1. Sodium will explode on contact with water. I know that sodium and water don’t contact but they’re just a heat-exchanger away. I’m told a nitrogen Brayton cycle would be a good alternative to steam.

      1. What is more dangerous – a fire or a high pressure steam leak?

        What is harder to isolate or repair, a sodium leak or a primary coolant leak from a PWR operating at 2500 PSI and 600 F?

        Yes, systems have to be designed to ensure integrity and fluid separation. EBR-II did that with a steam generator with double walled tubes with helium — I think — between the walls.

      2. If having a heat exchanger with sodium on one side and water on the other proves to be too much of a problem, a possible solution might be an intermediate loop using a molten metal or salt that is non-reactive with water and sodium.

          1. It has some issues (esp. regarding thermal decomposition and materials compatibility) but everything consided it’s the best option I’ve seen besides e.g. stone (which may make a cheaper storage medium with solar salt as the heat transfer fluid).

    2. @JohnGalt

      Would similar units still be viable options in much of the Western United States that does have the dry conditions like Richland? Maybe, another unit in Arizona that does not need Phoenix sewer water. The tertiary loop (or pool) with another coolant seems more complex.

  1. Anyone have any recent news on the BN800? It went critical this summer … don’t know if they have raised it to full power. Also, the design is complete for the BN1200 (I think). Have they broken ground for the construction of that design?

    Russia claims the BN1200 will be price competitive with the VVER (PWR’s) 1200 and is the reactor they plan to deploy in large numbers.

  2. @Rod

    A good article to get the discussion going … So

    Why did you use “fast” in the title instead of “breeder”? Many of the proponents of breeder reactors believe thermal neutrons are the way to go and that we should rapidly develop the thermal breeder technology. These reactors can generate both energy and kindling (U-233)
    and they can get us out of the enrichment business, a huge sink of resources. And these kinds of reactors will not be just on paper for much longer e.g. China. Time is of the essence. The US needs to encourage the thermal breeder business with all dispatch.

    The bottom line is that are many more “camps” in this argument than two.

    1. If I understand the thorium fuel cycle correctly we have a few issues with the U-233 direction:
      1) The intermediate Pa 233 production and 1/2 life of 27 days means we have a difficult reactor to operate on short time scales and at shutdown we have a positive reactivity coefficient into the bargain.
      2) the re-fabrication of fissile U-233 into fuel looks to be difficult with the 2.6Mev gamma from the accompanying U-232 creating a handling problem throughout all stages of the operation.
      3) Rather than just being desireable, U233 breeding seems absolutely essential if you are to have enough fissile material to light up the wet logs. How else do you expand the reactor fleet? Either that or you fall back to U235/Pu239 in which case you deploy both fuel cycles.
      4) Reprocessing is an integral part of the thorium fuel cycle which must be carried out with high energy gamma emissions. This cost needs to be carefully compared to the apparent issues surrounding uranium enrichment which, with laser enrichment will only get much cheaper. That’s if they let it out of the back shed.

      I agree with your observation that there are more “camps” than two but I’d like to know a lot more about the apparent limitations of sodium cooled fast reactors before dismissing them. EBR-II looked to be a spectacular success especially with its load following ability.

      I’d appreciate comments as a little knowledge is a dangerous thing.

      1. Robert,
        From your list, it seems that you are considering the Th-U fuel cycle using solid fuel pellets, however most of today’s proponents only consider this fuel cycle using a molten salt reactor. The Shippingport reactor did run with solid fuel pellets that had thorium, successfully, but handling them for reprocessing does pose challenges. ORNL did run the Molten Salt Reactor Experiment using uranium fluoride fuel salt but they did not breed U233 from Th232 – that was to be done by the Molten Salt Breeder Experiment but it lost out in the budget wars and was not performed.

        1. I haven’t got time to research this right now, but I’m suddenly curious whether anyone has investigated the use of rods containing molten salt fuels in water-cooled reactors.  This would eliminate issues of fuel element life limitation due to pellet swelling, and might even bubble some of the neutron poisons out of the active region of the core into a dead gas space at the top.  It would also make reprocessing and recovery of fissile and fertile materials much easier and cheaper.

          1. I would rather have water in tubes as moderator coolant with a dual mode fission. Areas near the tubes will have thermal neutrons and remain critical as long as water is present. The liquid fuel further away will absorb neutrons for conversion to more fissile as a net neutron absorber.

          2. I agree with you, Jagdish. Inverted geometry, fluid fuel would seem to have some great benefits: the inherently lower cost of fluid fuel and the inherent safety (as regards control) of fluid fuel. If you moderate AND cool with water, then you have additional reactivity safety.
            Also, one thought I’ve had regarding such a design would be the possibility of not using molten salt, but instead using a slurry of pulverized SNF in molten Pb (or LBE). One of the drawbacks to using an oxide slurry is that it is very much like moving wet sand around — and the abrasion tears at the metal tubing. BUT, if the “sand” is static, there is no abrasion. Also, one of the drawbacks for LBE is corrosion that is enhanced by MOVING the LBE at high speeds. Again, keeping the LBE static will mitigate that problem.

          3. And the world’s leader in LBE design and operation has chosen SFR’s for their long-life commercial designs.

            Sodium is cool stuff — just keep it dry. These proposals for three-loop systems with a non-sodium third loop aren’t really conceptually different from Rod’s description of the double-walled tubing used in the EBR-II secondary loop. Can’t hurt to armchair, but do realize any really new fuel technology will take at least several decades to commercialize. And we needed “energy cheaper than coal” yesterday.

            So back to Rod’s question: “What role can the U.S. government — as opposed to the Sino-Rooskies’ — play that would actually make a positive contribution?”

            We tried a government-sponsored fast-reactor program. It didn’t work in the 90’s and it won’t work any better today: see Yucca Mountain.

