Power Problems
The Costs of Nixing Nuclear
show comments
  • “Fortunately, there’s a new generation of nuclear reactors coming down the pipe, reactors that are fail-safe by design…”

    That sounds lovely, but existing reactors are already more-of-less fail-safe by design. They’re not passively fail-safe, but the failure chains all have very low probabilities. Still, they’re complex systems, just like the next generation of reactors will be, and therefore subject to unanticipated failure modes. I’m willing to believe that they might be somewhat more safe than the current generation, but we should be conservative. So: how dangerous are the existing set of reactors?

    A few weeks back, on the 3-year anniversary of Fukushima, I did a quick back of napkin:

    In the 60 years we’ve had commercial nuclear power, we’ve had three major disasters (>INES level 6): Chernobyl, Fukushima, and the Kyshtym disaster. These are the only accidents that have rendered any appreciable amount of land “unsafe for residency” (which IAEA sets at 5 Curies/km^2.) By that definition, the three disasters have rendered 28,000, 1100, and 650 km^2 uninhabitable, respectively. I’m fudging a little because I’m piecing together various maps, but I’m well within a factor of 2. Call it 30,000 km^2 in 60 years, or 500 km^2 per year. It’s always nice to have some idea of how big that is: 30,000 km^2 is a square 173 km on a side, and 500 km^2 is a square 22 km on a side.

    That’s a non-trivial amount of land rendered uninhabitable, but the half-life of Cs-137 and Sr-90, the two big gamma emitters that you have to worry about in nuclear accidents, are both about 30 years. So, assuming a major accident every 20 years, within 90 years you should have reached a steady state where 40,000 km^2 of land is uninhabitable at any one time, but land is coming back into service at about the same rate that it’s going out of service. That’s going to compare very, very favorably with the amount of land rendered uninhabitable by coal, oil, and gas production.

    (BTW, a lot of people prefer to use a 2 Curie/km^2 threshold. Those numbers were harder to come by–feel free to triple or even quadruple the numbers I have above. It’s still a relatively small number.)

    • louis_wheeler

      There are passive fail safe designs; we are just not using them. Liquid Fluoride Thorium Reactors is one. This design is not commercial, yet, but China is working on it.

      The US nuclear power industry made a wrong turn in the 1950s, because of politics. The world uses Light Water Reactors, because the US military wanted them to power Nuclear Submarines. The Navy insisted on this design, because it can breed Plutonium 239 out of U238 for making nuclear bombs. Commercial nuclear power plants used this design because it was proven and there was, already, a powerful lobby for Light Water Reactors in Washington. It was assumed that Light Water Reactors would be replaced by Fast Breeder Reactors, but those proved impractical.

      Liquid Fluoride Thorium Reactors were explored at Oakridge Labs in the 1960s. But, they were neglected because they could not be used for making bombs. A LFTR has many advantages which Light Water reactors do not have. Safety is just one.

      A LFTR converts Thorium 232, which the world has huge supplies of, into U233 and then burns it up for power. A LFTR, cannot, in practice, be used for making bombs. When U233 is bred from Thorium, a tiny amount of U232 is also created which is very hard to remove. U232 decays into a high gamma radiation emitter which is highly detectible from space and quickly degrades electronics.

      http://blogs.fas.org/sciencewonk/2012/08/back-to-thorium-the-thorium-cycle-and-non-proliferation/

      A LFTR’s core, mostly composed of Lithium and Barium fluoride salts, is a fluid at 1600 degrees F. In normal operation, if a LFTR started to overheat, the molten salts would expand. The expanded fluid would have fewer atoms of uranium exposed to neutrons, so the reaction would slow. No runaway reaction is possible.

      If all cooling ceased, such as from a break in the lines to the generators, a freeze plug in the bottom of the vessel, cooled by air, melts. This causes the fluid to fall into a wide area which cannot support a reacton. After the repair is made, the fluid is pumped up into the core and reaction restarts. The molten salt reactors at Oakridge were shut down over the weekend and were started up on Monday. A Light Water Reactor takes weeks to safely restart.

