Toward a White Paper on a Mass Global Deployment of Nuclear Power: Fuel

It is my intention to present a draft of a White Paper, which will lay out a plan for a global deployment of nuclear power plants in sufficient numbers to insure the goal of an 80% reduction of global CO2 emissions can be accomplished by 2050. The first section of this draft focused on nuclear safety. This section will focus on the role of deploying Generation IV molten salt and sodium cooled reactors in meeting the goal, and the role of Reactor Grade Plutonium drawn from nuclear waste, in starting the breeding process. In addition to the use of RGP, a drawdown of nuclear weapons grade HEU and Pu-239 to start Generation IV reactors is both possible and desirable, but this possibility is not addressed in the body of this paper.)

Finding fuel for a post-carbon energy future

Energy policy should be guided by well thought out goals, and be based on rational and realistic plans. Unfortunately, both American and International energy efforts have not bee based on well thought out goals, or rational and realistic energy plans. The problem is especially serious because of the frequent inclusion of renewable energy goals in the formulation of over all national and international energy goals. In order for goals to meet minimal standards of plausibility, it should be demonstrated that a realistic plan for the implementation of the goal must exist. A realistic plan would be one which would address objections, which could in fact be carried out to successful conclusions.

Climate scientist have argued that it is imperative to lower global Carbon Dioxide emissions by 80% before 2050. In addition, the rapid economic development of nations such as India and China suggest that Global energy demand will increase, at the very time when there appears to be an need to decrease global energy production from carbon based fuels. Shifting global energy production from fossil fuels, while at the same time filling the energy demands of several billion people from developing countries is an enormous and extremely problematic task, which must be done.

Nuclear power is a promising candidate to bridge at least part of the energy gap. However, any account of a massive global deployment of nuclear reactors as a major or the major post carbon energy source faces a number of challenges. Critics can be expected to raise issues related to the cost of nuclear power, the magnitude of the task, the time requirements for nuclear construction, the availability of sufficient nuclear fuel the proliferation issues, the problems related to the management of nuclear waste, nuclear safety issues, and the potential dangers of nuclear terrorism. A successful nuclear global deployment plan should foresee these objections and demonstrate that no single objection or set of objections points to an insurmountable problem. A goal for a nuclear deployment plan would be the inclusion of a demonstration that no set of objections could motivate rational objections to the global nuclear deployment scheme.

Of course, all renewables deployment schemes, should also face the same level of scrutiny, with questions focusing on capacity factors, cost, the use of fossil fuel powered renewable backup systems, grid stability, redundancy, and the cost and efficiency of energy storage schemes. Efficiency advocates need to explain why Jevons Paradox and/or the rebound effect would not apply to efficiency based carbon mitigation plans. Unfortunately to date, renewable/efficiency based post-carbon energy plans to date appear to be singularly lacking in candor in identifying and addressing their problems.

Senior Collell poses a problem for a uranium powered future

Marcel Coderch Collell, a distinguished Spanish technocrat, has reviewed the possibility of replacing much of the worlds fossil fuel generated electricity by 2030. He argues that the limited nuclear fuel supply mans that nuclear power cannot play a predominate role in a post carbon energy order. He is extremely critical of the French Model: Electricity from Nuclear and/or renewable power.

He describes the French Model

The French model: electricity, nuclear or renewable One of the first options to consider would be to follow the French model progressively increasing the reactor park to ensure that by 2030, or maybe a little later, much of the world's electricity is predicted to generate fossil fuels out of nuclear origin, since it does not require, in principle, some technical innovation. Thus, power generation would not produce broadcasts, because it was nuclear or renewable. This would save huge amounts natural gas and coal, and oil-well with a consequent reduction emissions, and could force down, or at least not to contribute upwards of prices of fossil fuels and expand its availability in time.

But there are, according to Senior Collell, problems with the French Model.

But leave aside the logistical difficulties (and financial) would be a nuclear construction program of this size and evaluate how much fuel would be needed to fuel a reactor park of this magnitude, and which could be another limiting factor. Surely it would be, mainly, of building thermal neutron reactors with a Third Generation quasi-cycle Open fuel (MOX fueled with uranium enriched with some plutonium). In the best case, not expected to be operational by 2030 the Fourth Generation of fast neutron reactors with closed fuel cycles (which is expected to reach 60 times the performance thereof, by the massive use of plutonium) and, therefore, in the coming decades should be the main nuclear fuel.