            Between YM and IFR it’s pretty clear the U.S. government — i.e. you and I — are a pretty unreliable business partner. In the face of looming global warming I think we do need a SNF policy the public and states can buy off on (and into) to continue new LWR builds. YM appears dead and any replacement will take at least as long. Meantime we’ve had plenty of discussion to the effect that deep-storage is cost effective only for permanent disposal: if you’re going to recycle the stuff you’re far better off leaving it on or near surface in its existing dry casks rather than playing needless shell games moving it around. SNF actually is fairly hazardous.

            So what to do about it? Prof MacFarlane has observed there’s no super rush, but we do have to get it right this time, which I think is Rod’s point as well. NRC thinks dry casks are good for at least sixty years but of course will have to be monitored throughout and beyond. IFR / PRISM is designed specifically to maximize actinide burnup and minimize radiotoxic lifetime, and of course Transatomic thinks they can contribute there as well. The problem is getting the production prototypes licensed and built.

            NRC seems locked into this “everything is a gigawatt LWR” bureaucratic mindset. Its not like the individual staff don’t know how to figure out how to license other designs, but it’ll take an act of congress to get them to actually move. So which act? What should it include?

            Our brave new republican overlairds like to think in terms of budgets. NRC has reduced new reactor licensing in its 2016 budget; this directly after Leslie Dewan of Transatomic, Mike McGough of NuScale, and Ashley Finan of Clean Air Task Force all testified before House Science Committee that NRC’s new reactor licensing is a real bottleneck. From Finan’s CATF testimony:

            As with any energy technology, the development and commercialization of advanced non-light water reactors requires a suite of supportive policies from early research through demonstration and adoption. I will focus on two elements that need more attention: First, a testing facility that would enable private companies to build prototypes in a DOE-supervised environment; and second, a clear and predictable regulatory pathway for licensing advanced reactors.

            So there’s two items: (1) NRC licensing overhaul and (2) DOE prototype supervision. Dewan’s and McGough’s testimony pretty much concurred. To those I’d add (3) energy subsidy reform.

            EPA’s Clean Power Plan is — or might be if suitably amended — probably this country’s first honest effort to address The Coal Problem. A start, but It won’t be enough. I’d suggest scrapping the thing and replacing it with something like Citizens Climate Lobby’s Carbon Fee and Dividend, or Climateand Prosperity’s Cap and Dividend.

            Of course, Clean Power Plan and Fad/Cad are not mutually exclusive, they might be combined in CBO’s worst nightmare. The point is any of these will go further on less pain if augmented by other reforms that will actually open lower-cost pathways to reliable low-carbon energy. See Climate change: Low CO2 emissions tax possible if clean technology is also implemented. We can tax carbon all we want, but if we keep a stranglehold on effective low-carbon technologies, all we’ll end up with is slightly more efficient use of slightly fewer fossils. And if that stranglehold is only maintained on dispatchable low-carbon technologies, all we’ll end up with is modestly respectable carbon reductions on a dead-end path that will not and can not go the required distance.

            We do have to get it right this time.

          4. @E-P,
            I once thought of the same idea – tubes of molten salts in water-cooled reactors, but discovered a materials issue. There does not seem to be an appropriate material to use for the tubes. Some are good with the salt but not the water, or vice versa; some that might be good for both turn out to be bad in the radiation/neutron environment (they either degrade or they absorb too many neutrons).

          5. There does not seem to be an appropriate material to use for the tubes. Some are good with the salt but not the water, or vice versa

            Silicon carbide, maybe?  If you insist on using metals, a cladding of zircaloy over Hasteloy (or concentric shrink-fit tubes) might do.  I think the bigger issue might be preventing any separation of uranium from the melt, which could create a core configuration that cannot be made critical and require measures to replace or at least re-mix the fuel.

            Straight zircaloy-clad metal fuel with lead for a thermal bond might do as well.  Lead is very plastic and would squeeze out as the fuel swelled; a central hole could vent gas.

  3. Terrestrial Energy is developing the IMSR “burner” thermal reactor using the U235 cycle that allows 6 times the fuel efficiency the of PWR/U235 system.

    Is it economical to create a sodium cooled (or other coolant) “burner” fast reactor using the more abundant U238 that would utilize the fuel more efficiently than PWR/U235 systems? Is the PRISM reactor more of a burner?

    1. From my book reading as opposed to practical experience the answer to your question is a profound “maybe”.

      It is certainly technically achievable because the operation has proven successful in EBR-11, BN600, Phenix and elsewhere but your question included the term “economic”.

      This question occupies much of Rod’s article.

      Fast reactors can breed U238 to become Pu239, 240, etc and even higher actinides or burn all of these isotopes and those produced in PWR’s or any thermal spectrum reactor. With allied technologies such as laser enrichment they can make seawater extraction of 4.5 billion tonnes of uranium entirely feasible because the energy utilisation of the uranium has gone up 60 fold or more.

      The uranium fuel component ceases to be a determining cost and is almost totally subsumed by the reactor process itself being the cost.

      In our new era of liberalised electricity markets relying on short term trading we see the destruction of long term low cost generation such as nuclear by short term “low cost of entry” players such as gas turbines or mandated “renewable” wind generation.
      Any of these new entrants can destroy the business model of the high capital cost player especially when regulation burdens them with a lot of high bars to jump.

      This is where a price on pollution becomes essential. The low cost gas doesn’t pay for wrecking our climate with unregulated greenhouse gas emissions, nor does the coal. The wind turbines and especially the solar PV’s doesn’t pay for their pollution either because most are produced with power generated by some of the filthiest power systems on the planet.

      So now we come back to the profound “maybe”. Short term liberalised markets don’t created opportunities. They are are opportunistic parasites that hollow out society’s successes until the edifice crumbles and we are left with – nothing.

      If we can put the appropriate cost on pollution into the system then the answer to your question will be a resounding “yes”. We just need to stop giving the parasites a free ride.