      A LFTR does not use water as a moderator, so hydrogen explosions are
      impossible. The LFTR can be run at much higher temperatures than a Light Water Reactor. That gives much higher efficiencies; 50% of the heat can be turned into electricity, using Braden Gas turbines. More efficiency means that less waste is created.

      No technology is 100% safe, but LFTR would be safer and cleaner than our current coal, oil and Natural Gas power generating systems. Obsolete 40 year old nuclear plants should be replace with much safer designs.

      • An LFTR is a concept, not a design. Designs have all of the materials science issues ironed out. Designs have all the systems issues evaluated and have elaborate failure chains constructed. Designs have prototype reactors with tens of thousands of hours of operational experience at production power levels.

        It would be great to live in a world–as we did back in the late fifties and early sixties–where it was acceptable to deploy these kinds of immature technologies and acquire experience in an operational setting. But we don’t–mostly because we learned that acquiring operational experience with fission nukes led to some unpleasant consequences. Since we don’t live in that world any more, I hope that we can one day deploy LFTRs and a whole bunch of other advanced reactor concepts. But we’re talking decades.

        Meanwhile, I’d be overjoyed if we could actually deploy some generation III+ designs in the US and start driving down their construction, certification, and operational costs. Nuclear power is expensive because people are scared of it, and their fear comes out in planning, licensing, and oversight costs. There are ways to reduce those costs, but they require scaling the deployments. But we’re never going to do that if we keep hanging our hopes on technologies that are probably half a century away. If I want to hang my hopes on something like that, I’d rather push for fusion. It’s nowhere near practical yet, but at least it has the advantage of being safe to deploy in an immature form when we do figure out how to make it practical.

        • louis_wheeler

          You have a nonstandard definition for “Design.” It isn’t the same as blueprint. A design can be a draft or a preliminary sketch. If we had used your definition, America would have never gotten the atomic bomb. We would been forced to wait until all the details were ironed out. No trial and error would have been allowed. People died building the bomb, because mistakes were made. But, people could learn from their mistakes, back then, and go on to success. Not now.

          Are there parts missing from a construction Blueprint of a LFTR plant? Yes. Has there been a prototype built to prove that the concept is commercial? No. The government wouldn’t spend the money to prototype it in 1975. They bet on fast breeder reactors, instead, and lost.

          Most of what is missing from the construction a LFTR plant has to do with its life expectancy. What I’ve heard is that $200 million dollars is needed for experiments on minor parts and about a billion for a one gigawatt prototype. Most of a completed design has been worked out, decades ago. We must work on the last 10 to 15% to make it
          practical.

          It seems clear that China is bent on finishing a LFTR prototype. Their time table seems to be less than a decade.

          Of course, there is no hope of doing anything nuclear in the US; the precautionary principle of our Anti-Nuke activist’s guarantee that. Besides, the utilities have given up, because the US government is so hostile to business. The Obama administration dragged it’s feet on the nuclear power plants which were approved by Bush.

          Fusion is a pipe dream; the problems seem insurmountable. Every experiment on it has failed.

          Our current reactors are awful. I have my doubts about the generation three and four types. Why complain about our generation one and two nuclear plants if the government won’t allow us to replace them? Just shutting them down is no solution. That instantly cuts everyone’s standard of living.

          What’s your alternative to nuclear power? Obama is bent on phasing out anything using fossil energy, whether we have a replacement for it or not.

          I moved out of LA and to a place that uses geothermal energy, partly, due to the threat of power outages. California imports 30 percent of it’s electricity. The state legislature won’t allow the construction of new power plants or allow the exploitation of the Monterrey Oil Shale deposit.

          This Luddite attitude will lead to catastrophic failure. I didn’t want to starve, freezing in the dark, so I left.

          • You seem to think that I’m anti-nuclear; nothing could be further from the truth. I would love to have nuclear providing most of base load power. I would love to have an aggressive LFTR program. But LFTR isn’t a panacea. There’s no such thing as a panacea in the nuclear industry, because people are scared of it. It would be great if people were less scared and therefore more relaxed about allowing experimental reactors to run at commercial scale, but that’s not going to happen.