Senior Collell then suggests that in order to follow the nuclear French model by 2030,

4,740 new 1GWe reactors would have to be built and [one] put in operationevery two days for the next 25 years

Senior Collell then suggests that this buildout would be very difficult in a business as usual world.

An optimistic estimate of construction times (five years) would mean having 950 teams of technical specialists, workers and machinery simultaneously working full time. This is hard to imagine, despite talk of standardising designs. In the previous period of nuclear construction (1963-88) only 423 reactors were built, at a rate of 17 per year.

He also argues that fuel shortages would constrict the depolyment of such a large reactor fleet.

A simple calculation suffices to show how an extension of the French model would collide with a scarcity of uranium. This is old news, given the serious doubts that already exist regarding the availability of uranium even to feed a few more reactors than now exist. In 2004, 365 GWe of nuclear capacity consumed about 67 kt of uranium (approximately 180 tons of uranium perGWe per year), of which 36 kt came from currently operating mines, while the rest came from recycled nuclear weapons and other secondary sources (that is, from prior production). Supply forecasts for the reactors currently in operation (plus foreseeable growth) put uranium mining production at 50 kt per year in 2015, with a significant shortfall developing in 2010, by which time Russia's nuclear weapons will have been dismantled and their uranium will have been consumed, . . .

If we assume linear growth from the current 365 GWe to 4,959 GWe in 2030, uranium demand would be around 400 kt in 2015 and 700 kt in 2030. This means multiplying by eight today’s estimates of production capacity in 2015, and multiplying by fifteen for 2030

Senior Collell sees these facts as casting the nuclear build out on the horns of a dilemma.

Let us suppose, however, for argument’s sake, that it were possible to achieve a production capacity of 700 kt/year by 2030. In the context of this analysis, two questions are raised: first, the CO2 emissions that would be generated in this phase of the nuclear cycle. Given the amount of uranium necessary, it would almost certainly be necessary to make use of hard rock deposits and low concentrations ore. There are fortunately multiple flaws in this delima argument. First the rock does not have to be moved. Uranium miners are increasingly adopting a mining technique called in situ leaching. When in situ leaching is practiced on uranium ore, the primarily the uranium is extracted, and the rock is left in place. Let us ignore for the moment these problems.

The problem that Senior Collell is pointing to is the limitation of the Light Water Reactor. Light Water Reactors were first developed as a means of powering American Nuclear submarines. In American Nuclear Submarines LWRs are small, they provide reliable power for 15 years, after which their cores can be replaced. There are also expensive, but nothing can serve as a substitute. Large power reactors can be even more expensive and there are very fuel inefficient. Part of the problem has to do with the flaws in the Uranium cycle. In LWRs as little as 0.3% of the potential fuel gets burned, and the rest falls into a category called "nuclear "waste." The problem is that uranium is relatively cheap, so it cost less to separate out the good stuff, the U-235 and use it for nuclear fuel. In LWRs a tiny fraction of the fuel gets converted to fissionable Pu-239, and a fraction of that gets burned as nuclear fuel. Unfortunately Pu-239 is not very good fuel in LWRs. When enough u-235 and Pu-239 is "burned" in a :WR, the fuel ceases to sustain a chain reaction, and has to be replaced. The now used fuel, still contains a significant amount of U-235 and an even larger amount of plutonium isotopes. Many of the plutonium isotopes are not fissionable, and plutonium produced in Light Water Reactors is not of weapons quality. It is also not very good nuclear fuel in Light Water Reactors.