      1. @ Robert

        Per your description of liberalized energy markets, one might conclude any nuclear energy technology might not be economical [clearly you are speaking of IFR]. At 14 billion per 2 GW, I think the AP-1000 is a steal. Especially if is able to operate 80-100 years or more.

        I am not that familiar with IFR technology. What specifically about the technology is creating the “maybe”? It operates at low pressure/high temp, has passive safety, and utilizes fuel efficiently. In Gordon Mcdowell’s video, the Thorium group visits the EBR-II and speaks with Roger Blomquist, PHD Nuclear Engineer.


        It is explained that the IFR’s “burning” of U238 and the recycling or reprocessing are segregated operations from each other [but done on the same site]. What are the challenges of IFR – is the burning of the fuel, the recycling of the fuel, or a combination of both?

        Also, for those that are interested, Gordon posted a new video of Kirk Sorensen explaining in detail the LFTR blanket/core Th-232/U233 fuel cycle. The LFTR breeder concept appears to be a connected loop, while the IFR process is segregated. (please correct me if I’m conceptualizing this wrong)


        1. I think you have it right, Tom. But keep in mind that IFR and LFTR are two very different beasts, and can do different things. Not to dampen anyone’s enthusiasm, but at present LFTR remains a concept — none have ever been built. Enough good engineers have worked on LFTR designs that I’ve no doubt they can be, but they haven’t yet.

          A U-Pu sodium cooled metal fueled fast reactor has been built, and operated for about thirty years – EBR-II. Fast reactors (aka fast spectrum reactors or fast neutron reactors) can burn all actinides as fuel, including the transuranics Np, Pu, Am, Cf, etc in addition to U-235, U-238, and Thorium. Fast reactors are attractive for that reason: they can burn LWR “spent” fuel, as well as the depleted uranium from which the LWR fuel came.

          There is *lot* of depleted uranium, mined, refined, and ready to go. At current electric generation rates, the U.S. has on hand enough depleted uranium and “spent” nuclear fuel to power the country for a bit over 900 years if burnt in fast-spectrum reactors, with no additional mining.

          Metal fuel is not necessary to do this, mixed oxides and nitrides may also be used. The EBR program chose metal fuel because it lent itself most economically to a compact integrated recycling facility: Electro-metallurgical pyro-processing works mostly with metals (electro-refined from molten salts) anyway, and the IFR designers wanted fuel pins that could reliably be cast robotically so that the radioactively hot transuranics would never be in a separate stream, assuring a (very) high degree of proliferation resistance.

          Metal fuel pins also allowed the noble gas neutron poisons (Xe, Kr) to escape without undue swelling, and also gave superior heat transfer to the sodium coolant, which reduced residual heat to where the primary sodium pool itself had enough thermal mass to contain all heat in a worst-case station-blackout scenario.

          GE-Hitachi S-PRISM is the IFR commercial design. Its fully ready for the NRC licensing process, whenever NRC and GE-H can both devote the manpower and dollars to the effort. GE-H inquired into this five years ago, NRC then was busy with the light-water SMR applications and essentially told them not to bother. GE-H took the hint and focused full-time on ESBWR instead, perhaps they’ll return to S-PRISM after they get a few of those 1.6 GW behemoths financed and under construction.

          ESBWR has some nice properties, and a good handful of potential customers. Who’ll probably wait until after the EPA’s Clean Power Plan is finalized this summer, and see its reception in Congress. Basically hold off until they get a better reading on how much this country actually does value very low-carbon generation.

        2. PRISM can act as a burner or breeder. It was designed to be a breeder. I think the main things holding it back are:
          1) the reprocessing technology has never been demonstrated at full-scale,
          2) I doubt regulatory authorities will be keen on using steam generation to make electricity, because sodium explodes on contact with water, when it has a clean surface. Sodium will always have a clean surface in a PRISM. PRISM is designed to use a primary and secondary sodium coolant. The secondary coolant loop then drives a steam turbine via another heat exchanger. A Nitrogen driven turbine would be far safer. You can read about about the PRISM design in the book Plentiful Energy, available as an e-book here.

          LFTR is a great idea but all the technology needs much more development than the PRISM. It’s at least a decade further away.

  4. OT:

    I just finished my own little investigation into heat-pumping and thermal storage for nuclear power plants.  By damn, it looks like it would work!  You could get a ratio of better than 2:1 between minimum and maximum output while maintaining max thermal output, while also generating temperatures sufficient for a number of processes for making pyrolytic biofuels.

    1. @ EP,

      I thought this was exactly the design that Carl Able was working on. (I think I have that name right). He was a guest on AI several times. Yes, using heat pump tech would increase temps to the point where process heat is available. Molten Salt storage of the heat would enable load following that is disconnected from the reactor operation thus bypassing the NRC concerns.

  5. One of the canards frequently used by those opposed to SFRs is the French experience with their 1,250MWe Superphénix plant.
    The claim is that it was shut down due to technical problems.
    Not so.
    It was shut down as part of a political deal between Frances Socialist and Green parties.

    For those who can read the language, here is a series of articles about Superphénix from the May 1997 issue of the French trade journal, Revue Générale Nucléaire.
    Note especially the second one – a statement by the French Nuclear Society (SFEN) to the attention of political leaders, in support of continued operation of the FBR, as well as the last one, describing the highly reliable operation of the plant in the year prior to the politically-mandated decommissioning.

    France’s current SFR development project, ASTRIDE, recently made an important decision to switch from steam Rankine cycle to nitrogen Brayton.


    1. I would love to see a translation of this material all in one place, Jaro. Up till now, virtually all the info I’ve read on Superphenix pointed to its failure, but it would seem it was on the brink of demonstrating its potential. No wonder the ideologues where so frantic to pull the plug, much like with IFR…

  6. Before the Tsunami, the Japanese were putting a lot of effort into “reduced moderation” reactors. As I read the descriptions of the concept, a breeding ratio slightly over 1 could be achieved and the production of weapons-grade material could be avoided.