            My definition of design is just fine. If you think that the Manhattan Project didn’t design their systems, you’re nuts. Yes, they did a lot of experimentation, but building a bomb is a completely different design process than building a power plant. But even to get a reliable nuclear weapon (the ultimate one-shot product), they spent $26B in 2010 dollars and employed 130,000 people to get there. They didn’t do that without understanding the physics, the materials, or systems. I’m sure you can swap expense for time on LFTR, but when you do that, you have to expect its development to take decades–which was my point in the first place.

            Again, I’m a big supporter of LFTR. But I’m a bigger supporter of a thriving nuclear industry, because there just might be something to all this climate change stuff and renewables simply aren’t going to have the power density necessary to run a modern civilization. And even if we don’t suddenly need to do a crash CO2 reduction exercise, nuclear energy allows us to invest in energy-intensive industries that will dramatically improve our standard of living. To do any of that, we need to get real nuclear deployments going now, and that’s not going to happen without lots of gen III+ deployments.

          • louis_wheeler

            There are no panaceas in technology; there are always tradeoffs. Compared to the problems of burning coal for electricity, then LFTR comes rather close to being one.

            Yes, Anti Nuke activists prevent any solution in America, at least, until times get truly desperate. That desperation is coming because the Luddites in the White House are endeavoring to shut down every “non-Green” energy source. Most Green’s don’t understand the pain of regressing in technology, or they think that other people will pay the price for them.

            I merely object to going back to using Kerosine for lighting. I moved out of LA to escape the effects of the next major depression, which is barely starting. Billions of people would need to die to go back to using kerosine technology and I’m too kind-hearted to like that.

            America is fortunate that wildcatters learned how to successfully frack shale. Otherwise, America would be in a world of hurt, now. We might be energy self sufficient, several years from now, when the Dollar goes belly up as the world’s reserve currency.

            Don’t get too hung up on words, like design. I gave you a definition right out of Webster’s New World dictionary. You need more precision in your word choice. You misread design as meaning Blueprint, layout, commercially proven technology, but I didn’t say that. The dictionary didn’t say that.

            The Manhattan Project was a “work in process.” They didn’t even know, for sure, that the bombs could explode. They made both U235 and Plutonium bombs, because they weren’t sure which one would work, but both did.

            You are correct that we don’t have the national will to solve this LFTR problem, because the Luddites are in control.

            I never thought that you were anti-nuclear. I merely disagreed with your statement: “… but existing reactors are already more-or-less fail-safe by design. They’re not passively fail-safe, but the failure chains all have very low probabilities. Still, they’re complex systems, just like the next generation of reactors will be, and therefore subject to unanticipated failure modes.”

            LFTR is passively fail safe and I gave both methods which keep it stable. But, it isn’t commercially proven and I said that. LFTR will come because the science is proven, but the technical details are not. The world will buy LFTR from China and you are probably fine with that.

            I am too, because China is a economic basket case. It is incredibly corrupt and counter productive, like all Fascist societies are. The US is Fascist, too, but we are likely to correct our ways. China doesn’t have our 300+ year history of personal freedom. The American authoritarians have been cracking down only for the last two or three decades. The Environmentalists are an aberration; they will destroy themselves.

            Ah! You sound like a pro-Nuke Green. That says that you have a working brain, at least.

            I think you are wrong about Anthropogenic Global Warming, because the evidence doesn’t support the theory — the earth warmed between 1850 and 1945 while the mass of the CO2 increased after the war.

            There have been plenty of epochs when the Earth was warmer than now. Google “Medieval Warm Period” and “Holocene Optimum.” We have evidence of how high the oceans were in both eras and life still flourished. There was no need to panic and there isn’t one now.

            Even so, The world needs to move on from fossil fuels as soon as there are cheap enough alternatives. Cheap enough nuclear technology could make cheap alternatives for gasoline and diesel. But, the market, not the government, needs to decide that.

