In particular:

• While not quite perfect, the Generation II light water nuclear power reactors in operation today pose nowhere near the hazard commonly believed. The two accidents (Chernobyl’s was not a Gen II LWR) not withstanding, they remain by far our safest energy source (measured by deaths per TWh produced) available – renewables included.
• Generation III+ light water reactors currently under construction are much simpler and even safer yet by several orders of magnitude, offering operator-free automatic passive shutdown assurance for three to seven days without additional intervention. If necessary, emergency cooling water may be easily replaced from external sources, and emergency circulation is via passive convection.
• Proposed Generation IV fast neutron reactors are “inherently safe”, requiring no operator intervention at all after automatic passive shutdown. They also may burn their uranium fuel sixty to one hundred times more efficiently than Generation II and III light-water reactors, may use current stockpiles of spent nuclear fuel, plutonium, and depleted uranium as fuel, and in the process reduce the final lifetime of high-level waste from the current 170,000 years to approximately 300.
• Our current supply of spent nuclear fuel is sufficient to provide the United States electric power for 100 years at present consumption rates, using such reactors. Our depleted uranium stockpile could power us for 900 more. It would be shortsighted to bury that much useful energy anyplace we can’t retrieve it.
• It is highly unlikely renewable sources alone will be able to satisfy the world’s energy needs in time to forestall the impending climate catastrophe. It is imperative our universities, utilities, and National Laboratories conclude modeling studies and cost-benefit analysis that will show just what combinations of all low-carbon energy sources, in conjunction with judicious carbon taxation or cap-and-trade, will best meet our needs for most rapid abatement of carbon emission, while maintaining economically competitive prices for energy.
• Commercial nuclear power is not going to go much further in the United States – not at the scale needed to meaningfully combat climate change – without a comprehensive National Plan for Nuclear Waste and public understanding of how it will work. The President’s Blue Ribbon Commission on America’s Nuclear Future issued its Final Report in January 2012. It deals with waste issues that include – but go far beyond – the spent nuclear fuel from commercial power reactors. We quote some of the report’s key conclusions. We must develop a National Plan.
• We tactfully suggest present wind and solar Production Tax Credits be modified or replaced in favor of something that might actually result in decreased global emissions of greenhouse gas.

I liked the film. Mr. Stone began research in 2009 and Pandora’s Promise was well underway at the time of the Fukushima disaster of May 2011. I was curious to see how he would deal with Fukushima and Three Mile Island and Chernobyl in a short documentary that advocates questioning and rethinking of many people’s views on nuclear electric power generation. I was not dissappointed: Pandora’s Promise confronts these issues head on, and continues with a brief history of commercial nuclear power in the context of cold war naval propulsion and bomb production, before biting into the meat of fast neutron reactors and how the early nuclear pioneers originally envisioned a future of abundant, safe, and low-cost energy.

Contents

List of Figures

List of Tables

Pandora’s technical contributors included Tom Blees and Barry Brook.

Professor Barry Brook holds the Sir Hubert Wilkins Chair of Climate Change at the School of Earth and Environmental Sciences at the University of Adelaide, is co-author of Why vs Why: Nuclear Power, and hosts the respected Brave New Climate blog. His site includes excerpts from Dr. Till’s book Plentiful Energy: the Story of the Integral Fast Reactor.

Tom Blees is president of the The Science Council for Global Initiatives, and is member of the selection committee for Russia’s Global Energy Prize for energy research. He is also on the board of The World Energy Forum, a UN and World Bank-affiliated organization whose goal is providing abundant energy for all mankind, and author of Prescription for the Planet. Tom has a 3-part YouTube video on Integral Fast Reactors.

Pandora’s Promise does not lack for technical expertise. After the spectacular opening sequence showing the dreams of Fukushima Daiichi going up in smoke, Pandora back traces sixty years to the dawn of the nuclear power age a few miles south of Arco, Idaho, with explanation of the EBR-I project by project engineer Leonard Koch. (See Koch: Remembering the EBR-I.)

The original “fast breeder” concept was developed by Enrico Fermi during the Manhatten Project of the mid 1940’s. Fermi is famous for having built the world’s first atomic piles at the University of Chicago. These were (relatively) primitive affairs that served as testbeds for research into reactor stability and the physics and chemistry of uranium transmutation to plutonium.

Naturally occuring uranium consists of two isotopes: it is 99.3% stable U-238 and 0.7% active U-235. The latter may be split by a slow “thermal” neutron to release two or three neutrons, two daughter decay nuclei of roughly similar mass, and about 200 Mev energy. The energy comes from the electrostatic repulsion of the two daughters, and some is distributed to the neutrons with a mean energy of about 2 Mev each. These are “fast” neutrons, and are too energetic to react readily with other heavy nuclei. “Not readily”, however, is not the same as “not at all”. There are two processes competing in a nuclear reactor core. One is fission, the second is neutron capture. Fission releases the energy for which the reactor was built. Neutron capture leads to transmutation to an element of higher atomic number and mass, and is the mechanism by which plutonium ${\mathrm{Pu}}_{94}^{239}$ is bred from ${\mathrm{U}}_{92}^{238}$.

• Fissile fuels may be fissioned (split) by slow “thermal” neutrons. The important fissile fuels are Uranium-233, Uranium-235, and Plutonium-239.
• Fissionable fuels may be fissioned by fast neutrons. U-238 may be used as a fissionable fuel in a fast neutron reactor. But fast neutrons may fission all other actinides as well. (This might be a good time to bookmark your local Periodic Table of the Elements, and keep it handy.)
• Fertile fuels do not fission directly, but instead first capture one of the neutrons released by a prior fission process to breed a heavier isotope of the same element, then undergo a series of two beta decays to form a fissile isotope of an element with atomic weight one greater and atomic number two greater than the original fertile element. The important fertile fuels are Thorium-232 (essentially all of mined Thorium) and Uranium-238 (99.3% of mined Uranium), which respectively breed Uranium-233 and Plutonium-239.

All commercial power reactors breed some fuel this way, but today’s solid oxide fueled light water moderated thermal reactors don’t breed enough new fissile Pu-239 from the fertile U-238 to make up for the amount of fissile U-235 they consume. This is because today’s reactors use the mined uranium-plutonium cycle in a moderated neutron spectrum. A faster spectrum is required to produce more Pu-239 than U-235 consumed. Nonetheless, a typical light-water reactor obtains about one-third of its energy from converted Plutonium.

Which in the overall scheme of things still isn’t very much. A light-water reactor’s primary fuel is U-235, enhanced about another 33% by bred plutonium. But the primary U-235 constituted only 0.7% of the original mined uranium ore. The other 99.3% was U-238 that either remained unused in the fuel rods (except for the little bred to plutonium), or was stored as a highly pure depleted uranium by-product of the fuel refining process. Overall a light-water reactor consumes less than 1% of all the energy available in the original uranium ore.

There are several natural negative reaction feedbacks. First, as the water heats up, it becomes less efficient as a moderator, which tends to stabilize the reaction at a given temperature. If additional power is desired, the plant operators have to further withdraw the neutron-absorbing control rods to obtain sufficient neutrons to sustain a hotter, more power-producing reaction.

Second, if the reactor leaks and loses coolant, it also loses moderator and efficient thermal neutrons. The chain-reaction then quenches and the reactor stops.¹

Except a solid-oxide fueled light water reactor doesn’t stop completely. The chain reaction itself stops completely. But all those daughter products that have accumulated in the fuel rods – Neodymium-148, Caesium-137, Strontium-90 and the like – are themselves radioactive with half-lives from 10 to 40 years. Plus there’s a host of much shorter-lived isotopes, all of which continue to decay and release heat. Sufficent heat to melt the fuel rods if cooling water is not promptly resumed. This is what happened at Three Mile Island and Fukushima Daichi. Solid-oxide fueled light-water reactors are not very forgiving.

The unforgiving nature of solid-oxide fueled LWRs was recognized at the outset. Though conceptually simple, they nonetheless require layered safety systems to prevent catastrophic core failure. And even those don’t always work. Fukushima was particlarly tragic in this regard, because everything did work. The earthquake struck. The seismic sensors tripped and automatically shut down the reactors (inserted control rods) and simultaneously started the emergency generators to run the cooling pumps if the AC grid went offline. Which it did. Then the tsunami topped the Fukushima seawall and flooded the air intakes of the running diesel generators. And their backup backup batteries as well.

Oops.²

Meanwhile, there was a Cold War to be won. There was a pressing need for submarine superiority, and Admiral Rickover deemed light-water reactor designs the fastest way to get there. One can hardly argue his decision: USS Nautilus was authorized in 1951 and launched just three years later with a reactor unit built by Westinghouse. Light-water reactors subsequently became the backbone of the U.S. Nuclear Navy.³

President Eisenhower delivered his famous Atoms For Peace speech to the United Nations General Assembly in December 1953. The proposal had several purposes, not all of them realised. Of those that were, one of the more far-reaching concerned the open sale of U.S. commercial nuclear power technology to other nations in need ot low-cost, reliable electric power. In the 1950’s “commercial” meant light-water reactors, as that was what U.S. industry had gained most experience with through naval propulsion. As result, apart from one or two prototype sodium fast reactors,⁴ and a handful more (originally) dual-purpose graphite-moderated thermal reactors, all in the Soviet Union or its remnants, most of the world has adopted light-water technology for commercial nuclear power generation. (See The Enduring Effects of Atoms for Peace.)

But light-water reactors are simple. They present engineering challenges galore, but the fundamentals are simple. What piqued Enrico’s Fermi’s interest was that other 99% of the energy in mined uranium ore that goes unused in light-water reactors. That, and the long-lived transuranic actinides that accumulate as waste by-product of the Uranium-Plutonium chain reaction. Plutonium-239 is only the first, and itself may undergo neutron capture to form heavier plutonium isotopes plus those of Americium and Curium and some Californium as well. All these elements are long-lived radioisotopes that significantly complicate light-water reactor waste management.

(At this point the reader may wish to re-visit her local Periodic Table of the Elements, if she hasn’t done so already.)

As hinted above, a reactor that incorporates a fast neutron spectrum (fast reactor) has the possibility of addressing both issues. Fast neutrons can fission all the isotopes of all the actinide elements. But none of them capture a fast neutron readily. The challenge is to create a reactor environment with a high enough fast neutron flux density to fission the tough-to-crack non-fissile isotopes (i.e., most of them) while simultaneously preserving sufficent slow neutrons to split the fissile U-235 (if any) and Pu-239 and sustain the reaction without breeding more transuranic actinides than the fast neutrons can burn, You might wish to do all this and breed more Pu-239 from U-238 than you had U-235 (or Pu-239) to start with. Because if you can, you may then perpetuate the entire cycle on bred Pu-239 alone while maintaining heavier transuranic actinides at a low and constant level.

Breeding more Pu-239 from U-238 than your original U-235 requires more fast neutrons than are available in a light-water moderated reactor. Fermi’s approach was to replace the water coolant with low-melting liquid sodium metal which is largely (but not completely) transparent to neutrons and does not slow them down nearly as much as does water. There are several major advantages to this liquid-metal fast neutron arrangement, and a relatively minor drawback.

Sodium metal melts at 98 C (208 F) and boils at 883 C (1621 F) at atmospheric pressure. Specific heat capacity is roughly one-fourth that of water but heat transfer is higher over a temperature gradient due to the higher thermal conductivity of metals relative to water. Molten sodium is non-corrosive to steel, making it easy to work with inside a reactor. It does, of course spontaneously ignite in air producing a relatively cool flame and dense smoke, and reacts rather more spectactularly with water. Precautions must be taken to prevent both.

In some designs sodium metal may be replaced with a lead-bismuth eutectic (lowest melting mixture, abrevieated LBE), which melts at 124 C. Sodium has the advantage of being very inert and unreactive under internal reactor operating conditions. On the other hand, while being more corrosive to steel while molten, lead-bismuth is very inert after it cools down and solidifies, making it convenient for long fuel-life small modular reactors intended to be shipped fully fueled to a power generation site, run for perhaps sixty years, then be shipped back to the factory for refueling. Lead also boils at much higher temperature – 1670 C (3,040F) – opening the possibility of reactors designed to produce process heat for chemical refining and production, including hydrogen. The Soviet Union powered their Alfa-class submarines with LBE fast reactors throughout the cold war, and Russia leads the world in current lead-cooled fast reactor design. One might note that HTGRs (High Temperature Gas Reactors) may offer much the same SMR and process heat advantages, whilst simulataneously getting the lead out. (See EM2.)

There are several advantages liquid-metal cooled fast neutron reactors enjoy over their light-water thermal reactor counterparts:

6.1 Fuel utilization

The fast neutron spectrum allows essentially complete burn-up of all actinide elements and long-lived waste products. This includes all the U-238 in the original fuel and mined ore, the Pu-239 bred in the reaction, plus all the long-lived transuranic radioisotopes of Americium, Curium, and Californium. (And Protactinium, should Thorium be part of the fuel cycle.) All that remains are the relatively short-lived ($<$ 40 year half life) fission daughter products.