  7. Why not accelerate the fusion reactor, such as ITER.
    No radio-active waste, no risk on disasters such as Chernobyl or Fukushima, endless cheap fuel, etc.

    The Japanese Monju breeder story is enough to provide all politicians cold feet.

    1. @Darius

      No radio-active waste, no risk on disasters such as Chernobyl or Fukushima, endless cheap fuel, etc.

      You forgot to mention that fusion also has no proven path to producing reliable electricity on any scale whatsoever.

      I have no personal interest in blue sky research. I’m an applied science guy through and through.

      1. Global fast breeder experience is a real bad story to build on.
        Capacity factors of:
        – Japanese Monju ~1%?
        – France Superphenix <10%. And that after gathering experience with two earlier reactors.
        – Russan BN-600 unknown. Their refusal to open up around accidents doesn't give confidence.
        Neither their marginal scaling up to only one new reactor of only 880MW.
        Those indicate the costs are prohibitive in the end.

        Little reason to assume a similar design will do better.
        We need something radical different and better.

        1. @Darius

          You have ignored successes like EBR-II, Phenix and the Light Water Breeder Reactor.

          Costs for the first of a kind of anything are always extreme. Lack of follow through is more due to politics and extreme, focused efforts to demonize plutonium — to the point of making fuel recycling illegal in the US after commercial companies invested hundreds of millions into a plant that was ready to begin operating.

          Low capacity factors for fast reactors still provide far more support for the notion that they can produce vast quantities of power than the experience we have of no net sustained power from fusion. Even on a lab scale, there is not a repeatable demonstration anywhere.

          1. We need much cheaper power as NPP’s are competed off the market: Once nuclear delivered 17% of the world’s electricity. Now it’s 10% going down.

            But cheaper is not what breeders deliver.

            The expected shortage of Uranium drove breeders. But that idea was wrong because far more reserves were detected, and nuclear didn’t boom as expected. Uranium is very cheap for NPP’s and that is expected to stay so.
            So we don’t need them in coming decades.

            Why then pushing expensive, risky breeders?

            Phenix, the predecessor of SuperPhenix wasn’t a success either. Check the costs, capacity factor, etc.

          2. @Darius

            Because we do need them. We need them to replace fossil fuels because of climate change, and we will definitely need them eventually if we want to keep living a decent life (short of fusion coming through, something I’m not holding my breath on). Why not start working on it now?

            Slightly of topic, someone please tell me what they think of this idea.

            I was reading this article about accelerator driven nuclear reactors and I had an idea. Maybe particle accelerators can be used to create nuclear fuel from uranium 238 which could then be reprocessed and used to start fast reactors. It seems like that should be something doable even with the particle accelerators that exist today.


          3. @ Evan

            The idea is not unheard of. A decade or so back, the concept of an “energy amplifier” was described: An accelerator driven nuclear reactor using Thorium, not Uranium (perhaps a selling point was that it didn’t produce Plutonium).
            I note in passing that the term “accelerator driven nuclear reactor” is sort of a misnomer under the US regulatory environment, where a “nuclear reactor” is defined to require a “self sustaining” chain reaction.
            There has been work at making less expensive particle accelerators. If and when that happens, then the concept will really have legs.

          4. @Evan:”Why not start working on it now?”

            Because we may hope that technology will progress in next 50yrs such that then developed designs have more chance to succeed (better controls and materials, etc).

            The accelerator driven nuclear reactor looks as if it will produce electricity for significant higher cost price. We need the opposite as wind and solar become cheaper.

          5. @Darius

            50 years seems way too long to me. We can burn a lot of fossil fuels in 50 years. Even with reprocessing it would be hard to replace fossil fuels without any breeder reactors for any significant period of time.

            I think the reason why Uranium supplies aren’t a problem now is because nuclear power is only supplying a small part of the world’s total energy. If we were to really ramp nuclear up I think we could go through the good ores before long, and the not so good ores require breeding to make work.

            There are reactors designed that can breed, or close to it, that we can build now and which seem good enough to me. I think we should start building them. We need the knowledge, the expertise and the infrastructure that can only be gained by doing. As for accelerator driven nuclear reactors, I’m not talking about build them. I’m talking about using accelerator to make nuclear fuel for starting fast breeder reactors. This way we could get them going faster. Maybe some of the accelerators that already exist can be used to create fuel when they aren’t being used for other things.

          6. In its last year of operation Superphenix had a CF of 95%.
            Does that sound like a to you, Darius ?
            See my post above for details.

          7. @Evan: I’m thinking you have a mis-perception of ADS (accelerator driven systems), sometimes called “energy amplifiers”, “sub-critical reactors” or “intermittent nuclear energy source.”

            At present there is no such thing as a reliable accelerator, not in the commercial power sense. ADS were proposed because their super-hard neutron spectra can, in principle, burn essentially every last transuranic atom, and quite a few of the fission daughter products as well. But that’s purely a political consideration: given just a couple or three centuries all those things are effectively going to go away anyway. Another term might be “Radiophobic Driven Systems.”

            But that wasn’t your question. You asked about using an accelerator to transmute U-238 to e.g. Pu-239 more efficiently than could be done in a fast rector. The short answer is “I don’t think so.”

            Accelerators are energetically very inefficient: it takes a lot of electricity to run them, and the actual beam currents in ampere’s are rather modest. And they typically accelerate protons that are then smashed into a lead target to generate the desired neutrons, losing a lot of energy in the process.

            And for U238 -> P239 conversion you don’t need really high energy neutrons. The neutron spectrum only needs be fast enough that P-239 is bred faster by fast neutrons than it is fissioned by slower neutrons. And sodium-cooled breeder reactors typically attain a breeding ratio of about 1.2: they breed five Pu239 for every four they split. Which isn’t bad at all, considering all the electricity you can generate from all those fissioned plutoniums.

