical reactor into commercial reality. Their shiny new startup, Advanced Reactor Concepts LLC, is betting it all on a "disruptive new technology": a refinement of the Integral Fast Reactor (IFR) pioneered at Argonne National Lab (ANL).

In short, their design is: the original IFR design and fuel cycle, shrunk down to a ~100 MW small modular reactor (SMR), with breeding gain reduced to ~1 (isobreeder), and the steam turbine replaced with a cutting-edge, supercritical CO2 gas turbine.

At a recent conference of the American Nuclear Society (ANS), they've released a white paper on their design (and a shorter brochure).

What I took from it:

Credit CRIEPI. Pyroprocessing experiment at CRIEPI in Japan. Here, an electrorefiner has successfully (?) separated uranium and plutonium from a molten salt bath; pictured is the cadmium metal cathode with U and Pu adsorbed.

• This is in short, a small, modular, sodium-cooled fast breeder reactor (SFR) of 50-100 MWe size
• They intend to close the fuel cycle with the original Argonne pyroprocessing system, developed for the IFR. This, as described briefly by the CEA, is a "dry" (non-aqueous) fuel reprocessing cycle, based on electrochemistry in molten salt solvents. Here's a slideshow about recent progress in it by Japanese researchers.
• To simplify fuel reprocessing, the reactor's fuel will be metallic (elemental -- 0 oxidation state) uranium and plutonium, like the original IFR design, unlike most reactors which use oxide fuels. This removes the oxide reduction step from reprocessing. The CEA published an accessible article on the tradeoffs between fuel types in SFRs.
• They plan for widespread, international deployment. To keep the reactor useless for weapons, the initially-loaded fuel is enriched to no higher than 17% U-235. Further, the electrochemical separation of pyroproccesing cannot be used to isolate weapons-grade material -- it can only separate actinides as a group, plutonium together with extremely hot transuranics.
• The fuel cycle is not only fully closed, but self-sustaining. The reactor is an isobreeder -- it converts enough U-238 to fissile fuel to burn through natural uranium entirely. (Ordinary LWRs use barely 1%.) This differs from the original IFR design, which has a large positive breeding ratio, creating more fissile material than it consumes. The breeding ratio is ~1.0.
• However, reprocessing is not integral to the power plant (as in the IFR design). Quite the opposite -- ARC has a very heavy fuel loading and high burnup (80-100 GWd/ton heavy metal), allowing for an extraordinary, 20-year refuelling interval! In fact, the white paper asserts there will be no fuel handling capability whatsoever at the power plant (which they intend to situate underground, like a missile silo). I guess the idea is a sort of proflieration-resistant "nuclear battery", a black box without user-servicable components.
• To my understanding (they don't clarify this), the reactor burns through all transurianic isotopes, leaving no long-lived TRU waste (well, excluding "< 0.1%" losses in the chemistry step). So this reactor produces no long-lived nuclear waste.
• They intend to commercialize a closed-cycle gas turbine with supercritical CO2 as working fluid -- currently only a research idea. They predict have a modest improvement in thermodynamic efficiency (electricity generated / reactor heat output), 38% at 510°C/950°F (coolant at reactor outlet), compared to typically 33% for operating light-water reactors (LWRs). Kirk Sorensen wrote about S-CO2 cycles, in the context of liquid fluoride thorium reactors (LFTRs). Apparently they are extremely compact, because supercritical fluids are much denser than gases.
• The founders are team of mostly nuclear engineers from US national labs. The CEO is an exception; Mr. Irfan Ali is an MBA with an MS in electrical engineering, and is also CEO of Lambda OpticalSystems, which does optical fiber networking (and apparently sells the world's only purely-optical switch). His bio is here.

ARC is hardly alone in trying to commercialize liquid-metal fast reactors. There's at least half a dozen such projects right now, some (unlike ARC) backed by giant, established nuclear suppliers. So far none of them are actually commercial, though they seem to be trying.

• General Electric and Hitachi are collaborating on the PRISM reactor, which is a small (300 MWe), modular, metal-fuelled sodium-cooled fast reactor -- just like ARC, it seems. According to this GE slideshow, the are also planning to commercialize the IFR pyroprocessing cycle. Also according to that source, this design has been in existence in various forms since either 1981 or 1985 (it hasn't been built yet).
• Toshiba is backing an extremely small (10 MWe), modular, metal-fuelled SFR (yet another one), the 4S. There is some hope one of them could be built in Alaska.
• Russian state nuclear company Rosatom and the En+ Group are backing a lead-bismuth-cooled modular reactor, SVBR-100, which is derived from the power plant of the formidable Alfa-class nuclear submarine. Sovietologist found more on this, including technical specs on the IAEA website and some renderings on a Russian-language site.
• Another small LFR is the SSTAR, this one nitride-fuelled. It is being sponsored by the US Deparment of Energy (and developed in their national labs), but has no commercial backing.
• Among these giants, a tiny startup called Hyperion is backing their horse, a nitride-fuelled, lead cooled small reactor. (Thanks to Idaho Samizdat for hosting these slides, delivered at ANS 2009).

For more on liquid-metal fast reactors, also see Steven Kirsch's page, the WNA historical overview, a couple of articles from the CEA, the international Gen IV alliance, the Idaho National Lab pages.