6.2 Waste management

The United States currently hosts some 70,000 tons of spent LWR fuel rods. The principle perceived problem with geological disposal is the relatively long half-lives of the transuranic minor actinides, necessitating safe disposal for time periods exceeding several tens of thousand and possibly several hundred thousand years before radiation levels drop beneath that of naturally occuring uranium. An oft-quoted figure is 170,000 years.

As explained by Gweneth Cravens in Pandora’s Promise, that 70,000 tons spent solid LWR fuel is enough to fill an entire football field to a depth of ten feet.⁵ In contrast, the U.S. currently emits some 5.3 billion metric tonnes CO2 gas into the atmosphere each year, of which over 2 billion tonnes are from electric power generation alone. ⁶ A single 560 MW LWR nuclear plant, such as the recently retired Kewaunee unit in Wisconsin, produces about 5 TWh of electric power each year (at 90% capacity factor). At 890 tonnes CO2/GWh for coal and 500 tonnes CO2/GWh for natural gas,⁷ this single plant saved an equivalent of 4.5 million (coal) or 2.5 million (NG) tonnes CO2 from being emitted into the atmosphere each year, or 180 (100) million tonnes over its forty-year lifetime.

Yet managing a tenth of one percent this weight in solid “waste” accumulated over the entire nearly sixty years of commercial power generation is beyond our present U.S. (political) capability. If there were enough of them, fast neutron breeder reactors could use that spent LWR fuel to provide the entire U.S. electric power requirement, at current levels, for 100 years. The total high-level waste resulting from such extended operation would amount to perhaps one-fifth the weight, and need to be stored for only 300 to 600 years before its radiation level subsided to beneath that of natural uranium.⁸ Many cities are older than that; the storage timescale would be reduced from geological to merely historic.⁹ (Update 12/22/2013: See figure 18).

For this reason some countries, notably Sweden, Canada, and France, have designed repositories to permit future recovery of the spent LWR fuel material should the need for fast breeder reactors be realized.¹⁰ U.S. reliance on light-water reactors with a uranium-oxide once-through fuel cycle results from combined political and technical decisions. U-236 buildup limits PUREX reprocessing (described below) to a single pass, but increases uranium fuel utilization by a not-insignificant 30%. However, the Ford administration felt the plutonium proliferation risks associated with PUREX outweighed this benefit, and decided commercial U.S. nuclear fuel utilization would be single-pass, after which the spent fuel should be either sequestered or disposed (buried). This policy was affirmed by the Carter administration and those that followed. Current NRC policy requires retrieval be possible for at least fifty years.¹¹ As highlighted in Pandora’s Promise, the U.S. government’s Integral Fast Reactor program at Argonne National Laboratory, intended to demonstrate commercial utilitization of spent light-water nuclear fuel as fast neutron reactor fuel, was cancelled in 1994¹² – with the predictable result that international collaboration on fast reactors and fuel cycles has since shifted to Russia.¹³

6.3 Fast Reactor Safety

There are several design differences that allow fast neutron reactors to be inherently safer than their light-water counterparts. Here we consider liquid metal-cooled fast reactors, defering discussion of High Temperature Gas Reactors for later.

6.3.1 Ambient pressure operation

The most popular light-water reactors are pressurized water reactors (PWR) that operate their primary water moderator–coolant at about 300 C, which requires several thousand psi to keep the water in its liquid state. That’s a lot of hot water under a lot of pressure, and requires a sturdy large-volume containment structure to contain the whole mess should the pressure vessel or its associated plumbing suddenly and inadvertently spring a massive leak. (To date none ever have, but one needs the containment just in case. It did prove useful at Fukushima.) In contrast liquid metal coolants boil at 883 C (sodium) and 1,670 C (lead-bisthmuth eutectic) at atmospheric pressure. Metal-cooled fast reactors operate at about 600 C and are pressurized with just a few psi inert gas to prevent oxidation of the coolant. A fast reactor vessel is not pressurized in the same sense as a PWR, and the safety containment structure may be of much smaller volume and lesser resistance to potential internal pressure stress.¹⁴

EBR-II and the Integral Fast Reactor were piped-pool designs in which the primary molten sodium coolant was held in a large “pool” tank with openings only at the top and smothered in a blanket of inert Argon gas. The reactor core and secondary loop heat exchanger and its associated piping were all lowered in through the top of the reactor vessel, into which there were no other openings. “A guard tank surrounded the primary tank with an annulus between them which allowed for detection of sodium leakage. The guard tank was in turn surrounded by concrete shielding which acted as a final containment vessel. Were leakage to occur in both the primary and guard tank, the core would not be uncovered and would be adequately cooled. An inert gas (argon) filled the space between the tanks and their cover.”¹⁵ This very simple unpressurized double-tank + concete smooth-wall design promoted safety from a loss-of-coolant accident (LoCA) by the simple expedient of providing no path or mechanism by which excessive primary coolant could be lost. Liquid sodium is itself non-corrosive and does not react with the steel vessel or metal fuel. In 30 years of operation there were no sodium leaks from the inner reactor vessel at EBR-II into the inert void between the guard tank.¹⁶

6.3.2 Ease of fuel rod replacement

Further, the unpressurized reactor vessel allows fuel rod repositioning and replacement on an on-going basis. In contrast, a light-water reactor must be shutdown for one or two months for fuel replacement. Fuel rods must be replaced, not because their nuclear fuel is depleted, but rather because many of the reaction decay products are themselves neutron absorbers that eventually poison and stop the chain reaction. At this point the fuel rods must be replaced, even though typically only 3% of their fissile fuel has been burned.¹⁷ In a light-water reactor this necessitates reactor shutdown and depressurization, and dissassembly of the pressure vessel cap, before the fuel rods can be repositioned or replaced. As consequence, fuel rod replacement is delayed until neutron poisoning becomes severe and daughter-decay a significant source of latent heat.¹⁸

Metal-cooled fast reactors operating at ambient pressure have no such constraint. Individual fuel rods may be withdrawn for reprocessing on an ongoing staggered basis, thus minimizing the total load of hot daughter products in the core. Further, the high thermal conductivity and high negative coefficient of reactivity of both the liquid metal coolant and the reactor core itself (due to thermal expansion of sodium coolant and the fuel rods and their support matrix) allows reactor designs that will passively limit their power output to low safe levels should power be lost to the secondary coolant loops that extract heat for the electric power generators. This “inherently safe” or “walkaway safe” property was demonstrated at the EBR-II reactor, and highlighted in Pandora’s Promise. These tests are described at Experimental Breeder Reactor II.

6.3.3 Ease of fuel rod fabrication

Fast reactors may have an additional distinction from light-water designs in the construction of their fuel rods. Commercial light-water reactors use fuel rods comprised of pellets of uranium oxide encased in zirconium cladding. Oxides are not good conductors of heat, so individual pellet temperatures tend to run high even during normal reactor operation, and decay heat can more readily lead to fuel element thermal damage in a loss-of-coolant situation. The Integral Fast Reactor design in particular promotes unclad solid metal fuel alloys of Uranuium-Plutonium-Zirconium rather than zirconium-clad oxides, which both greatly enhances the fuel’s thermal conductivity, and simplifies fuel rod construction. The latter is an especially important consideration in integral designs where the fuel rods are expected to be replaced at frequent intervals, reprocessed for decay-product removal, then refabricated to new rods, all done robotically on-site.

6.3.4 Ease of fuel rod reprocessing

The pyroprocessing methods to be used for fast reactor fuel recycling differ drastically from the older PUREX method used for conventional light-water fuels. PUREX (Plutonium-Uranium Extraction), dissolves the entire used fuel assembly in acid solution, then selectively extracts just the unburnt uranium and plutonium for reuse. The remaining solution contains both the long-lived transuranic actinide neutron capture products, and the short-lived reaction daughter decay products, lumped together in one place for disposal. PUREX, originally developed to extract plutonium for weapons, is expensive, has in the past produced copious amounts of liquid waste, and raises proliferation concerns in some quarters. (But PUREX is not completely without merit. See See Processing of Used Nuclear Fuel .)

Pandora’s Promise illustrated LWR waste volume reduction at a storage facility in Paris, France. However, the long-lived transuranic actinides still pose a radiation hazard for roughly 170,000 years, after which time any remaining Pu-239 and its U-235 decay product would (theoretically) be ripe for mining.

In constrast, in metal fuel pyroprocessing the used metal fuel is first simply melted. Volatile decay products such as xenon and iodine are recovered from the inert atmosphere, while metallic decay products (notably caesium and strontium) are removed by electrowinning (electroplating) techniques. The remaining molten fuel contains fissile uranium and plutonium, plus all the remaining transuranic actinides. Even solidified it is both thermally hot and highly radioactive, making an uninviting proliferation target. This metallic mixture is then recast into new fuel rods for reuse in the reactor, where all components – including the transuranics – are subject to fission by fast neutrons.¹⁹ The decay products that are removed mostly have half-lives less than 40 years. “There will be dramatic reductions in the toxicity of wastes to be disposed of. Best current estimates are that fast-reactor recycle will reduce net long term toxicity by something like two orders of magnitude. The final wastes can easily be tailored to an appropriate form for optimum security: long-lived isotopes in a metallic waste form (which can be highly corrosion resistant in the repository), shorter lived materials in ceramic waste forms. Radioactivity in a repository will reach background levels in less than 500 years.”²⁰

6.3.5 Proliferation resistance

Pyroprocessing cannot extract plutonium from spent nuclear fuel (SNF) with the chemical purity needed for bombs. Further processing would be required, even if the isotopic purity were acceptable, which it is not.²¹ Plutonium in a commerical power reactor accumulates about 25% of the unstable isotope Pu-240, which undergoes spontaneous fission with sufficient frequency to make bomb production highly impractical. Presence of thermally hot Pu-238 (used in radioisotope thermoelectric generators) further complicates matters, as heat alone can degrade and/or destabilize the chemical explosives intended to trigger the bomb.²² All in all, relative simplicity of uranium bomb construction and modern ultra-centrifuge designs make enrichment of readily-mined uranium a much more attractive target for proliferation than diversion of plutonium from spent commercial nuclear reactor fuel. That said, we aren’t talking horseshoes or hand grenades: impractical complexity notwithstanding, any reactor or fuel-cycle program must be subject to strict security and stringent international monitoring.

6.4 Cost

As only two fast reactor for commericial power generation – Russia’s BN-600 and France’s Phénix – have seen extended operation,²³ the actual capital cost and O&M (operations and management) estimates for such units contain some uncertainty. Their unpressurized design and resulting lack of need for pressure vessel and large containment structure should drastically reduce construction cost of the reactor assembly and containment themselves. On the other hand, whether done on-site (IFR) or at a central location (small modular reactors), fuel re-processing is expected to be considerably more expensive than our present once-through burn-it-and-bury-it approach to conventional LWR spent fuel (that has thus far not proved politically feasible in the U.S.). It is hoped the increased cost of fuel reprocessing will be largely offset by the cost saving of reactor vessel and containment building. But this will certainly not be the case for early production units, and while readily managable, the final decay waste must still be sequestered for a significant interval. (See Cost Comparison of IFR and Thermal Reactors.)

The Uranium-Plutonium (${\mathrm{U}}_{92}^{235}$ – ${\mathrm{U}}_{92}^{238}$ – ${\mathrm{Pu}}_{94}^{239}$) fuel cycle is not the only one suited to nuclear power reactors. The Thorium-Uranium (${\mathrm{Th}}_{90}^{232}$ – ${\mathrm{U}}_{92}^{233}$) cycle was proposed in 1950, and an experimental reactor to exploit an elegant liquid-fuel design was built and operated at Oak Ridge National Laboratoy in the 1960’s. Although the Thorium molten-salt fuel arrangement is in many respects much simpler than the solid metal fuel used in modern IFR designs, for a variety of political and weapons-related reasons the technology has recieved no government support since that early ORNL effort. It has, however, continued to attract theoretical attention from reactor designers worldwide, notably in China. Pandora’s Promise could only cover so much, and in any event proven liquid Thorium technology is much less advanced than Uranium-Plutonium fast breeders, of which over fifty have been built (both experimental and naval propulsion) and for which commercial power designs are ready for deployment today. Although moderated thermal thorium plants should produce but 2% the long-lived actinide waste of their uranium counterparts, thorium offers no particular advantage over uranium in most fast neutron reactor applications.²⁴ It may, however, provide unique properties in the Accelerator-Driven Systems (ADS) designs proposed for possible end stages of high-level waste burnup. In any event, choice of reactor technologies should be driven by the twin goals of minimizing both global carbon emissions and the lifetime of high-level nuclear waste.²⁵ A standard reference is Liquid Fluoride Thorium Reactors by Robert Hargraves and Ralph Moir. A brief history and description of thorium fuel cycles useful in various reactor designs (not just molten salt) is WNA’s Thorium. Thorium technology is chronicled at Energy From Thorium. See The IFR vs the LFTR: An Exchange for an informed expert discussion.

These issues are detailed in (a rather lengthy) WNA article Uranium and Depleted Uranium, from which we find world depleted uranium stock is about 1.5 million tonnes, increasing by 50,000 tonnes each year. Known Recoverable Uranium Resources were 5.3 Mt in 2011 at US $130/kg U. However:

The article goes on to explain the difficulties inherent in trying to make any kind of accurate assessment of the planet’s total recoverable land-based uranium. Regardless, these numbers are in rough agreement with those cited by Prof. David MacKay in “Sustainable” power from nuclear fission. MacKay likes to normalize his units to the amount of energy consumed per day per person, kWh/d/p. Europeans consume an average of 125 kWh/d, Americans, twice this (250 kWh/d). This is total energy consumption from all sources: electric, gasoline, home and industrial heat, residential as well as commercial.