            Heck — you probably needed it anyway 🙂

            The fly in the ointment is the part where fast neutrons aren’t particularly swift at fissioning U235 and Pu239, and its those fissions that sustain the reaction. So you need pretty concentrated fissiles to get these unmoderated fast reactors going in the first place, typically about 20% enriched U235, as opposed to 3 to 5% in a LWR and even less in a CANDU. And that high fissile concentration has to be maintained as U235 is fissioned breeds U238 to Plutonium. Enriching that initial U235 is going to cost, both in energy and mined ore, particularly if we’re going to power a planet on it. Its why we’ve got to plan this thing in advance.

            Its also why Rod questions the wisdom of configuring the UK’s first S-PRISMs (if that’s what they decide to do) as burners just to get rid of their 100 tonnes enriched plutonium, which one would think could be more economically used to breed more plutonium, assuming that’s how the Brits wanted to power the kingdom rather than off imported uranium.


        2. What was the cause of these low capacity factors? Was it just that these things were experimental and there wasn’t the incentive to keep them on line? Was it simply mechanical problems (teething problems) with new processes? Or,…..was it a bizarre politics that prevented success?

          Seems like there’s been a lot of lessons learned as to what worked and what didn’t already. They have a long history. If another one is ever built in North America, perhaps the wheel would not have to be reinvented. Taking baby step innovations could provide greater assurance of safe facilites. It could also provide greater assurance of successful operation.

        3. Darius, the operating record of the BN-600 is actually rather good, sometimes said to be the best of any Russian reactor – data can be seen here: http://www.iaea.org/PRIS/CountryStatistics/ReactorDetails.aspx?current=484 – lifetime load factor is 74%.

          More details on operating history, including issues during commissioning, leaks and other incidents etc. for the BN600 (and other SFRs) can be found in this IAEA TECDOC: http://www-pub.iaea.org/books/IAEABooks/7945/Liquid-Metal-Cooled-Reactors-Experience-in-Design-and-Operation

          You’ll see that there have been a number of sodium leaks, but they have all been dealt with effectively. Perhaps the best lesson is that leaks will happen but are not necessarily a significant risk with well thought-out design and operating procedures. Taking a zero tolerance approach to all sodium leaks makes no more sense than seeking to eliminate 100% of steam leaks or oil leaks – both have the potential to cause great harm (via fires, flooding, explosions, loss of essential supplies, direct injury to personnel etc) if combined with badly designed or operated equipment, but vast experience has enabled us to minimize the rate of incidents and handle failures gracefully. This is more sensible than just saying ‘no leaks ever’.

    2. @Darius: Are you implying ITER is being starved of funds? Maybe, but in Britain fusion research gets 100% of state funds directed to nuclear reactor R&D. I think nearly all of that ~ £35m/year, goes to ITER.

  8. Fast reactors and reprocessing are required to dispose off the used nuclear fuel from the present generation of nuclear reactors. It is becoming increasingly more difficult to mine uranium due to political opposition as well as a protest against the increasing quantities of used fuel. Closed cycle increases the energy from same amount of mined uranium by two orders of magnitude. Russians do not find the cost of fast reactors prohibitive. China and India also find closed cycle worthwhile. The rest of the world can make their own choice.

    1. “Fast reactors and reprocessing are required to dispose off the used nuclear fuel from the present generation of nuclear reactors…”

      This, I think, is the real reason to proceed with the fast reactors. Rebrand the current “waste problem” as “fuel supply.”

    1. So you think we should shut down the ONLY fast-spectrum breeders ever proven to work under actual commercial-scale conditions.

      Why not just admit that your name is really Amory Lovins?

      1. The original Wright Flyer is hanging in the Smithsonian. Does that prove that early airplanes were safe? Should the Wright Flyer have been put into mass production? How many would have died in Wright Flyer accidents if thousands were built? How many decades would that have set back aviation?

        At that time there were no other designs to choose from, but we can choose reactor designs that are safe from high energy criticality accidents.

        EP, glad to see your best technical response is a cheap insult, it reinforces my point.

        1. @Bill Hannahan

          Your analogy is interesting; please think a little more deeply about what you are advocating.

          You said that fast reactors that cannot prove they are immune to some kind of postulated accident should be shut down and have their fuel removed. That action would halt opportunities to learn by doing and would eliminate valuable fuel testing facilities that have unique neutron spectra available.

          It would turn all fast reactor innovation into a complex, unprovable modeling exercise. Please recall that there have never been any fast reactors put into mass production. They are operated and maintained by a large group of specialists, not by a minimal crew in a commercial setting.

          As you noted, the Wright Flyer was unsafe in the hands of anyone who did not have the proper training. It was also unsafe in the wrong weather conditions, above a certain altitude, without the right fuel, and a host of other limiting conditions. In fact, those statements remain true of any heavier than air flying machine, though we have made many improvements during the 111 years since the first flight at Kitty Hawk.

          Would you have advocated grounding the Flyer or most of the hundreds of iterations required between it and the Dreamliner?

          If you want to say that commercialized versions of solid fuel fast reactors need to assure regulators — perhaps through physical testing — that they are sufficiently immune to your postulated accident before they are mass produced, fine. That is a different, more consistent position to take.

          1. Rod wrote;

            “you said that fast reactors that cannot prove they are immune to some kind of postulated accident should be shut down and have their fuel removed. That action would halt opportunities to learn by doing and would eliminate valuable fuel testing facilities that have unique neutron spectra available.”

            Rod, your essay is not calling for a few experimental reactors to be built and fully tested in remote locations. I would approve of that. But first, it makes sense to run computer models of high reactivity insertion rate criticality accidents, and to explore the possibility of multiple criticality accidents in which a medium energy criticality could rapidly crush another portion of the core causing a high energy criticality.

            Or they could start with the experiment I suggested in one of the comments;

            “To prove fast neutron solid fuel reactors are safe, model an accident that envelopes all possible accidents. Start with a core early in startup, very low power, K slightly above 1.0 now crush the core into a pancake in 0.1 seconds, squeezing out all the sodium. If the resulting energy burst is reliably contained I’m satisfied. Feel free to do this with a water moderated reactor or MSR.