The new French model of the future from Grenoble

Another French Model, one not contemplated by Senior Collell, is offered by Scientists of the Reactor Physics Group at Laboratoire de Physique Subatomique et de Cosmologie, Grenoble, France, together with other associates, have made important contributions to our understanding of the fuel management problems that would be faced in a Global Nuclear Deployment. The RPG were until very recently among the handful of world scientists who understood the potential of Molten Salt Reactor technology. They are among the most important and far reaching energy thinkers in the world, yet outside of the narrow circle found in the Energy from Thorium Discussion Forum, their work is almost unknown in the United States. Scientists from the RPG wrote a number of papers examining the potential problems of a Global Nuclear Deployment. "Molten Salt Reactors and Possible Scenarios for Future Nuclear Power Deployment offers a good introduction to their work. as a means of further enhancing the intellectual legitimacy of the Energy from Thorium approach to climate change mitigation, and to further enhance American Awareness of the work of the LPSC on Molten Salt technology. LPSC researchers use the term Thorium Molten Salt Reactor. Energy annalists associated with the Energy from Thorium approach to climate change mitigation use the term Liquid Fluoride Thorium Reactor or LFTR. In Future Deployment they write:

The worldwide demand for primary energy is constantly increasing and, if it is to be satisfied, solutions must be thought out and the extent to which the responses are adapted to the issue must be examined. There are not so many options once it is agreed that recourse to fossil energies should be as reduced as possible in order to limit green house gas emissions. Fission based nuclear energy is, along with new renewable energies and, in the longer run, fusion based energy, one of the primary energy sources capable of contributing significantly to satisfying the demand. The scenarios studied in our group show the potential, and limitations, for a worldwide deployment of nuclear power, and demonstrate that the different reactor types are quite complementary. This study shows that fissile matter availability comes as a strong constraint if a fleet of reactors able to breed their own fuel is to be started. In addition, such breeder reactors will not be deployed industrially before the next 20 to 25 years so that any transition towards extensive and sustainable nuclear power production will have to call on second or third generation light water reactors, which will have to be built.

We have considered three main reactor types:

• Pressurized water reactors of the second generation (PWR) and third generation (EPR - European Pressurized Reactor). These reactors do not breed their fuel. PWRs are currently in operation while EPRs will begin production in 2010 in our scenarios.
• Fast neutron reactors with a liquid metal coolant (FNR). These are fourth generation reactors that are based on the ²³⁸U/Pu fuel cycle. Their breeding ratio varies with the scenario considered. FNRs begin production in 2025 in our scenarios.
• Molten salt reactors (MSR). These are fourth generation reactors based on the²³²Th/ ²³³Ufuel cycle, with a neutron spectrum that can be anything from thermal to fast. These begin production in 2030 in our scenarios.

Our studies show that an intensive nuclear power deployment is feasible but that it requires careful handling of fissile matter resources and of nuclear wastes. The scenario that combines the three reactor types is by far the one that gives the most flexibility in the deployment of nuclear power; if necessary, it could accommodate more intensive production than we have set in our scenarios. The three reactor types complement each other strikingly; the use of natural fissile matter is optimized (figure 3); the volume of trans-uranians produced is minimal; the option to stop, then restart nuclear power production remains open so that decisions are not irreversible. Intermediate scenarios, with a greater or lesser contribution of FNRs as compared to MSRs can be considered in order to satisfy regional or other criteria but, with these studies, it appears that the ²³²Th/ ²³³U fuel cycle will be needed early on.

Figure 3: Evolution of natural uranium resources for the three scenarios considered.

Figure 4: Amounts of plutonium and 233U present in the fuel cycles of the reactors for the three scenarios considered.

The French Scientists from the University of Grenoble offered a more detailed analysis of the fuel management problem elsewhere. In "Scenarios with an Intensive Contribution of Nuclear Energy to the World Energy Supply," H.Nifenecker, D.Heuer, S.David, J.M.Loiseaux1, J.M.Martin, O.Meplan, and A.Nuttin, maintain that

If carried out with PWR or BWR reactors, the important nuclear power deployment will make heavy demands on natural Uranium resources. Resources are, presently, estimated to be around 20 Million tons. Assuming PWR or BWR reactors, the cumulative needs in 2050 could reach 16 million tons. This shows that breeding reactors are necessary to meet the needs or, alternately, that Uranium would have to be extracted from sea water, at a significant cost.