MacKay assumes a 6 billion person planet and estimates that the 4.5 billion tons uranium in seawater could, if burned in fast breeder reactors, provide each person with 420 kWh/d for a thousand years.²⁶ Of course, since deep ocean water overturns only about once every 1600 years, it would not be possible to extract uranium this fast. But 420 kWh/d/p is at least three times as much total energy as any one person really needs, and after three thousand years one might hope humanity might unlock nuclear fusion or reduce its population to something less than a billion, or – who knows – even figure out how to use renewables effectively.

Lets depart the far future a moment and return to our present dillema.

In 2011 US electric consumption was 3,750,000 Gwh electric, for an average power of 430 GW electric or 1.3 TW thermal (at 33% efficiency).

The US has 63,000 tonnes of LWR spent fuel, increasing by 2 - 2.4 ktonnes annually Spent Fuel Storage.

The U.S. currently stores about 700,000 metric tons of depleted UF6, containing about 470,000 metric tons of uranium.

A fast reactor has a heavy metal burnup energy conversion of about 909 GW day/ton (GWd/t) or 2.5 GW year/ton (GWy/t or GWa/t). At current US average electric generation / consumption of 1.3 TW, this DU reserve would last us 2.5 GW a/t * 470,000 t / 1,300 GW = 900 years in fast reactors. The 63,000 tons of spent nuclear fuel could last another 100.

Then there is thorium, globally 3 to 4 times more abundant than uranium. And as with uranium, the earth’s total economically extractable thorium is not known. Estimates range up to 300 times the presently known 6 million tons, in which case thorium reactors could power 6 billion – should there still be that many of us – people at 120 kWh/d for 60,000 years.²⁷

We’ll return to these numbers a bit later,

9.1 Three Mile Island 1979

In 1979 what should have been a minor cooling malfunction ended up resulting in the partial core meltdown of Three Mile Island Unit 2 near Harrisburg, PA. The reactor was destroyed. Some radioactive gas was released several days after the initial accident. It was biologically inert (xenon-133, half-life 5 days) and there was not enough to cause any dose above background level. Chapter 6 of Prof. Bernard Cohen’s online book The Nuclear Energy Option gives an overview and history of safety considerations in commercial nuclear power, then details the accident at Three Mile Island. A more recent analysis is detailed at Three Mile Island Accident (March 2001, minor update Jan 2012):

9.2 Chernobyl 1986

Firstly, Chernobyl was most emphatically NOT a light-water reactor. Not in any conventional sense of the word. The RBMK-1000 was a water-cooled graphite-moderated boiling water design originally intended to simultaneously produce both electric power and weapons grade plutonium. This design would never have been implemented in the west. Nonetheless, it was implemented, and with catastrophic results. From Chernobyl Accident 1986 (Updated June 2013):

Which doesn’t mean Western reactor and safety specialists were not keenly interested in helping learn exactly what went wrong at Chernobyl Unit 4, and why.³³ Detailed discussions of the accident are given at both links. However, apart from the “apart from increased thyroid cancers” part, it looks, if not promising, then at least not as bad as one might have thought. Unless one were one of those with an “increased thyroid cancer”. So what is the epidemiology? What are the odds?

Briefly: 31 deaths as immediate result of the hydrogen explosions (3) and acute radiation poisoning to emergency responders (28). Of the affected civilian population (who received much lower radiation doses) there have been 9 confirmed thyroid cancer deaths. That does not tell the whole story. An additional 4,000 civilian cancer deaths might accumulate over the years as result of radiation received from Chernobyl. Or they might not: there is a very large civilian population, and it will be difficult (or impossible) to distinguish Chernobyl-caused cancers, should any occur, from the much larger number of “naturally” occuring background cancers. Doesn’t mean such cancers might not or will not happen, just that we may not be able to distinguish them if they do.³⁴

This is important stuff, as illustrated by Pandora’s charged exchanges with Physicians for Social Responsibility co-founder Dr. Helen Caldicott.

From Fact Sheet on Biological Effects of Radiation:

Similar estimates may be found at Radiation Information Network’s Radiation and Risk. The Belarus government and Pandora’s Promise producers appear to have their facts right. The 1 mSv/yr contamination radiation limit used by Belarus is one third the average U.S. background. As shown in Table 1 below, the resulting total 4 mSv/yr is also considerably less than background in many other parts of the world.

The 13 meter (42 ft) tsunami swept through Tokyo Electric Power Company’s woefully unprepared Daiichi Nuclear Power Station in Fukushima Prefecture, knocking out power lines, and disabling cooling systems and all emergency power. Some 160,000 people were ordered to evacute the vicinity from fear of radiation release from the stricken power reactors.

9.3.1 The Radiation Release

It was for good reason that Mr. Stone spent the Pandora’s Promise sequences measuring radiation levels around the striken Fukushima Daiichi plant. People are rightfully frightened whenever their local power plant explodes, and although “it has only happened twice”, it seems the most spectactular power plant suicides have been nuclear: Chernobyl in 1986 and again at Fukushima Daiichi in March 2011. Ionizing radiation is invisible, and we have an instinctive fear of hidden dangers we cannot see.

But here again Mr. Stone is absolutely correct: the actual danger posed to the public by the radiation released at Fukushima has been minimal. That does not mean there never was cause for alarm: the Japanese government ordered evacuation from the 10 km zone in the early morning of 12 March when it became clear there was emminent danger of hydrogen explosion at Unit 1, and expanded the zone to 20 km after the hydrogen buildup on the service floor did indeed explode.³⁵ There was no way of knowing beforehand how bad such explosions would be, nor the total radiation release.

We review the film’s look at Japanese government policy and present radiation levels.

The standard physical unit of radiation is the Becquerel (Bq), where 1 Bq corresponds to 1 disintegration per second. It does not account for the size of the radioactive mass undergoing decay, nor the energy or penetrability of the radiation. These effects are approximately accounted for in the Sievert (Sv), the standard measure for biological radiation effect. One Sv is actually fairly hazardous; one thousandth this amount, the millisievert (mSv) is usually used.

In the immediate aftermath of the tsunami, the Japanese governments decided to evacuate people living in areas that might receive radiation levels higher than 20 mSv/yr. A line had to be drawn somewhere. The following table gives some context:

Thus the Japanese government’s decision to evacuate people living in areas that might be higher than 20 mSv/yr and to not let them return (with some flexibility) until radiation drops beneath that level appears to be based on 20 mSv being roughly 10 times the 2 mSv/yr background radiation typical in Japan. As we saw in the Chernobyl section, background levels in the U.S. average about 3.1 mSv/yr. From the above table, 20 mSv/yr is less than half (40%) the background radiation level in some parts of Iran, India, and Europe. It is one eleventh the 220 mSv/yr public limit assigned by IAEA, and one twelveth the background level in Ramsar, Iran. The Pandora’s Promise measurements of current (2012) radiation levels near Fukeshima Daiichi were probably accurate, and some might suggest the Japanese government is being unduly cautious. But that is what they are elected for.

As of August 2013, there is concern about radiation being leaked to the sea. There is an estimate of 300 or 400 tonnes contaminated ground water each day, most of which is captured and placed in storage tanks. 300 tonnes in itself isn’t but a drop in the ocean and doesn’t mean much without the associated number of Curie or Bq of which radioisotopes. That information will undoubtedly be forthcoming, along with associated radiotoxicity.³⁶ From Concrete Actions to Address the Contaminated Water Issue at Fukushima Daiichi NPS (11 Sept 2013):

From Current Information on Radioactivity in Seawater as of 24 September 2013, the highest levels detected at reactor outlet ports was 5 Bq/l (combined Cs-137 and tritium) on Sept 16. See FAQ:Radiation from Fukushima for more on the sea-water controvery.

Update November 6: Tepco has opted for high-efficiency ion-exchange resin technology in its Advanced Liquid Processing System (ALPS), which will remove every radioisotope from the Fukushima-Daiichi storage tanks save tritium, which is part of the water itself. The left-over resin will be managed like other solid radioactive waste. Removing the tritium from the water molecules would be very hard, incredibly expensive, and totally unnecessary. At 6 kev tritium is one of the weakest beta emitters known, with half-life of 12 years. Tritium levels at Fukushima-Daiichi are estimated at 630 kBq/l and at that level are essentially harmless even if ingested. However to meet standards, after the other isotopes are separated, the remaining purified tritiated water will nonetheless be diluted 12 fold before release to the Pacific. Details at Fukushima Commentary. Mr. Corrice’s Fukushima and the Inevitable Tritium Controversy was added October 25, scroll down beneath “October 27, 2013” until you find it.

Let’s return to the situation inland:

Now 0.266 mSv/day is 97 mSv/y and 0.84 mSv/day is 307 mSv/y. From table 1 we see 50 mSv/y is natural background in several areas of Iran, India, and Europe, while 220 mSv/y is the long-term safe level for public exposure as set by IAEA (International Atomic Energy Agency). Namie was definitely punching above that limit, and the government was fully justified in being cautious and carefully locating the few hotspots before allowing evacuees to return. By March 2013 maximum Namie radiation appears to have subsided beneath 184 mSv/yr (Figure 2).

From Fukushima Accident 2011 (updated July 2013):

Summary: “There have been no harmful effects from radiation on local people, nor any doses approaching harmful levels. However, some 160,000 people were evacuated from their homes and only in 2012 were allowed limited return... As of October 2012, over 1000 disaster-related deaths that were not due to radiation-induced damage or to the earthquake or to the tsunami had been identified by the Reconstruction Agency, based on data for areas evacuated for no other reason than the nuclear accident. About 90% of deaths were for persons above 66 years of age. Of these, about 70% occurred within the first three months of the evacuations. (A similar number of deaths occurred among evacuees from tsunami- and earthquake-affected prefectures. These figures are additional to the 19,000 that died in the actual tsunami.)”³⁹

The casualties from the Fukushima evacuation itself exceeds 5% the total 19,000 who lost their lives across Japan as direct result of the earthquake and tsunami. This raises a few questions:

1. Given the extreme stress of the situation at the Fukushima-Daiichi nuclear station (and rapid deterioration of Unit 1 in particular), could the plant operators or government consultants have given the government a realistic upper bound to the radiation expected to be released by the impending hydrogen explosions?
2. Mass evacutions have mass casualty statistics of their own. By how much could the Fukushima casualties have been reduced had mass evacuation not been ordered 12 March 2011, in favor of orders for people to remain indoors until radiation teams could map the radiation release and assess more precisely just which hot-spots should be evacuated?
3. In hindsight the mass evacutation appears to have been unneccessary. Could it have been seen to be so in foresight?
4. Given that mass evacuation was ordered, by how much could the Fukushima casualties have been reduced had the majority of evacuees been allowed to return to their homes in a timely fashion?

9.3.2 What Happened

From the WNA’s description of the Fukushima Accident:

Whether or not it was ever necessary to evacuate that many people (or for how long) is a matter of ongoing debate. But faced with inadequate information and deteriorating conditions, such was the governmental perception at the time. See Fukushima’s Worst-Case Scenarios: Much of what you’ve heard about the nuclear accident is wrong, and a Roger Clifton comment on some findings in Mark Willacy’s Fukushima. ⁴¹

9.3.3 And Why

The nuclear consequence of the Great East Japan Earthquake of 2011 is well described in the above WNA link. It was not pretty. The World Nuclear Association article is dispassionate, objective, and does not assign blame. Of which there is plenty. Given the very public concern about the ongoing situation at Fukushima Daiichi, it is worth mentioning results of both Japan’s own investigation, and international review. A concise overview is given by Fukushima a disaster ’Made in Japan’, with the sordid details exposed at Why Fukushima Was Preventable.

Make no mistake: Fukushima Daiichi was not a random “Act of God” that could not be defended against. Yes, the GE-Hitachi boiling water reactors there were of a 1960’s design housed in the earliest Mark-I iteration of its containment structure. But that wasn’t the problem. By best estimation, although “The earthquake that preceded the tsunami exceeded the seismic design basis of the plant at units 2, 3, and 5. TEPCO and NISA (the Nuclear and Industrial Safety Agency) have stated that no critical safety-related equipment – such as emergency diesel generators, seawater pumps, and cooling systems – was damaged in the earthquake, although it seems that this claim cannot be conclusively verified until the plant can be inspected much more closely than is currently possible. Though the tsunami led to most – if not all – of the damage, the underestimation of the seismic hazard provides evidence of systemic problems in disaster prediction and management.”⁴³

It is telling that Tokyo Electric Power Company’s Fukushima Daiichi unit was not the nuclear power installation subject to the greatest tsunami energy after the Great East Japan Earthquake. That dubious distinction goes to Tohoku Electric’s Onagawa plant, 120 km (75 mi) to the north:⁴⁴

It’s not certain that is quite how it finally played out, as it seems there was a compromise whereby the plant retained its original site, but the seawall was increased to 14 m (46 ft) from its original 3.1. The tsunami of 11 March 2011 was 13 m at both Onagawa and Fukushima Daiichi. But the latter’s seawall was only 5.7 m:⁴⁵

Emphasis added. As we saw in section 9.2, even knowing it was possible, the USSR State Committee for the Supervision of Safety in Industry and Nuclear Power (SCSSINP) didn’t believe an explosive instability would ever actually arise at an operating RBMK-1000 power station.

