On 4-5 December 1998, Pugwash Meeting No. 243 was held in Paris, France.
The Prospects of Nuclear Energy
Report by Steve Fetter
A workshop on the prospects of nuclear energy was held 4-5 December in Paris. The meeting was organized and supported jointly by the Belgium, English, and French Pugwash Groups. Twenty seven individuals participated, about two-thirds from France and the United Kingdom.
The workshop opened with the organizers setting the context for the discussion. World energy consumption is expected to double or triple over the next half century, driven by substantial increases in population and in per-capita consumption in developing countries. What fraction of this energy demand will be supplied with hydrocarbons? renewable energy sources? nuclear energy?
The organizers observed that there has been no comprehensive review of the potential of nuclear energy in the last decade. During that time, the possibility of climate change due to greenhouse-gas emissions from the burning of fossil fuels has become a major concern. It was felt that Pugwash might have an advantage over other groups in starting a new dialogue on the potential of nuclear energy to supply a substantial fraction of future energy supply. The group agreed to publish a book based on the workshop proceedings.
[As of this writing, the book manuscript has been sent to the publisher, World Scientific. Readers interested in more detail than is provided here should refer to the forthcoming book. This report will focus more on the group discussion than on the content of the working papers presented.]
The possibility of additional Pugwash workshops on the subject was also mentioned. It was noted that better participation from developing countries would be required in any future meetings, particularly since most of the growth in energy demand will occur in these regions.
After these preliminaries, discussion moved to projections of future energy supply and demand and related economic and environmental issues. Today, world commercial energy consumption is about 12 terawatts (TW), of which 85 percent (10 TW) is supplied by fossil fuels. In a business-as-usual scenario, energy demand is expected to grow by a factor of two or three over the next fifty years, to 25 to 35 TW. If this demand is supplied primarily by hydrocarbons, with the products of combustion vented to the atmosphere, the carbon dioxide concentration of the atmosphere would soar, with possibly catastrophic consequences. If, however, we wish to stabilize greenhouse gas concentrations at a tolerable level—for example, an equivalent doubling of the preindustrial carbon dioxide concentration—traditional fossil fuels could supply no more than 10 TW in 2050, and perhaps much less.
There are two main approaches to reducing fossil-fuel carbon emissions. The first is to reduce the overall demand for energy by increasing the efficiency of energy use. We do not want energy per se, but the amenities or services that energy provides. Decreasing energy demand would reduce security risks related to the depletion of inexpensive oil and gas resources while making it easier to achieve environmental goals, such as preserving natural habitat and avoiding climate change. The amount of energy required to produce a dollar of economic product has been decreasing steadily in developed countries. With the proper policies, economic development might be achieved with only modest growth in world energy consumption. Such policies might include increasing the density of settlements to reduce transportation requirements, efficiency standards for buildings, vehicles, and appliances, and energy taxes.
Some participants were skeptical that efficiency improvements could supply a major fraction of the demand for energy services. It was noted, for example, that energy consumption in OECD countries had declined partly because energy-intensive industries had been exported to developing countries. Skepticism was also expressed about the willingness, ability, or desirability of having governments take the steps necessary to substantially increase energy efficiency. Most participants seemed to feel that while increased efficiency should be encouraged through increased taxes and efficiency standards, demand for energy will nevertheless increase substantially and that we must think about how to meet that increased demand.
This led to a discussion of the other main approach to reducing carbon emissions: increased supply of renewable and/or nuclear energy. As regards renewable energy technologies, the most promising are solar, wind, and biomass. The potential of other sources—tidal, wave, ocean-thermal, and geothermal energy, or a future of expansion of hydropower—is limited. One participant estimated the maximum contribution of renewable sources at 10 TW by 2050, with 2–5 TW being more likely. (For comparison, detailed scenarios developed by the World Energy Council and the International Institute of Applied Systems Analysis project a contribution from renewables of 6 to 10 TW by 2050.) If total demand is 25 TW or greater by 2050, and fossil fuels and renewables are limited to less than 10 TW each, this creates the opportunity for an expanded role for nuclear energy.
The technology and economics of nuclear powers was then reviewed. In 1996, 433 nuclear reactors produced electrical energy at a rate of 0.26 TW—about 17 percent of total electricity supply, and equivalent to about 0.8 TW of primary (i.e., thermal or fossil) power. Most of this energy is produced in light-water reactors, which use low-enriched uranium as fuel and water as coolant and moderator. In some countries there are plans to separate the plutonium from the spent fuel in reprocessing plants and to recycle the plutonium as fresh reactor fuel. Plans to develop fast breeder reactors, which produce more plutonium fuel than they consume, have been canceled in many countries. Concerns about the safety of nuclear reactors have led to development of several so-called “passive-safe” designs, although none has so far has received design approval by regulatory authorities.
Near-term prospects for nuclear power are not very favorable. Forecasts by various agencies range from a substantial decrease to a modest increase in installed capacity over the next 20 years, with fission’s share of total world electricity production falling to less than 10 percent by 2020. This is due a combination of factors: the availability of cheaper alternatives, the retirement of older plants, and public opposition to nuclear power in many countries due to concerns about accident and waste-disposal risks and potential links to the spread of nuclear weapons. The only region expected to experience significant growth in the near future is East Asia.
Overall, the safety record of nuclear power reactors in the OECD countries has been excellent. There has been one major accident in about 5,000 reactor-years of operation—the core meltdown at Three Mile Island—but this resulted in only a very small release of radioactivity, which had no direct public health consequences. Detailed calculations indicate that subsequent improvements in reactor design and operation have decreased the chance of a major accident to about 10–6 per reactor-year.
A recurring theme was one of public acceptability, and of perception versus reality. One participant noted that, although the real risks of nuclear power may be low, they are not perceived as such. In many democratic countries—Sweden, Germany, Netherlands, United States—nuclear power has been rejected. Some advocated education to correct misperception; others suggested that nuclear reactors and other facilities should be designed and operated in a manner that is transparently safe. According to one participant, the public has formed their opinions about nuclear power from the people in charge, who haven’t seemed to be terribly concerned about public health and safety.
There was extensive discussion of the prospects for nuclear power in developing countries. Some participants believed that nuclear power was appropriate and necessary to meet the growing energy demands of developing countries, including, in some cases, the development of indigenous reactor designs. Other participants reacted strongly against this notion. Some argued that nuclear power requires a degree of technical competence and vigilance that simply is not available in most developing countries. The probability of an accident will be substantially higher in developing countries, and another accident could be the death-knell of nuclear power. Another participant claimed that nuclear power is economically disastrous for developing countries because it is too capital intensive. Outside China, no nuclear plant has been ordered by a developing country since 1980.
There ensued a debate on the health effects of low-dose radiation. Most of the population dose from an accident—even a very severe accident, such as the Chernobyl accident—would be in the form of very low doses to very large populations. The standard assumption is that the probability of developing a fatal cancer follows a “linear, no-threshold” dose-response relationship. In other words, a dose of 1 milliSievert (mSv) to one million people is assumed to result in the same number of cancer deaths as a dose of 1000 mSv to one thousand people. (For comparison, the natural background dose rate is about 3 mSv/y.) It is impossible to confirm statistically whether or not the linear, no-threshold hypothesis is correct for doses below about 100 mSv. Indeed, based on currently available data one cannot rule out the possibility that small incremental doses of radiation have no health consequences at all, in which case releases of radioactivity from accidents or waste disposal would be far less worrisome. Other participants vigorously defended the linear, no-threshold model as plausible theoretically, and as reasonable for the protection of public health.
The discussion then moved to the back end of the fuel cycle. In general, there are two ways to deal with spent fuel: direct disposal as waste, or reprocessing to separate uranium and plutonium followed by disposal of the residual wastes. According to some participants, reprocessing is highly uneconomic, generates large stockpiles of surplus plutonium, and sets an unfortunate precedent for regions of proliferation concern. Others find reprocessing a vital link in the fuel cycle, making possible the recycling of valuable plutonium, thereby extending the uranium resource and decreasing waste-disposal costs and hazards. Opponents of reprocessing challenged both assumptions. Even if uranium may one day be scarce this does not mean that it makes sense to incur the costs of recovering plutonium today. And if the risks of waste disposal were truly negligible, it would not make sense to spend money to reduce these risks further. Although reprocessing avoids creating a “plutonium mine” that might be exploited by future generations for weapons purposes, it does so at the cost of exposing the separated plutonium to theft or diversion by current generations.
There followed a discussion of high-level radioactive waste disposal. A major barrier to a significant expansion of nuclear power is the lack of a permanent and acceptable solution to the disposal of either spent fuel or high-level reprocessing wastes. Today efforts are focused almost exclusively on geological disposal in mined repositories. The main problem is to demonstrate convincingly that, under all plausible scenarios—for example, earthquakes, changes in groundwater flow, human intrusion—wastes would not present a significant risk to future individuals, even a million or more years into the future. Although it is likely that such sites exist, there are many unknowns and it will take time to gain confidence that a particular site qualifies. But there is no pressing need to place wastes in a repository. Spent fuel or high-level reprocessing wastes can be stored safely on the surface for at least 50 to 100 years.
There was general agreement that, at least from a technical point of view, waste disposal should not pose a barrier to the expansion of nuclear power. In the words of one participant, if nature can contain billions of cubic meters of volatile methane in geological structures for a hundred million years, then we ought to be able to safely emplace thousands of tons of solid waste. But can this general confidence sufficient to support the further expansion of nuclear power, in the absence of a licensed repository that is accepting waste?
Several participants suggested that if would be a good idea if countries with licensed repositories were willing to accept, for a price, wastes from other countries. The current arrangement, in which each country—no matter how small or geologically unsuitable—must build its own repository, is nonsensical. There should, however, be IAEA standards and perhaps an international licensing mechanism to ensure that repositories are safe, particularly before they are allowed to accept wastes from other countries. The consolidation of spent-fuel wastes in a few repositories would also help reduce concerns about the widespread creation of “plutonium mines.”
Also discussed were separation-and-transmutation schemes to reduce the hazard of fission wastes. It is often observed that the most hazardous radionuclides in nuclear wastes are long-lived actinides, such as isotopes of plutonium and neptunium. It has been suggested that separating these long-lived actinides from the waste and transmuting them into short-lived or stable nuclides would greatly reduce waste-disposal risks. Transmutation could be accomplished in a reactor or accelerator, at some additional cost. However, the estimated risks of waste disposal, even over the very long term, are dominated in most scenarios by long-lived fission products, such as technetium-99 and iodine-129, which are far more soluble in water than are actinides. Transmuting the long-lived fission products generates additional problems. Again, if the risks of spent-fuel disposal can be made negligible, as promised, it would not be sensible to pay to reduce these risks still further.
Transmutation schemes involving accelerator-driven sub-critical reactors are also said to have safety and economic advantages, but these are dubious. Accident risks in light-water reactors are dominated by loss-of-coolant accidents, not criticality accidents. Some participants believe that a full probabilistic risk assessment would conclude that the accelerator-driven reactor was less safe than a light-water reactor. As regards cost, if an accelerator-driven sub-critical reactor is economical, then it is obvious that a critical reactor of the same design would be even more economical.
Another rationale for reprocessing, building accelerator-driven or fast breeder reactors, or moving to a thorium-based fuel cycle is to extend the uranium resource. Light-water reactors operating on a once-through cycle are inefficient users of uranium. Some participants maintained that light-water reactors would rapidly consume all stocks of low-cost uranium if there is a substantial growth in nuclear power. Thus, they argued, we must move to more fuel-efficient reactor and fuel-cycle technologies if nuclear power is to have a major role in world energy supply. Other participants found fault with this reasoning. First, uranium is very cheap today and likely will remain so for the foreseeable future, giving no economic incentive for the development of more fuel-efficient reactors or fuel cycles. Second, the cost of natural uranium is a small fraction of the total cost of nuclear-generated electricity. The price of uranium could rise by a factor of five or ten and still have little effect on the cost of electricity. Indeed, the price of uranium would have to rise above $200 per kilogram before reprocessing or alternative reactors were cost effective. Terrestrial uranium resources available at prices this high would be sufficient to support a substantial expansion of nuclear power for 50 to 100 years. Third, huge amounts of uranium exist in the oceans. One participant noted that it may be possible to extract uranium from seawater for $100 per kilogram. If so, fuel-efficient reactors would be cost-effective only if their capital cost was less than that of a comparable light-water reactor, which appears unlikely.
The discussion then turned to the links between nuclear power and the spread of nuclear weapons. To some participants, this link is weak to nonexistent. According to one participant, most countries do not want nuclear weapons, and those that do want nuclear weapons will not use civilian nuclear facilities for this purpose. The nine countries that have built nuclear weapons used dedicated military production facilities. In this view, proliferation is mainly a matter of politics—providing the proper incentives and environment so that countries wish to remain nonnuclear—than a matter of denying access to civilian nuclear technology. In any case, recently improved IAEA safeguards are adequate to ensure that facilities are used for only peaceful purposes.
Other participants were more concerned about the connection between nuclear power and proliferation. They believe that civilian nuclear facilities can used to facilitate and mask nuclear weapons programs. As examples, they point to the experiences in India, Iraq, and North Korea, in which each country used ostensibly civil facilities and activities for weapon-development programs. Furthermore, they contend that proliferation is not simply a matter of politics—access to nuclear facilities, materials, and technologies matters a great deal. This is particularly true for enrichment and reprocessing technologies, and for the use of fuels that contain high-enriched uranium or plutonium. Indeed, the primary reason that some participants oppose reprocessing is that the separation and use of plutonium creates proliferation risks. Even in countries with impeccable nonproliferation credentials, such as Japan, reprocessing generates apprehension in neighboring states that stockpiles of plutonium represent a latent nuclear arsenal. Such activity also creates a precedent that states of proliferation concern, such as North Korea, can use to justify their own plans to separate plutonium. Others objected to the focus on reprocessing, claiming that uranium enrichment or diversion of high-enriched uranium is an easier route to a bomb.
There was some discussion of ways to make the nuclear fuel-cycle more resistant to misuse. Ideas included: the use of the thorium fuel cycle, which generates less plutonium, and plutonium less suitable for weapons use; international storage and management of plutonium and spent fuel; the return of spent fuel from developing countries to countries that supplied the fuel or reactor.
In conclusion, although all participants agreed that Pugwash could play a valuable role in stimulated a renewed debate on nuclear energy, there was little consensus on what the result of that debate should be. Some participants favored a large expansion of nuclear energy, particularly in developing countries; others did not. Realistically, it will be difficult to maintain nuclear’s current share of world energy production during the next twenty or so years. Among those that favor a larger role for fission, there was disagreement about the means for pursuing that objective. Some embrace current technologies and a move toward fast breeder reactors or other advanced concepts; others promote technological fixes, such as safer reactors and more proliferation-resistant fuel cycles; others believe that the lack of public confidence in nuclear energy must be addressed more directly; still others believe that the main barrier to nuclear energy is its high cost.
Regardless of these differences, all agreed that carbon emissions must be curtailed and that supplies of carbon-free energy must be increased. Although we may not agree that nuclear energy will be an important part of the solution, we can agree that it is one of a small number of possibilities, and therefore deserves greater scrutiny.