These considerations may, however, probably exaggerate the Uranium shortage. It is by no means certain that the cost of extracting Uranium from sea water would exceed the cost of breeding. On the other hand, there are reasons to suspect that the cost of LFTRs and other Molten Salt Reactors might be significantly lower than the cost of Light Water Reactors. Certainly when the large global thorium stock is added to recoverable uranium there will be no shortage of nuclear for a long time to come. Alvin Weinberg relates how the possibility of a future global uranium shortage was understood by the founding fathers of the Nuclear age, including Enrico Fermi, and Eugene Wigner. It was understood even then that nuclear breeding technology would have to be introduced as a part of a global deployment of nuclear power.

In "Intensive Contribution," the French team reviewed

two possible breeding cycles:
* The U-Pu cycle using fast reactors
* The Th-U cycle using thermal reactors

This analysis was expanded with typical French thoroughness in "Scenarios for a Worldwide Deployment of Nuclear Power," if anyone is interested. Both "Intensive Contribution," and "Worldwide Deployment" came to the same conclusion, that a deployment of Light Water Reactors can only be sustained until 2030. Lets call this the conservative case. Conservative, in that it is based on very conservative estimates of global uranium resources. While far more generous estimates of Uranium resources are justifiable, their actual existence is by no means certain. A really plausible plan should make conservative assumptions. If generous assumptions do not pan out, then the plan can be altered in to reflect a better than expected resource picture.

The nuclear intensive plan would assume a nuclear build out to 3387 GWe of electrical generating capacity by 2030. This is, in itself an enormous and extremely daunting build out, and indeed suggests that a major revolution in nuclear manufacturing technology will be required. Fortunately many of the components of that revolution are already understood, and none of them represents a serious impediment to technological change. Factory production of reactor construction kits, together with on site labor saving machines, and new materials savings reactor designs can be expected to improve reactor manufacturing, labor, time and materials efficiencies during the next decade, and to be reinforced by a learning curve. Such a large build out will probably require a shift of many reactor manufacturing activities from the final manufacturing site to factories. The recycling of old steam plant locations as nuclear power stations sites, will also save money and time for the buildout.

Thus while ambitious, the 3387 GWe buildout by 2030 is still not impossible, but the goal must be set soon. Both "Intensive Contribution," and "Worldwide Distribution" then looked at the U-Pu fast reactor cycle. By 2030 an enormous amount of reactor grade plutonium will become available. This RGP can be put to use both in the production of nuclear power and in the breeding of more reactor fuel. Doing so would serve as at least a partial solution to what is commonly seen as a major problem for nuclear power, the so called nuclear waste problem. Indeed the reuse of nuclear fuel turns "nuclear waste," into an asset. "Intensive Contribution," argues that given the supply of plutonium for LWRs and fast breeders, a buildout to 9000 GWe by 2050 is possible.

"Worldwide Deployment" looks at a number of added options including burning recycled RGP in LWRs. This delays, perhaps for a hundred years, but does not prevent the eventual draw down of fissionable materials that are tied to a non-breeding nuclear economy. A better use of the RGP is

Thus the transition to some form of nuclear breeding will be inevitable, if a long term commitment to nuclear power becomes a matter of policy.

Fast sodium cooled reactors are conventionally viewed as the preferred method of nuclear breeding, although various Molten Salt Reactor breeding options exit, and include many attractive features that are more than competitive with what liquid sodium cooled breeder reactors such as the Integral Fast Reactor. IFR backers claim higher breeding ratios, but there are reasons to doubt that high breeding ratios are compatible with optimal safety.

"Worldwide Deployment" also reviews a gas cooled fast reactor option, but did not like it as well as the sodium cooled concept.

There is a significant problem with the start up charge of fast breeders. Fast neutron reactors require much more fissionable material to maintain a chain reaction. If anything the French team underestimated the amount of plutonium required to maintain a high breeding ratio in a Fast Breeder Reactor. A research report from the S-PRISM design team, indicated that the RGP in 40,000 tons of spent nuclear fuel (about 400 tons of RGP) would start 22 IFRs capable of producing 33,440 MWe output. This would suggest that the IFR would be a useful way to use and with a low conversion ration, use up RGP, but not a major adjunct to fighting AGW. A higher breeding ration is possible, but report author, Allen E. Dubberley of GE Nuclear Energy, and his associates, did not discuss the safety problems related to the high breeding ratio.