            In reality, the authors of NUREG 1368, presumably strong supporters of the IFR employed in its design, acknowledged that;

            “Regarding accommodation of HCDAs [criticality accidents}, there is not sufficient data to confidently predict the size of an HCDA in a metal fuel ALMR.”… “The major contributors to core melt all lead to energetic core disassembly accidents and Release Category R4A.”…

            “For the R4A no-evacuation case, prompt fatalities were shown to increase from 7 to 124, and latent fatalities increased from 1,520 to 3,320.”

            The point is that a mild low reactivity rate criticality accident can be far worse than Chernobyl, and the authors say nothing about the results of a high reactivity rate accident except this;

            “On this basis, the designers contend that core melt and sodium boiling do not have to be considered in the design”

            That sounds very similar to what we heard in the 70’s, We don’t have to worry about core melt accidents because the PRA proves it will probably never happen.

          2. Off Topic I’m wondering what people think of this idea.


            @BILL HANNAHAN

            It seems that in nuclear power discussions people are always referencing things that happened before I was born. I was never alive during the 70’s and the whole to cheap to meter thing was way before my time a well. I think we should try to keep things at least a little grounded in the present.

            Why is the burden of proof always on nuclear power to prove that can survive any imagined event, and why are the safety standards put on nuclear always so much higher than they are on other technologies? It seems unreasonable to me. No other industry is forced to accept responsibility for things that would normally be thought of as acts of God.

          3. Rod wrote;

            “Would you have advocated grounding the Flyer or most of the hundreds of iterations required between it and the Dreamliner?”

            No. At each point in time people choose the best option available. DeHavilland sold few Comet’s after multiple accidents traced to design defects, Boeing sold many 707’s.

            “If you want to say that commercialized versions of solid fuel fast reactors need to assure regulators — perhaps through physical testing — that they are sufficiently immune to your postulated accident before they are mass produced, fine. That is a different, more consistent position to take.”

            This is consistent with my position.

            Rod, do you think it makes sense to build very complex, expensive plants like the IFR that will take a long time to build, and will produce expensive kWhrs and have the potential to produce 124 prompt fatalities, and 3,320 latent fatalities from a slow gravity induced criticality accident, and a huge area of contaminated land far worse than Chernobyl and the possibility of a much more energetic accident unexplored?

            We have designs that can contain a full meltdown with less impact than 1 week of normal operation of a coal plant.

            Sea water uranium is estimated to cost about three times more than land based mines, but simple MSR’s can use 5-6 times less uranium per kWh than old plants. Progress is still being made despite a low level of R&D.


            Fuel will be cheap for a very long time, we will not NEED breeders for a thousand years or more, why take the additional risk?

          4. Evan wrote;

            “Why is the burden of proof always on nuclear power to prove that can survive any imagined event”

            Who should it be on?

            “why are the safety standards put on nuclear always so much higher than they are on other technologies? It seems unreasonable to me. No other industry is forced to accept responsibility for things that would normally be thought of as acts of God.”

            The root cause is the failure of our education system to teach people how to evaluate risk vs benefit in a high tech world. It’s emotion instead of science.

          5. @BILL HANNAHAN

            I think that my last comment was a little too abrasive. For that I apologize. As for who should have the burden of proof it depends. The people building nuclear power plants should do some kind of due diligence (i.e. some reasonable amount of investigation of possible threats) to insure that their power plants are safe, but I think that after that due diligence the burden of proof should shift to the people who believe the nuclear power plants are unsafe. Otherwise the burden put on nuclear is limitless as the human imagination which stifles the nuclear industry to no end. This doesn’t seem fair, either to the nuclear industry, or to people like me who wish to benefit from the many advantages nuclear power has over other forms of generation.

          6. Evan: ” I was never alive during the 70’s and the whole to cheap to meter thing was way before my time a well. I think we should try to keep things at least a little grounded in the present. ”

            Bringing up the “too cheap to meter” is an anti-nuclear tell. It comes straight out of the anti-nuclear play-book and has no relevancy to any current argument, but it’s one of the distractions that the anti-nuclear folks use to obfuscate the discussion and turn attention, when they might otherwise actually have to deal with facts not in their position’s favor.

            If you’re reading a discussion and a person brings that up, you can pretty much stop reading there. The person who brought it up has no interest in having a meaningful discussion based on an exchange of facts.

            Whether the people using this play book are paid shills or just misguided fanatics is left as an exercise for the reader.

            Of course, “too cheap to meter” was never a relevant part of the discussion, as the context of the original quote was regarding fusion, and there is every reason to believe that the speaker meant that one would just charge a flat rate for electricity service, but production of electricity would be so inexpensive that there’s be no economy to metering how much home users actually consume. Reading the meters and assembling the bills without automation was a substantial expense at the time.

        2. EP: “Why not just admit that your name is really Amory Lovins?”

          BH: “EP, glad to see your best technical response is a cheap insult, it reinforces my point.”

          It’s nice to see that people believe that being called ‘Amory Lovins’ is an unmitigated insult. That’s some progress. 🙂

          1. Thanks Jeff, it had never occurred to me that EP might be trying to pay me a complement. I’ll wait for confirmation to apologize.

        3. Bill – Good point. I like the way you think, so I demand that all airplanes should be defueled and grounded until proven incapable of a high energy gravitational accident (or “hitting the ground” in the layman’s vernacular) by fundamental principles of physics.

          The recent TransAsia accident in Taipei demonstrates just how unacceptably dangerous these airplanes are.

        4. The original Wright Flyer is hanging in the Smithsonian. Does that prove that early airplanes were safe?

          The BN-1200 is 64 years beyond the EBR-I, twice the span of time between the Wright Flyer (1903) and DC-3 (1935).  There are DC-3’s still flying today; they are more than adequately safe for their purpose.