9.3.4 U.S. Industry Response

Nuclear power plant operators across the globe closely watched Fukushima unfold. Thorough safety reviews were instigated at all 104 U.S. power reactors within days. The Nuclear Energy Institute summarize their findings in Safety: US Nuclear Energy Industry Responds to Accident in Japan. Key actions include:

• Acquiring additional safety backup systems, including diesel genrators and pumps.
• Increasing ability of nuclear facilities to remain safe in the face of extended loss of electric power.
• Improvements to ensure accessible and reliable hardened vents for Mark I and Mark II boiling water reactor containments, thus removing heat and maintaining of pressure control during an extended loss of off-site power. Of America’s 104 reactors, 31 have Mark I or Mark II containments. (Mark I containments were used at Fukushima. Failure of their emergency vent systems during extended station blackout allowed hydrogen to accumulate on the service floors of Units 1, 2, and 3. The hydrogen subsequently exploded, blowing the roofs off the containment buildings and releasing significant radiation off-site.)
• Improve plants’ ability to monitor water level and temperature in used fuel storage pools during an extended loss of electric power.
• Assess emergency communications and equipment to ensure that power is maintained during a large-scale natural event.

9.4 A Nuclear Reactor is not an Atomic Bomb

It isn’t. Alarmist propaganda not withstanding, a nuclear reactor cannot “Blow up like an atom bomb!!!” It just can’t. From Highly Enriched Uranium we find the minimum uranium enrichment for unmoderated criticality is 6%. Commerical light-water nuclear fuels are usually enriched to between 2% and 5% – but no higher for just this reason. From Fuel Fabrication Safety Considerations:

Of course, a nuclear reactor has to become critical in order to produce power. But criticality is designed in through use of a moderator, and the moderator – whatever it is – promotes fission so strongly that any combination of feul elements that should fall together in an extreme accident (as may have happened at Chernobyl after steam explosions destroyed their support matrix) will either overheat and slow the reaction, or violently blow themselves apart (with comparatively modest energy) long before a moderated critical mass can either accumulate or be substantially irradiated.

Here “long” is a relative term measuring a few to a few hundred microseconds.

Low uranium enrichment and moderation are just two of the myriad major differences between a nuclear reactor and an atomic bomb, but they alone are sufficient to guarantee the impossibility of a power reactor “blowing up like an atom bomb”.

Those who have read on the history of the Manhattan Project are well aware that atomic bombs do not just fall together. To get the high explosive yield obtained by fissioning a substantial fraction of the critical mass of uranium in the bomb requires careful fabrication of an explosive mechanism that can rapidly assemble pieces of the critical mass of metal into a critically dense configuration before energy from prompt fast neutron fissions can blow them back apart. The explosive mechanism may be either the “simple” gun arrangement used in uranium bombs, or the considerably more sophisticated implosion devices used with plutonium.

Considering just the uranium gun assembler, the arrangement traditionally consists of a thick doughnut “ring” of highly enriched weapons-grade uranium, and a spike of similar metal to be fired into the doughnut hole. Individually, neither the dougnut nor the spike make a critical mass, but together they do. The trick is to first get them together before the dougnut can reject the spike in a very low-yield explosion, then get them to stay together long enough for the chain reaction to sufficiently build to fission a substantial portion of the mass before it can similarly blow itself apart with very low yield.

These are bomb design considerations. Uranium has a small enough fast neutron fission cross-section that a gun mechanism may assemble the highly enriched parts before the reaction builds sufficiently to prevent insertion of the spike, after which the reaction might build rapidly enough to sufficiently irradiate the entire mass with a high fast neutron flux while it is still inertially confined. Its a good trick if you can do it, and as for moderation, to quote Werner Heisenberg “One can never make an explosive with slow neutrons, not even with the heavy water machine, as then the neutrons only go with thermal speed, with the result that the reaction is so slow that the thing explodes sooner, before the reaction is complete.”⁴⁷

The first explosion at Chernobyl was a steam explosion similar to a boiler explosion at a fossil fuel plant. Subsequent explosions are usually attributed either to steam, hydrogen generated by reaction of steam with zirconium feul-rod cladding, or perhaps carbon-monoxide and hydrogen from a hot-fuel driven reaction of graphite moderator with steam. Whatever their source, the explosions at Chernobyl pale to insignificance beside that of a real nuclear weapon. See Willacy’s Fukushima.

9.5.1 Risks of Nuclear Energy in Perspective

Prof Bernard Cohen was a radiation physicist. Not surprisingly, he had a few words to share in Understanding Risk

Fukushima isn’t going to change this any. Prof Cohen’s chapter is well-worth reading. Prof Wade Allison provides topical insight and analysis at Radiation and Reason – scroll to “Download Recent Articles” – and is currently active in the field. Also see Karen Street’s well documented and poignant essay Earthquake, Tsunami, and Nuclear Power in Japan: The Ocean of Light above the Ocean of Darkness.

It is important to realise the context of the origins of Gen III+ designs. Though its actual danger was minimal, the implications and effects of TMI on the nuclear industry were profound and cannot be understated. A good overview is given in Searching for Safety, part of the 1997 PBS series Nuclear Reaction: Why do Americans Fear Nuclear Power?. Though initially “cheaper than coal” in the 1970’s, the combined result of the industry’s re-thinking, and increased NRC regulations and public mistrust, was an over ten-fold increase in nuclear power plant costs over a period of just a few years.⁵¹

A brief overview of safety considerations in modern reactor design is given in Chapter 10 of Prof. Bernard Cohen’s The Nuclear Energy Option. The basic idea for both new pressurized (PWR) and boiling water (BWR) designs is to replace pump-based cooling circulation with natural convection, and to provide large amounts of gravity-flow and pre-pressurised redundant cooling water that may keep the reactor core and containment safely cool for at least three days in case of complete power loss such as happened at Fukushima. Allowance is made for ease-of-replacement of cooling water in such an emergency. (These designs predate Fukushima by over a decade.) A second consideration is simplification and standardization of design. Apart from their reactor vessels and cores, nuclear plants built during the 1960’s and ’70’s tended to each be a one-off, which eventually lead to regulatory and licensing delays, and massive cost over-runs during the 1980’s. The industry hopes pre-approved standard designs will promote predictable licensing times and construction costs.

In addition, most modern light-water designs are, like fast reactors, load following. Many existing reactors are as well. Rather than being able to supply just baseload, they may change power output rapidly enough to follow minute-to-minute changes in power demand from the grid – an increasingly important consideration in the face of deepening penetration of intermitent renewables. Such considerations are featured in each of the following:

1. AP1000: Advanced Passive 1000, Westinghouse. China has 12 units planned, four under construction with operation scheduled for 2014-15. U.S. has two under construction at Georgia Power’s Vogtle site and SCE&G’s V.C. Summer station. An AP1000 overview and design safety are given by Westinghouse in AP1000 Probabilistic Risk Assessment (PRA) estimates CDF (Core Damage Frequency) of $5\times 10^{-7}$ per plant per year, and LRF (Large Release Frequency) of $6\times 10^{-8}$ per plant per year (page 6).

   Load following: “The plant is designed to accept a step-load increase or decrease of 10 percent between 25 and 100 percent power without reactor trip or steam-dump system actuation... Further, the AP1000 is designed to accept a 100 percent load rejection from full power to house loads without a reactor trip or operation of the pressurizer or steam generator safety valves.” From Responding to System Demand, apparently quoting unspecified Westinghouse specs; these are allowed and exceed typical EU requirement for new plant.

2. ESBWR: Economically Simplified Boiling Water Reactor, GE-Hitachi. This large (1520 MWe) design awaits final NRC rule issuance.

   Next-generation nuclear energy: The ESBWR Focuses on simplicity and safety.

   Boiling Water Reactors are inherently load-following: see Nuclear Power Reactors and ESBWR Overview: ABWR Evolutionary Safety Improvement; ESBWR Improving Safety Passively page 3. But unlike GEH’s ABWR, ESBWR will not normally be expected to operate in a load-following mode because of its size.⁵²

3. EPR: European Pressurized Reactor, Areva. Two under construction in Finland and France, another two in China. European units face costly delays; Chinese units on schedule for operation in 2014.
4. ATMEA I: Areva-Mitsubishi. Also see ATMEA1. Passed first stage of Candian regulatory approval July 2013. Turkey has accepted a bid of four 1.1 GW units for $22 billion.

And a few Generation IV fast designs:

1. PRISM Integral Fast Reactor: GE-Hitachi. This is Gen IV, not III+, and a sodium cooled fast reactor, not light water. Link provides contains numerous references, as does Cost Comparison of IFR and Thermal Reactors – the last of a 4-part IFR series. No PRISM have yet been built, or are under construction. However, GEH have bid a two-unit system in a proposal to burn Britain’s accumulated plutonium surplus.
2. EM2 – GT-MHR. The EM2 Energy Multiplier Module is a modified version of General Atomic’s Gas-Turbine Modular Helium Reactor. It is a Gen IV Small Modular Reactor proposed as a candidate for the US DoE SMR funding program. Fuel cycle is similar to IFR – PRISM, though reprocessing would be done at a central factory location rather than onsite as befits the SMR concept. Interment time for final waste remains about 300 years. From General Atomics in contest for SMR funds:

   “The EM2 employs a 500 MWt, 265 MWe helium-cooled fast-neutron high-temperature reactor operating at 850C. This would be factory manufactured and transported to the plant site by truck. According to GA, the EM2 reactor would be fuelled with 20 tonnes of used PWR fuel or depleted uranium, plus 22 tonnes of uranium enriched to about 12% U-235 as the starter.

   “It is designed to operate for 30 years without requiring refuelling. Used fuel from the EM2 could be processed to remove fission products (about 4 tonnes) and the balance then recycled as fuel for subsequent cycles, each time topped up with four tonnes of used PWR fuel. The module also incorporates a truck-transportable high-speed gas turbine generator.”

   EM2 passive safety is obtained via high-temperature ceramic fuel cladding, together with low fuel density and high thermal conductivity of the support matrix, which combined may safely withstand residual decay heat even in case of complete coolant loss. High-pressure helium gas is the combined coolant, moderator, and working fluid for the turbine, and cannot itself become radioactive. As with all commercial reactors, EM2 operates with a negative thermal coefficient: heating the helium reduces its neutron moderation, slowing the reaction; substantial loss stops the reaction. Core expansion contributes: fast neutron reactors must operate at very high neutron flux just to maintain criticality; excessive thermal expansion of the core fuel rods and support matrix rapidly decreases the neutron density beneath criticality and stabilizes the reaction.

   The EM2 reactor module runs at an impressive 850 Deg C, resulting in turbine thermal efficiency of nearly 50% and making the system suitable for producing process heat for the chemical industry, including hydrogen production for transportation.

9.6.1 Load Following

From Responding to System Demand:

While true, the above assertion is not without qualification. The 3-part view of power generation gives a concise illustrated description of electric power variation, load demands, and resulting load following requirements. As do thermal fossil plants, commercial nuclear power plants usually have a design minimum power floor beneath which they typically may not operate without shutdown. (AP1000 may be an exception, but still has a nominal 25% load floor during normal operation.) Also, high initial finance and construction costs, and comparatively low fuel costs encourage baseload full-power operation of nuclear plant whenever possible. Cost per MWh generated, which is what you bill for, is minimized that way. Thus from the second ATMEA1 link under ”Load-follow requirements”, we find in daily operation ATMEA1 operates with a power floor of 25%, from which it is capable of rapid return to full power without notice at a rate of 5% per minute.

Atmea1’s load following figures are similar to currently operating French PWR’s,⁵³ which isn’t surprising as its a French design. Atmea1 is nominally a 1 GW plant, so 5% $P_r$ per minute corresponds to 50 MW per minute. 5% per minute is also a rate typical for fossil (e.g. natural gas) plants, which themselves have minimum partial loads of 30% - 50%, for which there are both emissions and efficiency penalties. Xcel Energy has some PC (Pulverized Coal) plants with a “hibernation” mode capable of turning down from 250 MW to 25 MW, still burning coal. (See Impact of Load Following on Power Plant Cost and Performance, Exhibit 1.) Cold start time for NGCC is given as “150 to 250 min with new technology, 100 starts/yr”

That last “100 starts/yr” limitation reflects a common weakness of all thermal plants: thermal stress. Rapid load fluctuations over a significant portion of a plant’s operating range can strain many elements of the heat transfer and steam system, including the steam generators in nuclear plants. Parts wear out faster and need to be replaced. This costs downtime and money. For this reason load following specifications often consider two modes of operation. The first is slow variation through large power swings in the intermediate load regime, where a plant might be specified as capable of swinging through 25% to 100% of rated power, provided such large swings are smoothly timed over an interval of perhaps half a day to follow diurnal load variation.