The problem with the entire sodium cooled fast reactor scheme is stated simply by Lawrence M. Lidsky and Marvin M. Miller of the Massachusetts Institute of Technology,

The LMFBR was chosen over other breeder reactor designs because it was, in theory, capable of very short fuel doubling times, shorter than that of any competing reactor design. The doubling time is the time required to produce an excess of fuel equal to the amount originally required to fuel the reactor itself. In other words, in one doubling time there would be enough fuel available to start up another reactor. In the absence of mined uranium, only a short doubling time would, it was believed, allow nuclear power to grow fast enough to compete with alternative sources of power. Unfortunately, the theoretical advantages of the LMFBR could not be achieved in practice. A successful commercial breeder reactor must have three attributes; it must breed, it must be economical, and it must be safe. Although any one or two of these attributes can be achieved in isolation by proper design, the laws of physics apparently make it impossible to achieve all three simultaneously, no matter how clever the design.

Lidsky and Miller conclude,

Strong support for plutonium recycle, with its associated technical risks and societal costs, in the face of increasing evidence that alternative strategies are superior, is clearly counterproductive.

The Lidsky and Miller superior alternative strategies involve the the employment of Molten Salt Thorium Breeders, that is LFTRs,

Thorium fuel cycles have also been promoted on the basis of lower long-term waste toxicity and greater proliferation resistance, . . . The initial rationale for introduction of the thorium cycle was the perception that it was more abundant than uranium, and that it could be used to breed U-233, an isotope with superior properties for use in thermal reactors.

The principle safety problem with Sodium cooled breeders is the possibility that a bubble in overheated sodium would lead to an uncontrolled increase of power, which interm enlarges the size of the bubble. I should here, in the interest of fairness, note that I have this week come across papers from Argonne National Laboratory which reports findings from simulation studies which suggest that Integral Fast Reactors are safe even in configurations which produce high void worth ratios. If IFRs are in fact safe despite high void worth ratios, then they can probably be pushed to higher breeding rations, but it should also be pointed out that current ANL IFR designs emphasize low breeding ratios, and in fact appear design at a ratio that is equivalent to the anticipated breeding ratio of the LFTR.

The new Indian Fast Breeder is expected to produce new fuel at a 1.12 breeding ratio, far less than its theoretical maximum. Such conservative ratios may at least partially be motivated by safety concerns. It should be noted that a 1.12 to 1 breeding ratio is quite good by LFTR standards, and would be quite satisfactory if a fast breeder could match thermal Molten Salt Breeder's start up charge. In fact as many as 12 MSBRs can be started for every LMFBR, and if the LMFBRs could breed at its theoretical maximum, they could over time produce more fissionable materials than the MSBR, were it not for safety concerns related to higher breeding ratios.

Unlike the sodium cooled LMFBR which the laws of nature appear to frown on, the face of nature positively shines on Molten Salt Nuclear technology, which is an extremely "Green" energy source. Small breeding start up charges, mean rapid scaleability of reactor production and start up. Small start up charges mean that te availability of fissionable materials will not be a factor in determining how many and how fast future MSRs are built.

In addition, low breeding ratios may be viewed as a desirable proliferation control measures. In hugh ratio breeders, breeding surpluses can be viewed a potential proliferation tools through diversion. In a 1 on 1 converter, fuel diversion for weapons purposes will lead to reactor shut down, an undesirable consequence that would decrease the likelihood of proloferation.

Given even conservative estimates of American reactor fleet growth before 2050, the United States will have produced more than enough RGP by then starting up a fleet of breeding capable MSRs capable of supplying 80% of its energy needs. In the absence of Molten Salt reactors, LMFBRs breeding at the reported Indian Breeding ratio of 1.12 to 1 would be helpful, but would probably lead to a somewhat slower nuclear deployment of nuclear power. If the report of IFR safety at higher breeding ratios is correct then the IFR might improve the fast reactor picture. It is likely that ambitious goals such as a 80% reduction of CO2 emissions by 2050 could not be meet by use of conventional nuclear alone. Were MSRs unavailable in 2050 there would likely to be significan consequences in terms of energy prices and availability in a post carbon world. .

(I plan to include an account of the proposed Indian nuclear system as an appendix to this section of the White Paper describing the ambitious and complex long term Indian nuclear program. In addition I plan to discuss Molten salt fast breeder options in a second appendix.)