          You don’t mention how anything could produce the level of compression in your scenario.  If your accident scenario requires the direct impact of a large meteor on the reactor core, why are you even bringing this up?  Not even a Chelyabinsk-level event would have had that effect on a reactor in a reasonable containment.  A chondrite mass would shatter against reinforced concrete, so even a half-ton direct hit wouldn’t squeeze a core the way your scenario requires.

          But I think most egregious of your errors is your disinformation about neutrons in fast vs. thermal reactors.  The actual lifetime of a free neutron is more or less irrelevant to the reactivity.  What matters is the fraction of delayed neutrons and how many are required to maintain a stable chain reaction.  Typically reactors are designed with a negative temperature coefficient, such as by Doppler broadening, under-moderation or change in leakage with temperature.  These things work just fine to control even fast-spectrum reactors; the EBR-II demonstrated thermal shutdown without any use of control rods close to 3 decades ago.

          glad to see your best technical response is a cheap insult

          I just thought you’d like to be compared to the one who’s famous for creating the disingenous arguments that you use here.  If you didn’t think well of him, you wouldn’t emulate him.

          1. EP wrote;

            “But I think most egregious of your errors is your disinformation about neutrons in fast vs. thermal reactors. The actual lifetime of a free neutron is more or less irrelevant to the reactivity. What matters is the fraction of delayed neutrons and how many are required to maintain a stable chain reaction.”

            EP, I can state two facts based on this comment.

            1… You did not follow all of links and read them carefully.

            2… Your memory is failing.

            I have explained that in a nuclear explosion the vast majority of fissions take place in a very short increment of time. Any delayed neutrons released during that pulse are not really delayed. The delayed neutrons are released long after the fissile material is dispersed, producing an insignificant number of fissions.


            The actual lifetime of a free neutron is extremely relevant in determining the energy yield, which explains why fast neutron criticality accidents can be far more energetic than those with thermal neutrons. It explains why nuclear weapons can only be made with fast neutron assemblies. It is the difference between the low energy BORAX explosion and the Hiroshima explosion.

          2. I have explained that in a nuclear explosion the vast majority of fissions take place in a very short increment of time.

            The meltdown of the core of a fast-spectrum reactor has almost no resemblance to the implosive collapse of a bomb pit.  You put your own amateur analysis above that of Hans Bethe himself, which is risible.  Several fast-spectrum reactors have melted their cores, with no explosions to date.  Last, we have examples from a number of fast-spectrum criticality accidents with bomb pits, including fatal ones, none of which caused explosions of significant energy.

            Bethe’s analysis was a worst-case scenario assuming that more or less everything that could go wrong, did (like the engineering of the BORAX test-to-destruction).  It showed that even the worst could easily be contained.

            You present yourself as an authority on this issue, but you aren’t.  You’re just another FUD-spreading ideologue.

    2. I believe SFRs will be incapable of high energy critical accidents should they use N2 to drive turbines instead of water. Sodium contacting water is the only major problem here.

      1. @Mark: That may depend upon one’s definition of “major”, and the relative thermal efficiencies of steam turbine systems vs. N2. One of the design goals of S-PRISM is conservative design, GE knows how to make steam turbines work.

        I think we need be careful about assessing Na-water reaction. It is indeed highly exothermic — but steam generators transfer heat. But Na + H2O -> Na2O and/or NaOH + H2 is explosive only in presence of air to ignite the H2. Its not as though the possibility of steam-generator failure hasn’t been considered. From http://gehitachiprism.com/wp-content/themes/geh_prism/resources/PRISM_Triplett_Loewen_Dooies.pdf

        A sodium leak detection system provides early warning of any sodium-to-air leaks from the IHTS. In the event of a steam generator ~SG! tube leak, the sodium-water reaction pressure relief system ~SWRPRS! provides overpressure protection of the IHTS and IHXs. The SWRPRS consists of a safety-grade rupture disk, a separator tank, a vent stack, and a hydrogen igniter. To separate the reactants, the SWRPRS initiates the waterside isolation and blowdown of the SGS and the purge of the SG tubes with nitrogen.16

        — although its unlikely a single paragraph will convince you. Neither am [i]I[/i] going to strenuously argue against N2 Brayton systems — I just don’t know enough about them.

        1. Rod just tweeted a link to a recent Kirk Sorensen U-Tube video: LFTR Chemical Processing & Power Conversion wherein Kirk explains recent advances in super-critical CO2 turbines — a DOE priority. In Kirk’s application they would run at 550 – 650 C. PRISM’s primary outlet temperature is 500 C, secondary 480 C. So its still a bit low.

          But only a bit. Sodium boils at 880 C.

        2. A much better understanding of the sodium-water reaction has recently been reported. Apparently the presence of oxygen and ignition of hydrogen are not the drivers for an accelating reaction. Rather, the electron transfer process and resultant coulombic repulsion are what are important. For a very accessible explanation please see the video at Nature: http://www.nature.com/news/sodium-s-explosive-secrets-revealed-1.16771 .
          Clean sodium in humid air has never struck me as particularly risky, but assiduously avoiding contact with liquid water seems worthy of careful engineering.

        3. FWIW, related factoids from the BN800 courtesy Google Translate:

          SG is intended for generation of superheated steam as a part of power unit of a three-loop BN-800 plant with fast neutron reactor and sodium coolant in the primary and secondary circuits. SG structure and fastening provide:

          *normal continuous operation under OBE of magnitude 6 according to MSK 64 scale;

          *normal operation of the secondary side under SSE of magnitude 7 according to MSK 64 scale, plant cooldown under all modes except for the cases of loss of feedwater supply in modes “water-stop”, “loss of power”;

          *keeping SG in operation under the condition of “small leak” of water into sodium by quick automatic disconnection of the section containing the failed module;

          *protection of the secondary equipment against overpressure under the condition of “large leak” and bringing SG out of operation with discharge of products of water-sodium interaction into the emergency protection system tanks;

          *integrity of module body, secondary equipment and pipelines under pressure increase in them in case of instantaneous guillotine break of one heat-exchanging tube.