The second load following mode provides peaker capability through smaller but more rapid load swings. Here a plant may follow rapid fluctuations of perhaps $\pm$ 5% around 95% of its rated power, resulting in a 10% total fluctuation, as opposed to 75% in the smoothly varying intermediate regime, and thereby reduce the thermal strain. Different plants complement each other by operating in different modes.

Rapid load fluctuations on a fossil plant introduce thermal (fuel) inefficiency and emissions problems as well.

It should be noted that some older nuclear plants might not have load-following capability without upgrades to their control systems, which requires NRC approval. And while respectable and usually adequate, nuclear load following is not generally in the same class as hydro. Gas turbines – whether from natural gas or high-temperature Brayton cycle fast reactors – are better than steam. For more detailed discussion see Technical and Economic Aspects of Load Following with Nuclear Power Plants.

And nothing succeeds like “Success!!!” Perhaps the only serious shortcoming of Pandora’s Promise is its failure to make a cogent statement for the need for nuclear power. Mostly idle windmills aside, it is not an easy argument to make, particularly in the large and windy United States. Energy technology, grid engineering, and power economics are extremely complex, and there is a lot of optimistic misinformation being passed about, mostly sincerely, by those who sincerely believe wind, water, and solar (WWS) can do the job by themselves. Several questions:

1. Just what is the job that must be done?
2. Can renewables actually do it?
3. How fast, and at what cost?

The job that must be done is the avertment of climate catastrophe. That is the singular goal that must be kept firmly in mind. Time is of the essence, and so is cost. Money is not unlimited, and when available at all, comes in incremental installments. The lower the cost of abating each ton of emitted CO2 or methane, the more that may be avoided, sooner. Safety is certainly a crucial consideration, as are reliability and cost. But the goal is rapid curtailment of greenhouse gasses.

The numbers are daunting. As mentioned above, the U.S. currently emits 5.3 billion metric tons CO2 each year into the atmosphere, of which over 2 billion tons are from electric power generation.⁵⁴ Globally,

EIA estimates that about 10% of world marketed energy consumption is from renewable energy sources (hydropower, biomass, biofuels, wind, geothermal, and solar), with a projection of 14% by 2035. About 19% of world electricity generation is from renewable energy – 16% from hydro – with a projection of 23% in 2035.⁵⁶ That 14% (23%) renewable penetration by 2035 is woefully inadequate, and does not in itself directly address the problem of ever-increasing fossil fuel consumption:

And that mythical “2 Deg C Global Warming Limit”, representing an increase of atmospheric CO2 concentration from its pre-industrial value of 280 ppmv to 450 ppmv, is itself a recipe for disaster:

Out of time. We are currently at 400 ppmv CO2 and rising. 1 ppmv CO2 = 2.13 Gt carbon. ⁵⁸ 100 ppmv CO2 = 213 Gt carbon = 781 Gt CO2. At 32 Gt/y we’ll emit this amount and reach 500 ppm CO2 in 24.5 yrs. But the “450 Scenario” is half this: at our going rate 450 ppm CO2 will be reached in but 12 years. And we won’t be able to just shut it off like a faucet come 2025: we must start reductions now.⁵⁹

The largest and most pressing problem is coal. How do we (global we) most rapidly and economically replace coal-fired electric generation with low-carbon equivalent? Massive increase in cost or decrease in reliability will not be acceptable, as those economic regions that accept either will be at severe economic disadvantage to those who do not. (Tragedy of the Commons.) Massive change in energy technology comes slowly. It takes several lifetimes to put a new energy system into place, and wishful thinking can’t speed things along. Vaclav Smil, in A Skeptic Looks at Alternative Energy and Can We Live again in 1964’s Energy World? writes:

Nuclear is pretty much a drop-in replacement for coal, and it is coal that must most immediately be dropped-out and replaced. As an example, Germany embarked upon Energy Change in 2011 and hopes to generate 80% of its electricity from renewables (wind, solar, hydro, and biofuel) by 2050, roughly forty years. In contrast, France currently gets 12% of its electricity from hydro,⁶⁰ and 78.8% from her 59 nuclear plants, 56 of which were built in just the fifteen year period between 1975 and 1990.⁶¹ In the process, Germany’s electricity prices have become second highest in Europe, 40% higher than in France.⁶² Progress is intermittent,⁶³ and the environmental cost is staggering.⁶⁴

10.1 But I Have a Dream...

Alternatively, Pandora’s promise is of but a few appropriately-sized nuclear plants optimally placed near major centers of consumption, with no more redundancy than necessary to cover scheduled (and unscheduled) down-time and maintenance. But nukes don’t come exactly cheap,⁶⁵ and their cost has risen rather disconcertingly over time.⁶⁶ Money is jobs, low cost energy is economic competitiveness. Which approach is lowest cost? Can they be effectively combined?

10.2 But the Wind Always Blows...

“The wind always blows somewhere. Build enough windmills everywhere and everyone will have a share.” Or that’s the theory. But there is no free lunch, or energy either: each windmill costs something and has a finite lifetime. How much overbuild will a given economic region require to provide requisite amounts of “free” energy, and how much will it cost? Wind is not necessarily an unlimited resource. How much of the demand should be met by solar? Of which varieties? How much additional hydro is available, and how much can the wind and solar overbuild – and associated new transmission – be decreased with hydro balancing and backup? How much more may the overbuild be decreased with fossil fuel backup, gas and coal?

The last is a critical question, because it turns out the wind doesn’t necessarily “always blow somewhere”, and as we saw in MarginalUtility, certainly not “always somewhere” on scales of less than 1000 km. If we are to place a firm absolute cap on carbon emissions, then baring a highly effective carbon capture and sequestration system, any dependence on fossil fuel backup of necessity places an absolute cap on the amount of energy an economic region (and its populace) may generate and use.

Proponents argue such energy impoverishment may be minimized with suitably draconian demand management⁶⁷ (sometimes euphemistically termed “demand side participation”), which will require some sort of “smart grid” that can dictate who gets how much energy, when, and at what price. We should welcome our smart new overlords, though I’ll personally feel better about the proposition should I have some assurance they will actually have sufficient energy to meter out. Then its just a matter of how much one wants to pay, and when. But any long-term requirement of fossil-fuel backup and balancing, simultaneous with a strict cap on carbon emission, does not lend encouragement in that regard.

And it is carbon emissions that must be the bottom line. Lowest short-term costs and greatest short-term economic competitiveness are obtained from business as usual: but if we continue to burn coal like there’s no tomorrow, there won’t be.

No. The only questions that matters are “How do we reduce carbon emissions rapidly enough to save the planet from catastrophic global warming?” What is the most economical way to meet that number one goal while still maintaining economic competiveness relative to the rest of the world? Can we develop and deploy such technologies cheaper than coal? Can the rest of the world use them?

At low penetration levels the low cost of intermittent rewables is both a blessing and a curse, exacerbated by the particular methods we have chosen to subsidise them.⁷⁰ Though subsidized differently, cheap natural gas has a similar effect. Low cost of lower carbon sources is fine when they displace high-carbon coal. But when they deplace even lower-carbon nuclear, their environmental benefit becomes less clear. From Nuclear Power Fracked Off:

But by way of comparison, a recent Delaware renewables study (Budischak et al. see below) estimates “for 90 percent (renewable penetration), the cost jumps to 19¢/kWh best case (which uses hydrogen storage), while the cost for 99.9 percent coverage rises to a best case (using vehicle storage) of 26¢/kWh. These rates include the cost of procuring and installing the energy infrastructure in the first place.” As at present neither hydrogen nor electric vehicles offer practical grid storage, the study has a certain amount of built-in optimism. Back at the oranges, from figure 4 below we see 2013 rates top out at 9¢/kWh in Texas, so if NRG couldn’t bring nuclear power in for less than 10¢/kWh (roughly speaking), it would lose money. But some new nuclear power is still being built. Noteably by Southern Co. at their Vogtle plant in Georgia, where we see 2013 rates topping out at 11.24¢/kWh. That isn’t much of a margin, but Southern has a sweetheart deal with the local PUC, and think they can afford to be in it for the long haul.

And over the long haul, its that 99.9% renewables coverage that counts, as any residual is covered by fossil fuel backup and even 10% can’t be squandered on electric: it will be needed for process and transportation. So: 11¢ to 13¢/kWh for nuclear vs. 26¢/kWh for renewables would look like a no-brainer, were “all us or all them” not still the wrong question:

Because while it may look like nuclear is the clear winner and renewables should pack their tent and go home, time is of the essence, and it is only so fast that any technology can be deployed. As Prof. Smil explained, replacing our global fossil fuel infrastructure is a monumental undertaking. And even pristine nukes have their opponents.

A rough nuclear-is-half-the-cost-of-renewable for large grid penetration argument has also been made for Australia, albeit with considerably more rigor: see Section 10.5.3 below.

10.4 But What About China...

What about China? And India and Pakistan as well. China is at last beginning to take its coal addiction seriously. Although it still accounted for 59% of China’s newly-added capacity in 2012, coal use in China is reaching its peak.⁷¹ From Nuclear Power in China:

• Mainland China has 17 nuclear power reactors in operation, 28 under construction, and more about to start construction.
• Additional reactors are planned, including some of the world’s most advanced, to give a four-fold increase in nuclear capacity to at least 58 GWe by 2020, then possibly 200 GWe by 2030, and 400 GWe by 2050.
• China has become largely self-sufficient in reactor design and construction, as well as other aspects of the fuel cycle, but is making full use of western technology while adapting and improving it.
• China’s policy is for closed fuel cycle.

The United States currently has about 100 commerical power reactors. China is working with Westinghouse to increase AP1000 from its nominal 1.1 GW to 1.4 GW and then 1.7 GW. China will have the intellectual property rights to manufacture and market them itself. Advanced reactors + closed fuel cycle means fast neutron reactors.⁷² And where is America? Our public refusal to recognize the necessity and benefits of nuclear power does not mean the rest of the world will. “American leadership” is predicated on the assumption America actually leads. Thus far we have lead the world in safety of reactor design and operation. The lessons of TMI were hard learned, but they were learned – and much cheaper than they might have been. America’s resulting nuclear safety culture is second to none, and if we wish to see it shared by the rest of the world, we must participate in solutions to the rest of the world’s problems. Climate change is a global problem. In the context of waste management alone, the President’s Blue Ribbon Commission on America’s Nuclear Future concludes

Emphasis added. You don’t play the game, you don’t make the rules. Or even influence them.⁷⁴ The world won’t stop just because America wants to get off.

Not really. My search has hardly been exhaustive, but the three renewable modeling studies I have found look positively spin. Each starts with the proposition that their selected grid is candidate for all-renewables electricity generation, then proceeds to find a least-cost mixture of wind, sun, energy storage, and fossil backup and carbon tax that meets a desired emissions goal. Each also makes the explicit assumption that nuclear is a Really Bad Thing to be avoided at all costs. Which is a Good Thing, as the final costs they come up with may exceed EIA nuclear estimates 1.5 to 3 times.

Such studies are useful in the sense they place a dollars-and-cents ballpark around the price of avoiding nuclear. To the extent they minimize the serious infrastructure and market changes actually required, allowing casual readers to lull themselves a false sense an “all renewables all the time” solution is realistically feasible, they are perhaps a bit less so. In the interest of balanced reporting I’ll also discuss three additional studies that forgo the rose-coloured glasses in favour of the bean-counter’s green eyeshades.⁷⁵ The last of these, “RCP4.5: Pathway for Stabilization of Radiative Forcing by 2100” (section 10.5.6) is truely cost-minimizing, and includes all available technology options that can cost-effectively contribute to mitigation. In the least-cost solutions, nuclear ends up playing a dominant role.

We’ll consider a few fundamentals and a few constraints, then the models:

“The market” finds a local minimum to an optimization problem. The problem it solves is finding the most efficient (lowest cost) production and distribution of goods and services, subject to various constraints. In absence of any constraints on CO2 emission or fossil fuel production, and but $10/ton coal⁷⁶ and gas less than $5/MMBtu⁷⁷ in the cost function, its little wonder the optimal solution the market finds does not significantly reduce U.S. GHG emissions (if it reduces them at all). Wind and Solar PTC (Production Tax Credit) lowers cost of wind and solar relative to fossil fuels and nuclear, and promotes growth of the wind and solar industries, but has at best a secondary effect on CO2.⁷⁸

From Is the US Exporting Coal Pollution?:

These findings are expanded in Coal: the Ignored Juggernaut:

For such reasons the modeling groups generally eschew government subsidies for all energy sources, and rely upon modeled carbon taxes to increase the relative costs of gas and coal.

Finally, there are a few questions about some basic physical assumptions underlying the economics of wind. The first is the “the wind always blows somewhere” myth widely known to be false by intermittant modeling groups and grid operators, but nonetheless occasionally still perpetuated by intermittent renewal proponents. As discussed in section 10.2, market penetration beyond the marginal low-penetration capacity factor of wind – about 33% in the United States – will of necessity require energy storage and/or demand management to absorb the windy day excess. That, or feather some of the turbines and forgo even more capacity factor and efficiency (which turns out to be cheaper). Either way, full demand-managed load storage and/or backup generation availability will be required at all times to fill in the calms, which will be stochastic in nature.