        • EngineerPoet

          Designs have all of the materials science issues ironed out. Designs
          have all the systems issues evaluated and have elaborate failure chains
          constructed. Designs have prototype reactors with tens of thousands of
          hours of operational experience at production power levels.

          You mean, like the MSRE’s 17655 hours critical?

          One of the MSRE’s goals was to test materials.  IIUC, a problem with tellurium corrosion and cracking at grain boundaries in Hastelloy N was discovered, and the metallurgists found that a bit of titanium in the alloy would fix it.

          • Yeah, like that, only at something approaching full power for a commercially viable design. 7.4 MW(t) isn’t exactly there. MSRE was a good first step, but I doubt that it generated enough data to go straight to a pilot plant. So we’re at least two iterations from something that the NRC would even consider licensing. Call it 15 years, best case. That’s the Gen IV timeframe.

          • EngineerPoet

            MSRE was designed for 10 MW(th), but the radiator could only dissipate 7.4 MW and the negative temperature coefficient enforced that rather firmly.

            I’d say that the successful MSRE run merits a 50-100 MW(th) unit.  A number of the SMR concepts are also around 100 MW(th), so the “pilot scale” reactor could also serve as the prototype unit for a full production series for industrial process heat, thermochemical processing and all the other things besides electricity that we need to de-carbonize.

            If we were serious about this, we shouldn’t have any difficulty producing a working unit in just a few years.  The problem is the getting serious part.

          • A little spreadsheet-smithing against Wikipedia’s list of nuclear power plants shows that the average unit is 940 MW(e), which is–what?–2 GW(t), assuming 45% turbine efficiency? (I was actually a little surprised at how big those numbers are–I was expecting something closer to half a gigawatt electric.)

            I’d buy a 500 MW(t) plant as a viable pilot, but I’m not sure 100 MW(t) is gonna do it. That’s a scale-up factor of 50, which is a pretty big single step. Even a scale-up of 10 to 100 MW(t) is a pretty big step.

            If we were serious, I’d say we could do it in 15 years. If we were panic-stricken, we could do it in 5 years, because the risk profile would be adjusted. But the only way I can think of us becoming panic-stricken is if somebody proves that we’re going to heat up 10 degrees C by 2050 unless we chop carbon emissions to 20% of today’s value–and that’s not going to happen. (I suppose a general Middle East War and a simultaneous magnitude 8.0 quake caused by fracking could do it, too, but that’s about as likely as the climate change scenario.)

          • EngineerPoet

            the average unit is 940 MW(e), which is–what?–2 GW(t), assuming 45% turbine efficiency?

            The thermal efficiency of LWRs is around 33% (the AP1000 is 1110/3415 = 0.325).  Figure 3+ GW(t) for a 1 GW(e) plant.

            I’d buy a 500 MW(t) plant as a viable pilot, but I’m not sure 100 MW(t) is gonna do it.

            100 MW(t) is only ten times the design rating of the MSRE, but it’s close to the thermal rating of the NuScale (160 MW(t)).  There are plenty of industrial users with heat requirements far below 500 MW(t), and if they came out sufficiently cheap you could just assemble a bunch of them.

            The EIA’s annual energy report data for 2011 says that the electric power sector consumed 18,035 trillion BTU of coal (table 2.6) and generated 1687.9 billion kWh from it (table 8.2c).  That’s an average heat rate of 10,685 BTU/kWh or a thermal efficiency of 32%.  If you manufactured such units as drop-in steam supplies to retrofit existing plants, a trio would feed a 100 MW(e) powerplant.  The MSRE was specced to generate 10 MW(t) from a core volume of all of 2 cubic meters, so if you kept the same power density you’d have a total core volume of just 60 m³ or smaller than a cube 4 meters on a side.  Pumps and shielding and steam generators and stuff would multiply that several times, but it would still be a very small building to replace a stories-tall boiler and exhaust stack.  Small is beautiful, no?

