IPFM International Panel on Fissile Materials

[This entry is drawn from Chapter Seven of the 2007 Global Fissile Material Report: "Managing the Civilian Nuclear Fuel Cycle." The printed version includes endnotes and, in some cases, additional figures. Entries are updated to reflect current data.]

Over the past twenty years, there has been little construction of new nuclear-power plants, with the exception of in Asia, where there has been some limited building. There is now, however, an active debate about the possibility of a dramatic nuclear "renaissance," driven in part by concerns over climate change. This chapter examines the potential implications of an expansion in nuclear power for fissile-material controls. The main concerns relate to the proliferation of national enrichment and reprocessing capabilities, which give states the capability to produce fissile materials for weapons. Overall, we emphasize that:

Nuclear power worldwide would have to expand five-fold or more to make a significant contribution to greenhouse-gas reductions. Such an expansion is far from certain, however, and even industry optimists do not see it being achieved before 2050.
Even if nuclear power expands substantially, there is no economic rationale for reprocessing, for the recycling of plutonium in light water reactors (LWRs), or for the adoption of closed fuel cycles of any type. Furthermore, there are compelling security reasons to avoid reprocessing and recycling.
Concern that some countries could use gas-centrifuge uranium-enrichment plants to make material for nuclear weapons has led to calls for dividing the world permanently into fuel-supplier states—basically, the NPT weapon states plus Europe and Japan—and fuel-recipient states. Such a division is in all likelihood unworkable. Using multinational ownership to protect against proliferation may be politically more feasible and is already happening to some degree.

Nuclear Power Today

As of March 2010, 436 nuclear-power plants, with a generating capacity of 370 gigawatts-electric (GWe) were in operation in 30 countries (see Table). These units provide about 14 percent of electrical energy worldwide. Eight countries accounted for 80 percent of global nuclear capacity: the United States, France, Japan, Germany, Russia, South Korea, Ukraine, and Canada.


Country	No. of Units	GWI(e)
Argentina	2	0.9
Armenia	1	0.4
Belgium	7	5.9
Brazil	2	1.9
Bulgaria	2	1.9
Canada	18	12.6
China	11	8.4
Czech Republic	6	3.7
Finland	4	2.7
France	58	63.1
Germany	17	20.5
Hungary	4	1.9
India	18	4.0
Japan	54	46.9
Republic of Korea	20	17.7
Mexico	2	1.3
Netherlands	1	0.5
Pakistan	2	0.4
Romania	2	1.3
Russian Federation	31	21.7
Slovak Republic	4	1.8
Slovenia	1	0.7
South Africa	2	1.8
Spain	8	7.5
Sweden	10	9.0
Switzerland	5	3.2
Taiwan, China	6	5.0
Ukraine	15	13.1
United Kingdom	19	10.1
United States	104	100.7
Total	436	370.5

Operating reactors and nuclear capacities by country, 2010. Data from the International Atomic Energy Agency, PRIS database, Nuclear Power Plants Information: Operational and Long Term Shutdown Reactors by Country. Update from March 5, 2010.

Nuclear Capacity Growth Projections

Projections to 2050 and Beyond. For those advocating or expecting a serious nuclear renaissance, the period after 2030 is of greatest interest. The 2003 MIT interdisciplinary study on the future of nuclear power presented one high scenario, in which nuclear power capacity reaches 1500 GWe in 2050 (see Figure).

MIT's 1500 GWe high-growth nuclear scenario predicts reactors in 58 countries. Map drawn by IPFM based on MIT, The Future of Nuclear Power, 2003. The nuclear scenarios are described in detail in the MIT report’s Appendix to Chapter 2. The capacities given are "nuclear equivalent capacities" defined as the needed electricity consumption divided by 8760 hours per year. For a 90% capacity factor, actual capacities would be 11 percent greater. The MIT scenarios are based on a Masters thesis submitted to the Department of Nuclear Engineering and the Engineering Systems Division of MIT by C. M. Jones, June 2003.

The MIT study estimated the distribution of this nuclear capacity by dividing the countries of the world into different groups based on their level of economic development. For the developed countries and Russia, the study then assumed that nuclear power would provide on average 51 percent of total electric power in 2050. In the large or advanced developing countries that already have nuclear power (including Argentina, Brazil, China, India, Mexico and South Africa) it was assumed to provide 30 percent of total electric energy in 2050.