The second has to do with wind energy density. Most modeling studies – including the massive NREL undertaking discussed in section 10.5.5 – assume an average wind areal energy density of 4 or 5 W/${\mathrm{m}}^2$. But from Are global wind power resource estimates overstated?

This may be an important consideration for windfarm planning and spacing, as deep penetration models place optimal wind contribution at over 80%. However, with our large land area and relatively low population density this is probably not a fundamental limitation within the United States. But it may well be elsewhere, such as the geographically smaller and population dense United Kingdom. While by no means “one size fits all”, there must be viable low-carbon energy rich solutions for everyone.

Update Tables 4 and 5 reproduce Budischak et al. Tables 3 and 4. GIV are “Grid Integrated Vehicles” – electric vehicles and plug-in hybrids integrated into a smart grid. From Table 4 we see that in the 90% solution fossil generation accounts for 7% (2.18/31.5) of total power, while for 99.9% fossils contribute but 0.05%. The former may be acceptible, the latter certainly is. Again, the PJM modeling study did not include transmission costs, which have been estimated to add another 1.5¢/kWh to the cost of renewables. There are uncertainties involved in projecting future technology costs, particularly for electrochemical storage technologies that do not enjoy appreciable commercial deployment today. Overnight 2008 pricings are most certain: for the 90% solution and hydrogen storage one gets approximately 23.5¢/kWh vs. about 10.5¢/kWh for 2008 - 2020 nuclear, a 124% penalty. If one accepts substantial breakthrough in hydrogen storage cost over next 17 years – and it could happen – then we would estimate 11.5¢/kWh for renewables vs. 8.5¢/kWh for nuclear, only a 35% penalty. I’m a bit skeptical that GIV storage cost will ever dip beneath central battery storage, but again can’t rule out the possibility, because while EV batteries are optimized for substantially different goals (e.g. energy/weight) than a central battery, there will (eventually, just not by 2030) be a lot of EV’s, with no reason not to connect them to a smart grid with certain priorities and caveats.

As an aside, in their Table 5 Budischak et al. estimate further cost reductions that might be obtained if it were posible to sell surplus renewable electricity for its thermal equivalent rather than just dumping it, for example by retro-fitting natural gas homes with redundant electric baseboard heat that might be used instead of gas during times of surplus wintery wind. It’s certainly a worthwhile question, but one possibly better addressed within the context of a study more specifically focused on hitting sufficient overall societal greenhouse gas emission targets at minimal cost – of which electric power comprises merely the easiest 33% (Figure 13).

In summary, the authors ignore wind and solar grid and transmission costs, which Germany’s experience suggests are non-negligible,⁸⁶ and exclude any possible nuclear power contribution from their optimization “...because it cannot be ramped up and down quickly and its high capital costs make it economically inefficient for occasional use.” But as seen in 9.6.1 many nuclear plants – including all Gen III+ and IV – can follow load at least as well as their fossil counterparts, and being dispatchable with 90+% capacity factors and carbon footprint comparable to wind,⁸⁷ there is no reason whatsoever to restrict nuclear to occasional use. And new build aside, by reducing the net amount of required new renewables generation, it is very likely employing current nuclear plant for their intended baseload use will decrease the future cost of a low-carbon grid. This possibility too was excluded. In short, although their current study is not without merit, “Cost-minimized combinations of wind power, solar power, and electrochemical storage, powering the grid up to 99.9% of the time” does not demonstrate a minimal cost low-carbon solution to powering the grid up to 99.9% of the time, only that the cost of wind power, solar power, and electrochemical storage needed to fill that bill by themselves will be exorbitant: 18.5¢/kWh vs 8.5¢/kWh for nuclear.⁸⁸ (Note 15 Feb 2014: this section needs to be revisited. The 8.5¢/kWh for nuclear figure assumes 90% capacity factor, which certainly cannot be attained when powering the entire grid. Some combination of gas, storage, and renewables will also be required. See comments on NREL’s Renewable Energy Futures, section 10.5.5 below.)

Likewise, the PJM study assumed constant load at 2008 values whereas Lang tries to meet estimated load growth out to 2050, which inevitably is what we must reach. Lang merely illustrates some easy-to-follow cost vs. emissions tradeoffs for different combinations of generation technologies we might employ while getting there. There are other differences as well, and optimization and more refined models can always make a good answer better – after one thoroughly understands the question. Lang suggests some likely results:

It is at least suggestive that of the six energy-mix options Lang considered, the only one that provided effective carbon limitation (gas+nuclear) contained no renewables.⁹²

10.5.3 Australia: Optimized AEMO Model, Draft Report April 2013

100 Per Cent Renewables Study – Draft Modelling Outcomes. From the Introduction:

Emphasis added. And just to ensure there are no internicene hard feelings,⁹⁴ AEMO helpfully adds:

So it’s all right then. Needless to say, DCCEE explicitly excluded nuclear from its definition of “renewable,”⁹⁵ with the now-predictable result AEMO’s $219 - $332 billion estimate eases in at over twice that of those definitions that do not. From New critique of AEMO 100% renewable electricity for Australia report by Dr. Ted Trainer, University of NSW:

And in 100 Per Cent Renewables Study Needs a Makeover, Martin Nicholson estimates

In regards to which, the AEMO Draft reports “results indicate that a 100 per cent renewable system is likely to require much higher energy reserves than a conventional power system. It is anticipated that generation with a nameplate capacity of over twice the maximum customer demand could be required. This results from the prevalence of intermittent technologies such as PV, wind and wave, which operate at lower capacity factors than other technologies less dominant in the forecast generation mix”⁹⁸ – similar to the Budischak et al. estimate of 290% overbuild for PJM.

As with the Budischak group’s PJM study, this Australian AEMO report has its strong points. Optimization is via Monte-Carlo sampling, and the cost function included increased transmission.⁹⁹ Nonetheless, the AEMO authors clearly felt constrained by DCCEE’s artificial constraints. Their results feel that way as well.

See Her Majesty’s Government’s The Carbon Plan: Delivering our Low Carbon Future (220 page pdf). From the Executive Summary:

So. The 86% Solution. The “By 2050, emissions from the power sector need to be close to zero” and “gas and coal-fired power stations fitted with CCS technology” requirements are completely consistent with those found by the global optimization study to be discussed in section 10.5.6. And to anyone who wonders “What were these guys thinking?”, I would once again remind them that Prof MacKay has gone to some length over the years to publicly explain. Sustainable Energy: Without the Hot Air is actually a lively and entertaining read. In the present context see Can we live on renewables?.

But just as I disfavor shuttering perfectly good nuclear plant just because current gas price is low and Production Tax Credit counterproductive, neither do I favor scrapping operational wind and solar farms just because they offend my personal esthetic and environmental sensibilities. All these optimal solutions include some degree of wind and solar. The goal is to most rapidly minimize carbon emissions at lowest cost with the tools available.

Emphasis added. Sounds much like the introduction to the AEMO study, perhaps sanitized a bit for American sensibilities. The authors are quite clear: “The scenarios were not constructed to find the optimal GHG mitigation or clean energy pathway e.g., to minimize carbons emissions or the cost of mitigating these emissions.”¹⁰¹

As for nuclear, that is part of the “full portfolio of clean technologies” not considered: “RE Futures did not allow new nuclear plants, fossil technologies with CCS, as well as gasified coal without CCS (integrated gasification combined cycle) to be built in any of the scenarios presented in this report. Existing nuclear (and integrated gasification combined cycle) units, however, were included in the analysis, as were assumptions for the retirement of those units.”¹⁰² With such limitations the study therefore does not and cannot address the fundamental question “How do we most rapidly minimize carbon emissions at lowest cost with the tools available? The authors recognize as much. From their Conclusions (pages 30-31):

While RE Futures might not press the questions that most need pressing, it’s what we’ve got and is fairly comprehensive at what it is. Let’s see what we can get from it:

RE Futures is based upon two low electricity demand scenarios: Low-Demand assumes 0.17% yearly growth; High-Demand assumes 0.84% as illustrated in Figure 1-2 (5).

RE Futures Low-Demand assumed a very-low demand growth of 0.17%/yr, High-Demand assumed a low 0.84%/yr, significantly lower than the historical annual growth rate of approximately 2.4% from 1970 to 2010.¹⁰³ RE Futures used 3656 TWh total demand in base year 2010, increasing by 7% to 3913 TWh in 2050 under their “low-demand” scenario and by 39.7% to 5109 TWh under high. Figure 2-2 (6) (or ES-3 of the Executive Summary) illustrates installed generation in 2050 on the (low-demand, RE-ITI technology improvement) scenario, and is reproduced here as Figure (6):

Update 12/15/2013:
For comparison, we have simply substituted nuclear first for coal, then for both coal and NG in REF’s Baseline scenario to obtain a final “no-fossils” mix in the above tables. We see that emissions are reduced more (by 85%) merely by replacing coal with nuclear in the baseline scenario than they are with REF’s “optimal” 80% RE low-demand solution, and are within 2% of the 90% RE solution if 25% of the NG is then replaced by an equivalent capacity of hydropower. Capital costs are less for nuclear, hidden in the “Details” below.
(End update.)
Save for hydro, source emissions are obtained by dividing column H by column G of RE Futures Table C-3: Nuclear 10.6 (4), Coal $\sim$1000, NG $\sim$500, Biomass 0, Geo 9.7, Hydro 26, CSP 78.5, PV 37.4, Wind 4.6, all as tCO2e/GWh. The hydro value is from Lifecycle GHG Emissions of Electricity Generation. The alternate 4 tCO2e/GWh for nuclear is from Energy Balances and CO2 Implications, citing Vattenfall’s Forsmark plant in Sweden achieving 3.1 tonne CO2e/GWh. Total GHG emissions/GWh is just the sum of the products of each component’s source emission and the component’s fraction in the mix. For the Baseline case this is (10.6$\times$0.11 + 1000$\times$0.53 + 500$\times$0.16 + 9.7$\times$0.03 + 26$\times$0.08 + 4.6$\times$0.06) = 614 tCO2e/GWh.

The Ratio column of the Capacity Mix table gives the ratio of Total capacity to the 446.6 GW average demand anticipated for 2050 under the low-demand scenario (3912TWh/y / 8760h/y = 446.6GW). The last column gives its inverse, the grid’s total net Capacity Factor. The REF study assumes the 80% solution is consistent with maintaining atmospheric CO2 beneath 460 ppm CO2, an assumption (also questioned at table 6 below) Prof MacKay finds optimistic: the Rest of The World wants its piece of the pie,¹⁰⁴ and in any event we will have a fixed CO2 emissions budget and squandering a full 50% (coal+gas) of it on fixed-source electricity generation seems shortsighted. High energy-density, compact, portable fossil resources should be purposed for what they serve their very best: transportation fuels. And that 15% biomass may be a solid DoE estimate, but I’ll unlikely be the only one to call it in question. That’s a lot of weed going up in smoke.

But neither am I the one doing the modeling, and 2050 is well beyond my own expiration date. I’m nonetheless going to suggest biomass beyond 5% (if that much) will not prove sustainable and unless captured, that carbon from coal will be better spent elsewhere. RE Futures does not assume CCS, as it is not yet a commercial technology. If the resulting deficit were to be met with nuclear, that would push its contribution to 20% - 27% of the national mix, at least double what it is today.¹⁰⁵

The study’s low-demand scenario places electric cost growth at 1.1%/y 2015 - 2050, high-demand at 1.3% resulting in respectively 47% and 58% price increase over the thirty-five year period, up to a final 16¢/kWh in 2050 (Figure 7). Which might seem reasonably bearable, were it not that EIA expects nuclear costs to actually decrease somewhat from about 10.5¢/kWh down to about 8.5¢/kWh over the same period¹⁰⁶ and grid integration of nuclear is much simpler and cheaper. ¹⁰⁷ (8.5¢/kWh would be for baseload nuclear. See updated Costs in “Details” below.) As we see in the above table, nuclear may also emit one tenth the CO2 as the 80% renewables combination obtained in the REF study.

All in all, one might mistakenly conclude RE Futures is rather more an argument in favor of nuclear than it is renewable energy. But as its authors point out, it is actually neither: that argument must be resolved via thorough cost-benefit analysis that is TBD. All in good time. When it is done, I expect our U.S. Carbon Plan will look much like the United Kingdom’s, though perhaps a bit more renewable as befits our breezy nature and sunny disposition.

The marginal cost of wind at current low market penetration is significantly lower than nuclear. At low market penetration existing fossil load following and peaking plants can balance all of wind’s intermittency, so every MWh of wind can be sold and used. (Not necessarily efficiently.) But in the previous subsection we cited five¹¹⁹ modeling studies that each suggest at the high (80+%) penetrations needed to meet anticipated carbon targets, marginal cost of renewables + storage will exceed that of nuclear by at a factor of one-and-a-half to two. The implication is clear: if we continue to subsidize intermittent renewables preferentially over nuclear – which is precisely what the Production Tax Credit (PTC) does – then we (a) reduce or delay nuclear design advances and construction while (b) possibly increasing intermittent renewable deployment beyond what is optimal for meeting CO2 targets at least cost. Carbon reduction is reduced or delayed either way.

An implication is not a proof; this one may be verified or refuted by modeling.