          • Small is beautiful if the economics are right, but I have to believe that there are compelling reasons why big is the norm. I think it would be great to see nukes move down-scale, and there are a fair number of Gen III+ designs that do just that. Maybe when we get some online experience with those designs, and the NRC’s pre-approved design program gets the kinks ironed out of it, we’ll start to see smaller deployments. That in turn could shorten the work needed to do LFTRs, but to a large extent all you’re doing is trading one kind of qualification for another. You’re still looking at 15-25 years.

            Again, I’m not saying that we couldn’t cut a lot of corners and do LFTR much quicker, but you would have to a) convince the public to accept a more aggressive risk profile and b) convince investors that that risk profile gave them returns large enough to offset that risk. The facts on the ground have to change a lot before that happens.

            You’re right on the Rankine cycle efficiency–I’d derived the efficiency number through the “rectal extraction” method. You can get close to 50% on a Brayton cycle gas turbine, however. That’s yet another advantage of LFTRs.

          • EngineerPoet

            I’m not sure why big became the norm.  Fewer units reduce the number of per-unit costs, and in a regulatory environment where the regulator bills per-reactor that’s going to loom large in anyone’s calculation.  mPower and NuScale turn this on its head, proposing to cut construction schedules and minimize interest-rate and market uncertainty by adding generation more quickly and in smaller increments with pre-fabricated reactor modules ready to produce steam when they arrive on the site.  The NRC would have to sign off on regulatory changes to make the personel costs work, though.

            I’m rooting for the supercritical CO2 turbine.  The specific heat of a fluid around the critical point is far more than in an ideal gas, and that minimizes the temperature rise in compression and the resulting back-work.  That’s also a good match for a MSR, or for a gas-cooled solid-fuel reactor.

      • EngineerPoet

        The world uses Light Water Reactors, because the US military wanted them to power Nuclear Submarines. The Navy insisted on this design, because it can breed Plutonium 239 out of U238 for making nuclear bombs.

        This is a piece of anti-nuke propaganda.  It could hardly be further from the truth.

        Any uranium-fuel reactor will breed some plutonium.  Weapons-grade Pu requires that 93% or more be Pu-239, and especially requires that Pu-240 concentrations be kept low.  This requires special processing, which PWRs and BWRs cannot do while fulfilling their main mission.  The US produced its weapons Pu in graphite-moderated reactors, like the Hanford N reactor.  Un-enriched uranium slugs were irradiated for a few weeks to get about 0.1% Np-239, then left to “cool” for several months before reprocessing.

        The Pu from power reactors has as much as 25% Pu-240.  It’s essentially useless for weapons, and no weapons proliferator has even tried using spent LWR fuel for bombs.

        The fuel in US naval reactors is essentially weapons-grade uranium when it goes in.  It isn’t when it comes out.

        • louis_wheeler

          No, not true. You may be thinking that Light Water reactors are the only way
          to produce fission. But, there are ways which don’t make materials for bombs. A reactor burning U233 would not necessarily do so.

          It depends on the design of the reactor. In essence, U238 must be exposed to neutrons to produce Plutonium 239. The Navy did not think this was a design flaw.

          Spent fuel rods from a commercial Light Water reactor are typically 0.7 % U235
          and 4% Plutonium 239 after being in a reactor for 18 months. This design was copied from the Navy, but details could have varied, because there are no space constraints like in a submarine. A power plant could lower the concentration.

          The Light Water Reactors were assumed to be temporary; fast breeder reactors were assumed to take over. But, no one could get them to work.

          • EngineerPoet

            You may be thinking that Light Water reactors are the only way
            to produce fission.

            Your reading ability isn’t so good.  I said straight out that the Hanford N reactor was moderated by graphite, not water.  Light water absorbs too many neutrons to sustain a chain reaction in natural uranium; nuclear graphite does not.

            It depends on the design of the reactor. In essence, U238 must be exposed to neutrons to produce Plutonium 239.

            All uranium derived from natural uranium retains some U-238.  Weapons-grade material may reduce it to a few percent, but doesn’t get rid of it entirely.