Among the most populous, less-advanced developing countries, India was assumed to have 175 and Indonesia 39 equivalent GWe of nuclear power capacity in 2050. None of the least developed countries were assumed to have any nuclear power in 2050. However, several developing countries that have no or negligible nuclear power today--including Algeria, Armenia, Azerbaijan, Belarus, Georgia, Indonesia, Iran, North Korea, Malaysia, Pakistan, the Philippines, Poland, Thailand, Turkey, Turkmenistan, Uzbekistan, Venezuela, and Vietnam--were assumed to acquire nuclear-power plants by 2050. In fact, some of these countries are already expressing an interest in nuclear power.

Constraints on Nuclear Growth

Throughout most of the nuclear era, projections of future nuclear growth have been consistently too optimistic. The Figure below shows the history of IAEA nuclear-power projections for OECD countries.

IAEA forecasts made in 1973, 1977, 1982, and 2001 for nuclear capacity growth in OECD countries. Average of high and low projections in the IAEA/OECD series, Uranium: Resources, Production and Demand of the indicated years, courtesy of Tony McCormick, Urenco, August 2007.

Projections for nuclear-power growth outside of the OECD have been overoptimistic as well. For example, in 1985, the Chinese government projected a nuclear capacity of 20 GWe by the year 2000. At the end of 2005, China had only 6.4 GWe in operation. Similarly, in 1962, the Indian Atomic Energy Commission projected a capacity of 20-25 GWe in 1987. As of May 2007, India’s nuclear capacity was only 4.1 GWe.

Many of the factors that constrained nuclear power in the past--high capital costs, slower-than-projected growth in demand for electricity, scarcity of capital in developing countries, and problems with public acceptability--are likely to continue to dampen its growth. We find it unlikely that nuclear capacity will reach even the 1000 GWe of MIT’s low-growth scenario by 2050.

High capital costs. The figure below compares the International Energy Agency’s estimates of the cost of nuclear power with the costs of power generated by gas, coal, and wind. The cost estimates for gas and coal do not include the extra cost of capturing and sequestering carbon dioxide, which may become part of a future climate-change mitigation strategy. For the integrated gasification combined cycle (IGCC) system, carbon-capture costs are estimated to add about 1.5 cents per kWh.

Electricity generating costs for low and high-cost nuclear plants and some alternatives. The alternatives considered were combined cycle gas turbine (CCGT), coal steam, integrated gasification combined cycle coal (IGCC), and on-shore wind. Data from International Energy Agency, World Energy Outlook 2006, p. 368.

The "overnight" capital costs assumed for nuclear power in the figure (i.e., costs excluding interest charged during construction) were $2000/kWe (in 2006 dollars) for the low case and $2500/kWe for the high case. Nuclear power would be in the same cost range as coal and wind for the $2000/kWe case. The MIT study found, somewhat less optimistically, that nuclear power would be roughly competitive with coal if nuclear power’s overnight costs could be kept to $2000/kWe and countries enacted a substantial tax on carbon dioxide emissions to the atmosphere.

The capital charge for the plant is the most important cost element for nuclear power and is affected by the economic conditions of each country. For developing countries, in which investors require high real interest rates and returns on capital, every additional $500 per kWe capacity in the "overnight" capital costs adds about 1.5 cents per kWh to the cost of electricity. Other costs are less, but can still be significant. Since 9/11, concerns about terrorist attacks have driven up insurance and security costs. The interest charged during construction also adds significantly to costs, especially if there are delays.

A recent estimate for the cost of building the first nuclear unit at a new U.S. site was $2400-3500 per kilowatt (in 2006 dollars). The uncertainties are large because no new plants have been built in the United States in recent decades, and only a few elsewhere. In Asia, the overnight costs for recent plants (in 2002 dollars) ranged from $1800/kWe to $2800/kWe. In Europe, the Olkiluoto-3 reactor now under construction in Finland has an estimated overnight cost of $2500-3000/kWe. Construction of this reactor is already behind schedule by a year and a half.

The U.S. Energy Policy Act of 2005 sought to reduce investor risks for the first six new nuclear power plants built in the United States through two billion dollars of government guarantees and incentives. Nevertheless, Standard and Poor’s, which sets corporate credit ratings, stated in January 2006 that, "from a credit perspective, [the] provisions may not be substantial enough to sustain credit quality and make [nuclear generation] a practical strategy."

Slower-than-projected growth in electricity demand. The 2006 IAEA nuclear projection of 414-679 GWe in 2030 was based on an assumed growth rate of total global electricity consumption of between 2 and 3 percent per year. This is the range in which consumption grew during the 1990s. However, as analyzed by Goldemberg and Lucon, growth rates in both OECD and non-OECD countries declined between 1971 and 2003 owing to increased efficiency in electricity use and the saturation of electrification. If this second-order trend continues, the lower end of the IAEA’s range for electricity demand in 2030 is more likely to be realized and global electric consumption in 2050 would be roughly two thirds that assumed in the MIT scenarios.