      2. Any EBR-II-derived LMFBR with a secondary metal coolant loop is incapable regardless.  If you lose cooling the core temperature increases, the structure expands, and the increased neutron leakage shuts it down.

        1. To expand a bit on E-P’s remark, the secondary sodium loop is not radioactive. The tertiary steam loop is not radioactive. There is only a finite amount of sodium and water in either. Even in the absolute worst-case beyond-any-basis-event-imaginable scenario where all shut-off valves and fire suppressors and nitrogen floods fail and all the secondary sodium leaked out in smoke, flames, and loud noise, the reactor itself will be unaffected beyond automatic natural-expansion shutdown. Even if a steam generator were totally destroyed and the sodium reactant products completely escaped the SG containment (???), it would be far from the end of the world.

          Caustic soda and lye would indeed be a mess to mop up. With vinegar or dilute HCL and wet-filter respirators. . But that’s all it would take: hoses, mops and weak acid. All that’s left is table salt. We aren’t talking Bhopal here, or Lac Megantic, or the tens of thousands of gas wells leaking methane and cyclic aromatics over all oblivion and gone.

          Its probably the inevitable

          “Smoke and flames and loud noise reported at NPP, residents evacuated. Governor demands answers, announces inquest. We’re all gonna DIE!!!”

          headlines that have everyone all worried. Is why NRC triple-checks the SG design, along with everything else.

        2. The Sodium thing with water does not give a lot of people a warm fuzzy.

          Both with these postulated fast reactors and the molten salt reactors (Thorium or otherwise), the discussion has been of higher temperature units. This has led to people talking about Carbon Dioxide combined cycle turbines and Jaro mentioned that the French are looking at a Nitrogen gas turbine.

          Many years ago when I went to school the idea of Magneto hydro dynamics (MHD) was discussed.


          It’s been tried out a few times with limited success. The beauty of it is that there are less moving parts. If you could get it to work, the KISS principle would apply. From the link:

          :”A coal-fueled MHD generator is a type of Brayton power cycle, similar to the power cycle of a combustion turbine. However, unlike the combustion turbine, there are no moving mechanical parts; the electrically conducting plasma:
          provides the moving electrical conductor.”

          I was just wondering if you could replace the experimental Carbon Dioxide Bayton cycle turbine with an experimental MHD unit. The liquid Sodium would need to heat an intermediate loop of hot gas. I suspect that too many new things being tried at once is looking for trouble.

  9. If I may weigh in on this question of sodium vs. lead (or lead alloys) in fast reactors, I would like to note that the real limitation of sodium is its low boiling point, 883 C. Of course the boiling point is merely the point at which the vapor pressure is 101 kPa, there are considerable pressures at lower temperatures. (One can easily vaporize sodium in the lab, and in fact see it boiling at low pressures.)

    Unfortunately, many hydrogen cycles, as well as carbon dioxide splitting cycles and even water gas shifts and Boudouard reaction shifts (at ideal equilibria) involve temperatures that are higher than 883C, and thus to access these using sodium cooled reactors one is talking about pressurized sodium.

    Speaking only for myself, I don’t like it, particularly with the accumulation of 24Na. While fully conceding that any pressurized sodium reactor would still be safer than the cleanest coal plant ever built, I can certainly imagine a festival of idiots complaining about that kind of reactor.

    In recent years, the more I think about, the more I’m a lead cooling kind of guy, although certainly there are an infinite number of options, effectively for cooling, but I like neat lead, as opposed to LBE.

    Historically lead had a set of corrosion properties, some of which have been addressed to some extent by the use of alloys like T91 and 316SS. Unfortunately these tests have been run, to my knowledge, at fairly low temperatures, and they use LBE.

    I’m for doing away with the bismuth to eliminate or minimize to 210Po problem and considering neat lead.

    I would note that many materials science developments in recent years offer wonderful possibilities for new (or in the case of lead, old) coolants. Some of these are ceramic or metalloceramic coatings, such as “MAX phases” – Ti3SiC2 has been extensively tested with lead, but there are many other types which may be even better.

    By the way, I don’t think the “we’ve already built reactor x or reactor y” argument is a very strong one. It’s conservative in the sense that it is an argument that nothing should be tried for the first time. It’s a trap, and a bad trap for nuclear energy.

    We had wonderful ideas at the dawn of the nuclear age, and some were abandoned not because they were bad ideas but because the technology of the time was not sufficient to scale them.

    I wrote about one of these interesting reactor concepts I recently had a good time reading about on another website: http://bravenewclimate.com/2014/12/29/current-world-energy-demand-ethical-world-energy-demand-depleted-uranium-and-the-centuries-to-come/

    I actually downloaded some pictures of the LAMPRE reactor. The whole thing operated at 1MW and and one could imagine it, with a little modification, actually sitting in a locomotive. It was a very cool little device, but, unfortunately, ahead of its time.

    It was sodium cooled, and therefore limited to operating in the 500C-600C range. We could do better.

    I really have nothing against sodium cooled reactors, and as is the case with any type of nuclear reactor built, they will never be as dangerous or as deadly as a coal or gas plant, but that said, I confess to being a lead guy myself, but I would never be one to argue that the world of possibilities is limited to that.

    As for molten salts, there is a universe of those as well, with FLIBE and FLINAK being a tiny subset of what is possible.

    It’s quite possible that we who know and understand nuclear energy, who recognize it for what it is, the best possible source of energy, have been bludgeoned into insensibility by the tiresome and ignorant opponents of our efforts to save the world.

    Let us think broadly.

Comments are closed.

Recent Comments from our Readers

  1. Avatar
  2. Avatar
  3. Avatar
  4. Avatar
  5. Avatar

Similar Posts