            Spent fuel rods from a commercial Light Water reactor are typically 0.7 % U235 and 4% Plutonium 239 after being in a reactor for 18 months.

            Since it’s obvious that you have no idea what you’re talking about, I’ll just send you straight to the source on spent nuclear fuel:

            http://www.world-nuclear.org/info/Nuclear-Fuel-Cycle/Fuel-Recycling/Plutonium/

            Fresh LWR uranium is about 96% U-238, 4% U-235.  Spent LWR fuel is about 1% each U-235 and various isotopes of Pu (broken down in the table by reactor type).  In that table you’ll find two LWR types, one HWR type (CANDU), and two graphite-moderated types (AGR and Magnox).  All of them produce fission quite successfully.

          • louis_wheeler

            “I said straight out that the Hanford N reactor was moderated by graphite, not water.”

            You mean like Chernobyl or the Chicago Pile-1? Obsolete junk like that? LOL

            Sure I knew what you were talking about, but it was so old fashioned and dangerous that I neglected to mention it. None of that has anything to do with our current problems with Light Water Reactors and their waste.

            You do remember what the topic of this thread is, don’t you? “The Cost of Nixing Nuclear in Japan.”

            Yes, there have been reactors with positive temperature coefficients and no means shutting them down without inserting cadmium rods. They are bad designs, very unstable designs. Chernobyl went into meltdown within a minute of having its cooling water turned off. It had no natural fail safes.

            “All uranium derived from natural uranium retains some U-238.”

            Natural uranium is 99.93% U238. Reactors use between 0.5 % and 3% U235. (Your source says up to 4% and I’ll accept that.) You can run a reactor without U238 at all, but is more expensive.

            “Fresh LWR uranium is about 96% U-238, 4% U-235. Spent LWR fuel is about 1% each U-235 and various isotopes of Pu (broken down in the table by reactor type).”

            It depends on where you get your info. I listed my source. Of course, it is biased toward LFTR. I don’t pretend to be an expert. But, this stuff isn’t rocket science. It would be simple if there weren’t so much disinformation out about nuclear.

  • Pete

    When will Germany wise up?

  • submandave

    “But when nuclear goes wrong, it can go very wrong.”

    As RM details below, though, the historical evidence does not support an assumption of worst-case for nuclear power. Very rarely has nuclear power gone “very” wrong. While much is made of the potential long-term health risks of nuclear by-products or contamination, something hardly never looked at is a comparison of these risks against known long-term (and even near-term) health effects of competing energy technologies to accurately appraise relative dangers.

    The greatest thing nuclear power has against it is that you can’t see or feel radiation, only its effects. The fact that one can see and feel something like smoke or solid waste gives the false impression that this danger is controllable, while radiation’s invisibility makes it seem more sinister.

  • Dantes

    We have yet to come up with a solution for nuclear waste besides on site storage, which is not a great idea.

  • GlobalTrvlr

    A few observations. Fukushima, a >40 year old plant, right on a seismic fault, did not fail due to the earthquake- it failed due to the backup diesel plant being too close to the water and too close to each other for redundancy sake.

    The other thing that is not often discussed is that you CANNOT run a grid off of solar and wind. You have to have large inertia base load plants in order to be able to start large motors for industry, withstand large power swings up and down due to lost loads or swinging loads to another system in an emergency. Running a grid off of solar and wind would be like trying to fly a 787 with a few hundred Beechcraft propellor engines strapped to it.

    The increased costs in places that have gotten up to the mid teens or a little more, such as Germany, Spain, Denmark comes not just from the increase $/KW first costs of the generation. Very quietly so as not to upset the subsidy spigot or get on the wrong side of the greens, the utilities have had to install massive synchronous condensers that help to provide the inertia and system stability, and massive transmission interties to other countries in order to have backup when the winds aren’t blowing. These additions have cost in the low $billions per country, which never gets factored in when people are trying to prove the economic trade-offs of wind or solar.

© The American Interest LLC 2005-2017 About Us Masthead Submissions Advertise Customer Service
We are a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for us to earn fees by linking to Amazon.com and affiliated sites.