11.1 Production Tax Credits

Although since 2005 U.S. nuclear has been granted an $18/MWh Production Tax Credit (PTC) that partially compensates for the traditional $22/MWh granted wind and solar, the compensation is only partial and the effects are not the same. (Update 12/20/2013: see errata). As demonstrated by direct observation of operating energy markets, wind PTC can and does result in negative energy prices during up to 10% of operating hours in some markets. From Negative Electricity Prices And the Production Tax Credit we may take home the following:

• “Wind producers can readily turn wind turbines on and off, but have no incentive to do so because they still receive positive margins during negative price hours due to the PTC subsidy they earn when they generate... In the short term, the failure of wind producers to curtail output makes it more difficult for system operators to maintain reliability, and also makes it more costly for them to operate the regional electric grid.”
• “In the long run, the PTC destabilizes the market for conventional electricity as generators that are not eligible for the PTC are significantly harmed by negative prices, both in terms of near-term daily operational decisions, as well as long-term decisions to build or retire generation.”
• “America’s continued reliance on the PTC subsidy therefore will invariably deter investments in the conventional power generation needed to maintain a reliable electric system. Conventional generation is critical to reliability because wind generation often does not produce energy during times of peak electricity demand, while producing at high levels (and driving negative prices) when demand is low. In recent years, about 85% of total wind capacity has not operated during the peak hours on the highest demand days of the year, on average. Controllable conventional generation is thus needed to backstop wind and ensure the lights stay on.”¹²⁰

Emphasis added. Here “conventional generation” means gas, coal, and nuclear. In light of climate change, few tears might be shed over the premature demise of the odd coal plant. But as illustrated in figure 12, it is nuclear that Production Tax Credits and other renewables subsidies hit hardest. A conventional nuclear power plant, whether load-following or not, has a minimum power floor (typically 25% maximum load if load following, much higher if not) beneath which it may not operate without full shutdown. A restart may take several days. Faced with a negative energy price trough, the nuke operator is faced with Hobson’s choice: he may either pay the grid operator up to $4/MWh¹²¹ just for the privilege of keeping his plant operating and on line while the wind is brisk, or he must shutdown, bear restart costs, and be unable to bid (and receive) positive prices in the interim. You know – unable to do what he’s ostensibly in the business of doing.

Consider the (seemingly) only marginally less onerous case where the PTC stops at zero energy price and the grid operator is free to accept zero cost electricity from two producers: a wind farm and a nuclear power plant. Which choice minimizes the grid operator’s cost and most benefits his customers? The answer is the nuclear plant. Sure, the nuclear plant owner still loses money giving energy away free when he has fixed operating and fuel costs. But at least he can keep his plant running at its minimum power level and be able to bid positive prices when the wind dies down and the grid’s customers still need electricity that wind is not blowing hard enough to provide. And the grid operator wins because he’s retained the nuclear option in his mix. Unlike gas, he knows the long term trend of his cost for nuclear electricity. And unlike wind, he knows that the nuclear plant will much more likely be there when he needs it. But his customers and PUC expect him to buy electricity at the lowest cost, and if a fixed PTC allows a wind farmer to pay the grid operator to take its energy, then that is what the grid operator must do: the nuclear plant can either pay him more, or shut down. And if it shuts down, natural gas will be burnt in its stead.

Make no mistake: negative energy prices benefit nobody but the wind and solar farm owners. Even the grid operator loses on that one. Robin may justify robbing Peter to pay Paul only if Paul can spend Peter’s money more productively than can Peter. But any electricity consumer who needs must say “Sure, I’ll burn your free energy – but you gotta pay me to do it” by definition cannot burn his free money more productively than the taxpayers from whence it came.

Low natural gas prices exacerbate the problems, because they result in a lower bid price when wind is weaker and energy prices are positive. As consequence you have results such as the permanent shutdown of Dominion Resources’ Kewaunee nuclear plant in Wisconsin this past May. 560 MW carbon free energy – an equivalent gas plant emits 2.5 million tonnes of CO2 each year. No mechanical, maintenance, or operational issues. 90% capacity factor. Exemplary operating record. Recent 20 yr. extension to its NRC operating license. Plant paid for. Gone like a puff of smoke in a fitful breeze.¹²²

See Nuclear Power Cannot Compete with Cheap Shale Gas and Nuclear Plants Vexed at Prices That Shift as Demand Does. For such reasons it is argued that nuclear will never be competitive in unregulated electricity markets, which have become the norm this century in the United States. But to my knowlege (still looking) no one has yet shown that nuclear does not have an absolutely essential role in avoiding climate catastrophe. So something’s got to give, and the most obvious somethings are (a) The Climate, or (b) Production Tax Credits.

It would be one thing if one’s final goal were to build out renewables – that’s what their PTC is there to encourage. But as shown in the previous section(s), if what one really wants is to minimize emission of CO2, then it’s hard to fathom how wind + solar + any amount of natural gas will produce less greenhouse gas than the equivalent MWh delivered from nuclear. Even a 40 year old treasure like Kewaunee.

Wind and Solar Production Tax Credits are a literal license to print money, and should be either modified such that they do not apply beneath some minimum price floor, replaced with Investment Tax Credits or cash grants, or eliminated completely.¹²³ Production Tax Credits were introduced in the late 80’s to encourage growth in nascent new industries. This they have done: wind and solar are now quite mature international business, and no longer need that kind of subsidy.¹²⁴ Neither does nuclear. None of them do. What we all need is less CO2 – a lot less – and tweaking around the edges not only isn’t going to get us there, it quite likely is making matters worse.

Such is often the case when faced with a difficult problem that admits several solutions, one of which addresses the problem directly but is itself difficult to impose, while the others are second-best patches that nonetheless seemingly appear to allow at least a partial solution: it frequently transpires that the second-best “solutions” turn out to be worse than no solution at all. From which arises the Theory of the Second Best. Lets look at another example:

11.2 Carbon Taxes and the War on Coal

Well, its not really a war. Because war is not healthy for children and other living things, and we’re talking about coal. Coal means jobs. And jobs mean votes. Jobs and votes are healthy. So as a second-best, President Obama has directed the U.S. EPA to wage a proxy war on coal power plants instead. And the Supreme Court has ruled the EPA has authority to do it.

And to what end? Ostensibly, its because coal is by far the worst emitter of green-house gas, polluting the atmosphere with CO2 and causing uncountable future costs, harm, and death via climate change. So we’ll stop burning it here in the United States.

Not stop mining it, mind. Just stop burning it. Here. But we’ll continue mining and exporting and $elling the stuff to Europe and Asia, so they can burn it for us. There.¹²⁶ Lowering their energy costs in the process and sucking our heavy industry and electricity-intensive manufacturing jobs overseas along with our coal.

You know: precisely those industries and jobs we’d otherwise need to build-out wind and nuclear sufficient to displace enough natural gas to meet the remainder of our CO2 target, because the coal is being burnt anyway so anything else we won’t be able to do wasn’t going to make any difference either. And the lower coal costs abroad will give them furreners that much more disincentive to not deploy renewables and nuclear of their own, so why should we bother?

Seriously. You cannot make this stuff up.

No. The only way to stop CO2 concentration from increasing in the atmosphere is to stop emitting CO2 into the atmosphere. That’s the only way. Coal is mined to be burnt. That’s where over nine tenths of it ends up.¹²⁷ If you aren’t going to burn it, don’t mine it. If you are going to burn it and you don’t want more CO2 in the atmosphere, then you’ve got to put your toys away when you’re done. Your mother taught you that. Probably by withholding your toy allowance when you didn’t:

Carbon Capture and Sequestration. Place high enough severance and end-use tax on coal that none is mined for energy whose CO2 isn’t captured and stored. Place high enough severance on natural gas that only enough is piped domestically to fulfill minimal domestic peaking, transportation, process and chemical feedstock needs. Tax end-use preferentially to give CCS healthy incentive. Tax exports to allow like purpose only, and imports to penalize those who don’t play by our rules. Fine-tune and sauté to taste. When we’re done, any carbon tax revenue should be minimal because we won’t be emitting enough carbon to tax.¹²⁸ Until then we can keep any appreciable carbon tax revenue neutral, which will free up other taxes to create more jobs.

The beauty (if that’s what it is) of carbon taxes is that in addition to minimizing GHG emissions, in the process they can also minimize much of the “us-vs-them” acrimony in the renewables-vs-nuclear debate. A sufficently high carbon tax will prevent unwarranted shutdown of currently operating nuclear plants before their natural time. Beyond that, the market will find a cost-optimized mix of renewables, storage, nuclear, and gas. How much gas will depend on efficiencies of the low-carbon sources and storage, and the carbon tax penalties in the cost function. The tax penalties can be adjusted as necessary to meet CO2 and methane emissions goals. There will be enough energy for everyone, and everyone will have a share.

12.1 Load Growth Happens: Plan for it

We previously estimated the amount of energy available from current U.S. supply of depleted uranium and spent nuclear fuel. Realistically, we’re sitting on nowhere close to 1000 years of already mined and refined U-Pu-transuranic fuel. More like only 500 years, if we’re lucky. The reason we’re talking about 500 years fast-fuel supply on-hand rather than 1000 is because finding the wind, water, solar and requisite backup sufficient to reduce current US electric CO2 emissions by 80% isn’t the final goal. It isn’t even halfway there. The final goal is to reduce total US GHG emissions by 80% (or more) of 1990 levels. Of which electric power currently comprises only 33%:

From where will the other 67% be obtained? Increased efficiency is all very well and good – necessary even – but will only go so far, after which the large majority of the slack must perforce be taken up by (increased) electric power and possibly hydrogen. Electric cars. Electric trucks. Electric trains. Electric heat pumps. And that “20% Industry” – of what does that comprise? At least 17% is process heat – heat used for making chemicals (including hydrogen) at 800 - 1000 C.¹²⁹ Population grows as well, with the US estimated to increase by another sixth just by 2025.¹³⁰ We may have to double current US electric production, at relative carbon emission now 90% lower than today. Does one really think wind, water, solar, and NG will go that distance on their own? The National Research Council doesn’t think so. Their 2010 report, Limiting the Magnitude of Future Climate Change very strongly advocates immediate deployment of new nuclear technologies (as well as renewables), subsidy reform, and implementation of carbon taxes and/or cap-and-trade. It fully deserves an article of its own. Meanwhile, lets look at one of its graphics:

To this point we’ve been dealing only with CO2. The NRC takes a broader view, and Figure 16 reflects total US total GHG emissions. Of these CO2 comprises only 84%,¹³¹ so the figure is in close agreement with our previous 5.3 Gt/y CO2 of section 10. From Summary page 2:

Compare Figures 16 and 8 for RCP 4.5 and 2.6. They are not all that different. To achieve such lofty-yet-necessary goals we need clear-eyed, realistic modeling and cost-benefits analysis. We need a Plan. Toward which end the NRC panel also recognizes the urgent need to resolve the issue of nuclear waste.

12.2 Waste Happens: Deal with it

We need a Plan. France has one. Even thirty years ago when it was less certain, based in some part upon U.S. success with EBR-II,¹³² the French public was confident a long-term plan for nuclear waste could be developed.¹³³ France, of course, then enjoyed a certain confidence in science and engineering¹³⁴ that no longer encumbers the United States. As result France today enjoys very low carbon emissions per kWh, second only to deep-hydro countries Norway, Switzerland, and Sweden,¹³⁵ and correspondingly low price for electricity.¹³⁶ French commercial reactor waste is recoverable, and France has lead the global effort in researching how to transmute it entirely to shorter-lived decay products prior to permanent disposal.

The United States must come up with something similar. Whether we bury it, burn it, or sell it to someone who can, we’re sitting on 65 kilotonnes spent nuclear fuel that isn’t just going to go away on its own, not for another 170,000 years at least. What can we do? Permanent burial may be both a geologically and engineeringly sound solution – but not in my backyard. 170,000 years is something the earth may be able to swallow, but not anything we Amercans have yet been able to wrap our collective heads around. Spent fuel reprocessing, though popular in Europe, was an idea rejected here by the Ford and Carter administrations. The Integral Fast Reactor closed fuel cyle program was cancelled by Clinton. The government contracted with industry to have a permanent reposity open for business twenty years ago. Yucca Mountain, paid for by nuclear ratepayer taxes, was shuttered this past January. Meanwhile spent fuel rods pile up on-site in above-ground dry-cask storage at both operating and retired nuclear power stations. Its a situation that pleases no one – save perhaps a few radical no-nukes activists – and it’s unlikely the American public will buy off on another massive round of nuclear power plant build without a coherent plan for dealing with waste. Meanwhile, Rome burns.