Lack of capital for nuclear-power investments in developing countries. Unlike dams and other infrastructure, nuclear power plants are not underwritten by the World Bank or most other international lending organizations. Nuclear energy is also not included in the Kyoto protocol mechanisms, under which the industrialized (Annex 1) states can obtain credits against their own greenhouse-gas emissions for investments that reduce emissions in developing countries. The large investments required for nuclear power would therefore compete in developing-country budgets with investments for health, education, and poverty reduction.

Public acceptability. Simply to replace retiring nuclear capacity will require building a large number of new plants in the coming decades. Given continuing public skepticism about nuclear power, this may be challenging. An IAEA-sponsored opinion poll of 18 countries in 2005 found that about two-thirds of those expressing an opinion opposed shutting down nuclear power, but about the same fraction opposed building additional reactors. When asked specifically about the possible use of nuclear energy to combat climate change, only 38 percent expressed support for an expanded reliance on nuclear power.

Nuclear Power and Climate Change

Nuclear power’s environmental appeal is that it emits less carbon dioxide to the atmosphere than does coal or natural gas. When compared to an equivalent modern coal plant, 1 GWe of nuclear capacity operating at an average capacity factor of 90% reduces the amount of carbon released to the atmosphere by about 1.5 million metric tons annually.

Total global carbon emissions to the atmosphere in 2006 from fossil fuels were approximately 7 billion metric tons per year. Assuming business as usual, emissions are projected to approximately double in 50 years (a 1.6 percent average annual growth rate). The deployment of an additional 700 GWe nuclear capacity by 2050--in place of building 700 GWe of modern coal-electric plants--would lessen projected emissions by one billion tons of carbon per year. If the rate of carbon emissions is to be stabilized and then reduced, other technologies will have to be deployed as well. These technologies will both complement and compete with nuclear power.

Energy efficiency is likely to be the most important. In the International Energy Agency’s "Alternative Scenario"--in which governments adopt an array of policies to reduce greenhouse gas emissions--energy efficiency accounted for two-thirds of the potential emissions reduction by 2030. Other studies of opportunities to reduce greenhouse gas emissions have reached similar conclusions.

On the supply side, wind power and integrated gasification combined cycle (IGCC) plants burning coal with carbon capture and storage currently appear to be the most economically promising among the non-nuclear technologies that could reduce carbon emissions from electricity production.

Efficiency improvements in the power sector could also have a substantial impact. In its business-as-usual scenario, the IEA estimated that coal-based electricity production would roughly double by 2030, with an average efficiency reaching about 40%. Today, the worldwide average efficiency of coal-based plants is below 30%, but newer coal plants have efficiencies up to 46%. By 2030, efficiencies could reach 50% or higher. Using technologies to shift the average efficiency of the world’s coal-based plants from 40% to 45% in 2030 would save roughly the same amount of carbon emissions as would replacing 266 GWe of 50%-efficient coal plants with nuclear power, assuming both operated at a 90% capacity factor.

At a national level, the average efficiency of China’s 307 GWe of coal-fired plants was only 23 percent in 2004. The IEA predicts an efficiency of about 37% in 2030. If this could be raised to 42% for the 1040 GWe of coal-fired capacity that China is expected to have online by 2030, that would save 3.5 times as much carbon as would the 31 GWe of nuclear capacity that the IEA expects China to deploy by then.

Conclusion

If nuclear power grew approximately three-fold to about 1000 GWe in 2050, the increase in global greenhouse-gas emissions projected in business-as-usual scenarios could be reduced by about 10 to 20 percent.

Even a modest expansion of nuclear power would be accompanied by a substantial increase in the number of countries with nuclear reactors. Some of these countries would likely seek gas-centrifuge uranium-enrichment plants as well. Centrifuge-enrichment plants can be quickly converted to the production of highly enriched uranium for weapons. It is therefore critical to find multinational alternatives to the proliferation of national enrichment plants.

If a large-scale expansion of nuclear power were accompanied by a shift to reprocessing and plutonium recycle in light-water or fast reactors, it would involve annual flows of separated plutonium on the scale of a thousand metric tons per year--enough for 100,000 nuclear bombs. Fortunately, while there are strong security reasons to avoid plutonium recycling, there appears to be no economic rationale for such recycling for at least 50 years.