Dry-cask storage is good for decades, perhaps centuries. Those casks are tough. But they aren’t forever tough and on-site real estate is limited. Uncertainty breeds contempt, and we have put up with the situation quite long enough. We need to know where the stuff is going to go, and when. A few options:

1. Bury-it-and-forget-it. This has been Plan A for thirty-five years and isn’t likely to change. But whatever its science and engineering merits, politically it hasn’t washed.
2. Declare a National Fast Reactor Plan and implement it. Decide where at least some IFRs and SMR reprocessing sites will be located, and make credible plans to ship spent fuel casks to those locations when they are ready. It may take centuries to burn all our accumulated spent fuel, and if some of those casks must be stored at secure intermediate repositories in the interim, make plans for that. And plans for 300-year disposal of the final decay product waste when the fast reactors are done.
3. Leave our spent fuel where it is until the BRICS have advanced their own fast reactor programs far enough to accept it. Then sell it to them.¹³⁷

I’ve tried not to hide my personal preference for Plan B. There’s just too much already-mined energy in this country’s spent nuclear fuel to just Plan A throw-it-away. Somebody will find a use for it. And Plan C – selling it all to somebody who isn’t American – somehow seems un-American. But charged with formulating a National Plan for Nuclear Waste that must involve considerably more than just Spent Nuclear Fuel (SNF) from commercial reactors, perhaps not surprisingly The President’s Blue Ribbon Commission on America’s Nuclear Future judiciously chooses mostly from Plan A while explicitly leaving Plan B open for future generations to persue should they so choose. From the the Commission’s Final Report Executive Summary:

I couldn’t agree more.

Update: U.S. Congress has held hearings this summer on the draft Nuclear Waste Administration Act (NWAA S.1240), a product of the President’s Blue Ribbon Commission and Congress, especially Senators Feinstein (D-CA), Alexander (R-TN), Wyden (D-OR) and Murkowski (R-AK). See Congress Needs To Take The Nuclear Option.

12.3 Toward a National Carbon Plan

As we saw in sections 10.5.5 and 10.5.6, American Universities and the United States national labs have carefully developed the tools – ReEDS, GridView, SAM, GCAM and friends – to concoct a credible United States National Carbon Plan. The National Research Council and IPCC contributors have clearly outlined what must be done. The modeling tools are specifically written to include nuclear as part of their optimization solution should we so choose. Our laboratories and universities have the organizational infrastructure and knowlegable personnel with experience with similar studies to do it. Nuclear may not only be the fastest deployable technology to address the problem, it may be least cost as well, perhaps by as much as a factor of two or three. That remains to be seen. What has been seen is that rushing renewables can lead to economic and ecologic disaster. Germany has taught us that. Part of that painful lesson is that as currently implemented, PTC and feed-in tariffs introduce chaos into the electricity market, and do little to actually decrease carbon emissions. By encouraging early retirement of existing nuclear and deferred deployment of new, they may actually result in increased emissions. If wind and solar must continue to be subsidised, they should be so through less disruptive means. At this point in their game carbon taxes alone (or cap-and-trade) should suffice.

Early utility studies showed renewable contributions could be economic up to about 30% penetration – a bit less than the capacity factor for onshore wind. That’s what Xcel showed for Colorado when we embarked on our renewables program in 2004, and is the target set by Colorado’s Renewable Energy Standard for 2020. Xcel has thus far done quite well by it, and expects to slightly exceed the 2020 target. Currently, wind contributes about 12% of the total power in the Xcel grid.¹³⁹ Interestingly, Xcel also operates three baseload nuclear units at two facilities in Minnesota,¹⁴⁰ together providing 1.7 GW and providing nearly 30% of the energy in Xcel’s upper midwest grid.¹⁴¹ Stretching from Minnesota through the Dakotas and Colorado down to Texas, Xcel taps the most reliable onshore wind resources in the country, and can probably provide invaluable insight on how best to integrate nuclear with intermittent renewables.

But we shall need far better than 30%, and industry and the national labs owe it to us to show how that might most rapidly be accomplished at greatest reliability and least cost. We’ve argued nuclear is a critical part of that solution, and waste management is an absolutely critical part of nuclear. Yes, existing commercial SNF may be burnt down to 300 yr levels in fast neutron reactors – but 300 years remains 300 years and at that could still take over a century to burn through just for what we have on hand: the public must know what we are going to do with our spent fuel in the meantime. Uncertaintly breeds only cynicism and contempt. But a national waste management program must manage all nuclear waste, and some of the non-commercial (e.g. defense related) waste is unlikely to ever be candidate for partitioning and transmutation. It will live a long time – hundreds of thousands of years – and must be securely dealt with.

Towards the view of a National Carbon Plan, it might be best to make a public distinction between management of commercial Spent Nuclear Fuel and management of the cold-war leftovers. Though there is much management overlap, one is a past mistake to be dealt with, the other a valuable resource for the future.

I’d hope my fellow citizens may eventually see it that way as well. Education is invaluable, and Pandora’s Promise is a valuable step on that path. Further up the road we need

• A National Nuclear Waste Management Plan that distinguishes past defense waste from ongoing and forward-looking civilian fuel cycles. Three hundred years may still be a long time, but it’s not forever. We need to get behind Congressional efforts with the draft Nuclear Waste Administration Act (NWAA S.1240) to ensure it encompasses the flexible storage envisioned by the Blue Ribbon Commision on America’s Nuclear Future.
• A National Carbon Plan whose purpose is a cost-optimized route to a least-carbon future. I say “least” because we are certain to overrun our 450ppm CO2e limit, and 600 ppm as well if we don’t get moving. If there is to be salvation, we must reduce CO2 emissions beneath the level where the biosphere, perhaps in conjunction with CCS of burnt biomass, can begin to absorb more than we emit. That’s a tall order, but we’ve served it upon ourselves.
• Anticipate a final mix of about 10% hydro, 25% other renewables, 40% nuclear, and 25% gas plus coal co-fired with biomass, all subject to CCS. That’s just a guess: the actual values are for the National Carbon Plan modeling to suggest, and of course “the market” to actually determine as new technologies emerge and ongoing cost trends stabilize. If the models and informed public input say “no nukes”, fine. But whatever it is we need to know how much its going to cost, how long its going to take, and how much total carbon will be emitted in the process.
• Most immediately: Production Tax Credit reform. PTC currently introduce negative price excursions into our unregulated markets, and this disruptive effect must be eliminated. Given carbon tax or cap-and-trade flexibility and sane sustainable energy incentives, it must then be determined whether any unregulated electricity market can handle the distributional and reliability complexities inherent in a high renewables penetration. If one can, then the models might suggest market mechanisms by which our carbon goals can be realized.

• Three Mile Island and Fukushima notwithstanding, current generation light-water reactors are by any measure very, very, safe. Statistically safer than any other energy source, including renewables. Generation III reactors will be safer by at least an order of magnitude.
• Generation IV fast neutron reactors will be safer yet, and extend the lifetime of uranium fuels effectively indefinitely. Thorium can stretch “indefinite” by at least another factor of three.
• Spent nuclear fuel Partitioning and Transmutation in fast reactors can reduce the radioactive lifetime of sequestered final waste from 170,000 years to a managable 300 years.
• In the process and at present (2013) demand, our current 65,000 tonnes spent nuclear fuel could by themselves, if burnt in fast reactors, supply all U.S. electric needs for 100 years. The 470,000 tonnes accumulated depleted uranium by-product of LWR fuel production could similarly supply all our electric needs for an additional 900 years.
• Realistically, nobody is going to bury that much usable energy anywhere they can’t retrieve it. Not even us: NRC regulations require any such deposition be recoverable for at least 50 years. If we (or our kids) can’t figure out how to use it by then, we can always do as with the rest of our tech waste: sell it to China.
• Nuclear deployment is probably much cheaper than renewables for the same amount of reliable power, possibly by factors of two or three, and with far less environmental impact at the grid penetration levels required for low-carbon generation sources to seriously impact climate change. We have cited six independent academic and national lab studies to support this conclusion. Although all sources of sustainable energy and storage will be important, the formulation of any National Carbon Plan must seriously consider nuclear as a major contributor to electric power.
• Current U.S. Production Tax Credits possibly result in increased greenhouse gas emissions by favoring moderate-carbon renewable+gas combinations that encourage early retirement and/or deferred deployment of very-low-carbon nuclear.
• Absent U.S. carbon taxes or cap-and-trade, our proposed coal plant carbon restrictions will likely serve only to further subsidize global Business As Usual.
• The previous three assertions are each amenable to economic analysis and modeling.
• Realistically, any deployment of nuclear power reactors on a scale large enough to seriously dent the carbon problem will not be politically acceptable in absence of a politically acceptible National Plan for Nuclear Waste. Congress has finally made a first effort with the draft Nuclear Waste Administration Act (NWAA S.1240). It can be done, and the American public needs to know how it will be done. All of it.

Articles:

Liquid metal cooled reactor.

Sodium-cooled fast reactor.

Nuclear Fuel Cycle.

Nuclear transmutation.

Options for the Treatment of Spent Fuel from Nuclear Power Plants - Partitioning and Transmutation .

Actinide and Fission Product Partitioning and Transmutation (conference proceedings, pp 85-92)

Advanced Nuclear Fuel Cycles (M.Salvatores’ PowerPoint presentation as pdf.)

Economic Case for the Pyroprocessing of Spent Nuclear Fuel.

Smarter Use of Nuclear Waste. Scientific American 2005. Readable overview of fast reactor fuel cycles and waste processing.

Fast Neutron Reactors.

Operating and Test Experience for the Experimental Breeder Reactor II (EBR-II).

Reactors Designed by Argonne National Laboratory: Integral Fast Reactor and links therein.

The Integral Fast Reactor. Y.I. Chang, Argonne National Lab (6 pages, 1988)

What is the IFR?. A short FAQ by George S. Stanford (May 2013)

Plentiful Energy – The story of the Integral Fast Reactor, is excerpted at Cost Comparison of IFR and Thermal Reactors and links therein.

Response to an Integral Fast Reactor (IFR) Critique and links therein.

Impact of Load Following on Power Plant Cost and Performance (51 page pdf)

Technical and Economic Aspects of Load Following with Nuclear Power Plants

Earthquake, Tsunami, and Nuclear Power in Japan.

Sites:

Carnival of Nuclear Energy 169 – a compendium of articles.

Touring ‘Nuclear Energy’ on the World Wide Web. A collection assembled by PBS.

World Nuclear Association Information Library is an extensive online reference.

World Nuclear News is the World Nuclear Association’s daily news site.

Radiation and Reason: the Impact of Science on a Culture of Fear Wade Allison, Emeritus Professor of Physics at the University of Oxford. Radiation and Reason is the title of Dr. Allison’s book. Visit his website and scroll down to “Download Recent Articles” for a wealth of accessible scholarly information.

Science Council for Global Initiatives. Tom Blees, President.

Brave New Climate. Prof Barry Brook and contributors.

The Breakthrough Institute. Michael Shellenberger, Ted Nordhaus, Michael Lind, Matthew Nisbet, and Roger Pielke Jr.

Atomic Insights by Rod Adams.

The Hiroshima Syndrome. Regular Fukushima Commentary and Accident Updates by L. Corrice.

New (Jan 2014): SARI: Scientists for Accurate Radiation Information

Policy:

Blue Ribbon Commission on America’s Nuclear Future Final Report 2012. Closest we have to a National Plan for Nuclear Waste – and by extension power – is this 180 page (pdf) bi-partisan report to then Energy Secretary Stephen Chu. Present Secretary Ernest Moniz is one of the authors.

Limiting the Magnitude of Future Climate Change, National Research Council 2010. Closest we have to a National Carbon Plan, advocates a very urgent and aggressive all-of-the-above deployment of renewable, nuclear, and carbon-capture-and-storage resources, together with early retirement of high-emissions infrastructure and a system of carbon taxes and/or cap-and-trade.

Effects of U.S. Tax Policy on Greenhouse Gas Emissions, National Research Council 2013.

Books:

The Nuclear Energy Option (Plenum Press 1990) is a comprehensive online book by radiation physicist Bernard Cohen.

Sustainable Energy Without the Hot Air is a comprehensive online book by Cambridge physicist and government consultant David MacKay, FRS.

Prescription for the Planet, by Tom Blees (Sept 2008, 426 pages, free pdf download at Science Council for Global Initiatives

Plentiful Energy – The story of the Integral Fast Reactor, by Dr. Charles E. Till and Dr. Yoon Il Chang. (Create Space, Dec 2011, 404 pages)

THORIUM: energy cheaper than coal, by Robert Hargraves (Aug. 2012)

WHY vs WHY Nuclear Power, by Barry Brook and Ian Lowe (May 2010, 128 pages)

(12/22/2013) Gt should have been Mt in the “Details” CO2 paragraph of section 10.5.5. (However, the paragraph has been removed in lieu of the 12/22/2013 update to this section.)

(12/20/2013) Nuclear Production Tax Credits. The Energy Policy Act of 2005 does not grant any subsidy to existing nuclear plant. Rather, its $18/MWh is a future provision that will apply to the first 6 GW of new-generation capacity such as the four AP1000 reactors now under construction in Georgia and South Carolina. (Thanks to Rod Adams’ Atomic Insight.)

(02/05/2014) Sec 10.5.5: 3.5 GW new nuclear capacity will save us but 25 Mt CO2e/y, not 25 Gt.

C Addenda

(12/15/2013) Updated the “no-fossils” table following figure 6 to include scenarios with no coal, but still utilizing natural gas. These are discussed in an update to section 10.5.5’s “Details” paragraph on Cost.

(12/22/2013) Added tabulated values of Electic System Costs to Figure (7), and estimated cost of an “85% Sustainable” scenario obtained by simply replacing RE Futures’s Baseline coal generation with nuclear.

(12/22/2013) We illustrate section 6.2’s assertion that “the storage timescale would be reduced from geological to merely historic”: