The Future of Nuclear Power in the United States

The Future of Nuclear Power in the United States
Terrorism Analysis Report 1
June 2011
Federation of
American Scientists
Washington and Lee
The Future of
Nuclear Power
in the United
By John F. Ahearne, Albert V.
Carr, Jr, Harold A. Feiveson,
Daniel Ingersoll, Andrew C.
Klein, Stephen Maloney, Ivan
Oelrich, Sharon Squassoni, and
Richard Wolfson
Edited by Charles D. Ferguson and
1 Federation of American Scientists
The Future of Nuclear Power in the United States
February 2012
The Future of Nuclear
Power in the United States
By John F. Ahearne, Albert V. Carr, Jr, Harold A. Feiveson,
Daniel Ingersoll, Andrew C. Klein, Stephen Maloney, Ivan
Oelrich, Sharon Squassoni, and Richard Wolfson
Edited by Charles D. Ferguson and Frank A. Settle
3 Federation of American Scientists
About FAS
Founded in 1945 by many of the scientists who built the first atomic bombs, the
Federation of American Scientists (FAS) is devoted to the belief that scientists,
engineers, and other technically trained people have the ethical obligation to ensure
that the technological fruits of their intellect and labor are applied to the benefit of
humankind. e founding mission was to prevent nuclear war. While nuclear security
remains a major objective of FAS today, the organization has expanded its critical work
to issues at the intersection of science and security.
FAS publications are produced to increase the understanding of policymakers, the
public, and the press about urgent issues in science and security policy. Individual
authors who may be FAS staff or acknowledged experts from outside the institution
write these reports. us, these reports do not represent an FAS institutional position
on policy issues. All statements of fact and expressions of opinion contained in this and
other FAS Special Reports are the sole responsibility of the author or authors.
About Washington and Lee University
Washington and Lee University, the nation’s ninth oldest institution of higher education, is among the nation’s premier liberal arts colleges and universities. Washington
and Lee University provides a liberal arts education that develops students’ capacity to
think freely, critically, and humanely and to conduct themselves with honor, integrity,
and civility. Graduates are prepared for life-long learning, personal achievement,
responsible leadership, service to others, and engaged citizenship in a global and diverse
For more information about FAS or publications and reports, please call
1-202-546-3300, email [email protected], or visit the website at
Copyright © 2012 by the Federation of American Scientists and Washington and Lee
University. All rights reserved. Printed in the United States of America.
ISBN 978-1-938187-00-1
Federation of American Scientists 4
The Future of Nuclear Power in the United States
February 2012
In early 2010, when we began producing this report, we could not have predicted that one
year later there would be a major accident at the Fukushima Daiichi Nuclear Power Plant in
Japan in March 2011. Writing this foreword in February 2012, it is still too soon to know
the full implications of this accident for the United States and the global nuclear industry.
Instead of focusing on this issue, we, the principal investigators, will discuss the original
motivations for this report. ese motivations are still relevant regardless of the accident. If
anything, the accident further underscores that constant vigilance is needed to ensure nuclear safety. e primary motivation is to educate policymakers and the public about where
nuclear power in the United States appears to be headed in light of the economic hurdles
confronting construction of nuclear power plants, the aging reactors (most of which were
built more than 30 years ago), and the graying workforce (many of whom are nearing retirement age). A corollary motivation is to provide guidance to policymakers in their decisions about the complex subject of nuclear power.
To acquire sage advice, we asked a distinguished group of experts to provide their
insights about the safety, security, building, financing, licensing, regulating, and fueling of
nuclear power plants. ese experts also addressed the issues of managing spent nuclear fuel
and associated wastes, comparing nuclear energy to other energy sources, and assessing the
potential commercialization of technologies such as small, modular reactors and Generation IV reactors.
Will nuclear power in the United States experience a revival in this and the following decade? It is too difficult to know at this time considering the complicated set of factors
analyzed by this report’s group of experts. We did not ask the group to reach a consensus.
Instead, each author focused his or her expertise on a particular aspect of nuclear power.
Nonetheless, we believe that it is worth highlighting here some insights from the
individual chapters. In the opening overview chapter, Sharon Squassoni assesses that “U.S.
nuclear energy growth can only be achieved with a combination of aggressive government
support and a complete revamping of the U.S. nuclear industry to stress standardization
and modularization in construction. e best approach for the U.S. nuclear industry over
the next five years will be to demonstrate that it can manage each stage of the licensing,
construction, and operating processes of the first four [new] reactors competently and efficiently.” Concerning the challenges of financing the new reactors, Stephen Maloney underscores the need for the United States to have a “viable bond market.” He advises that these
“markets require Federal Reserve policies, low corporate tax and capital gains rates, a strong
currency, and stable growth,” but the “deficit spending policies in place today do very little
to create a viable bond market.” Even if the financing risk were more manageable, licensing
and regulating are two additional requirements that can pose significant challenges. Albert
V. Carr, Jr. provides a masterful explanation of the U.S. history of licensing and regulating
nuclear plants and is cautiously optimistic that a nuclear revival is underway given the recent applications for new licenses.
5 Federation of American Scientists
Nuclear power also has to meet high safety standards. John F. Ahearne examines the
question of whether nuclear plants are safe enough and emphasizes that safety “remains a mixture of design, construction, maintenance, and what has been called a ‘safety culture’: the need
for all involved personnel to stress safety in all their practices.” Moreover, after “the two major
reactor accidents of TMI and Chernobyl, designs have been scrutinized and improved, operating
practices improved, and personnel training stressed,” and that reviews of the Fukushima Daiichi
accident “may spur further design changes and safety retrofits.”
While safety relates to unintentional accidents, security depends on keeping nuclear
plants protected against intentional attacks or sabotage. Harold Feiveson provides an in-depth
analysis of the design-basis-threat (DBT), which is an assessment of the plausible threats that
nuclear plants confront and must defend against, but he points out that despite improvements in
the DBT after the 9/11 terrorist attacks, “questions remain whether the DBT is yet realistic
enough to capture plausible threats by terrorist groups, and whether the DBT and associated
reactor security operations have been adjusted to accommodate industry concerns with cost.”
Furthermore, “there will always be the possibility of a beyond-DBT attack on a reactor,” and he
consequently recommends that the industry pursue new reactor designs, reactor site locations,
and operational procedures that would boost the inherent safety and security of the plants.
If nuclear power is to have a viable future in the United States, the plants will adequate
supplies of fuel. Presently, U.S. nuclear power plants are fueled with uranium-based fuels. But in
the future, they could use recycled plutonium for fuel. Ivan Oelrich addresses the reliability of
uranium supplies and the possibility of a plutonium fuel economy. He concludes that “allowing
for robust growth in nuclear-electric power generation and using fairly conservative assumptions
about current and future reserves, decisions about building nuclear reactors should not today be
constrained by concerns about fuel availability. The long-term fuel situation will be constantly
reevaluated but, for decades to come, uranium availability will most likely not be the factor limiting nuclear growth.” Concerning a plutonium economy, he determines that because of “the
technical uncertainties, making an irreversible decision today is ill- advised and it is unnecessary.
While there is universal agreement that some form of longterm waste repository will be required, wastes may not need to be committed to a repository right away. … Pending resolution of
questions regarding long-term geological storage or fuel for fast reactors, the plutonium can sit
in the used fuel rods where it is safe from theft and cannot be used for weapons.”
A robust supply chain and highly skilled personnel are needed to ensure the future of
nuclear power in the United States. Andrew C. Klein points out the reality that the U.S. nuclear industry is embedded in an international supply chain. He underscores that “U.S. utilities and consumers of electricity will benefit from a competitive market for the supply of
nuclear reactor systems, parts, and components. U.S. suppliers of these systems, parts, and
components must be enabled to effectively compete if they are to remain strong participants
in this market place.” Regarding the future nuclear workforce, he discusses that competition
“will come from various sectors, both inside and outside of the nuclear industry. The electric
utility industry, including all means of production and distribution of electricity will look for
Federation of American Scientists 6
The Future of Nuclear Power in the United States
February 2012
similarly educated and trained personnel.”
Further examining the theme of competition, Richard Wolfson analyzes the criteria from the perspectives of different groups of people for choosing nuclear energy compared to other energy sources. at is, consumers and businesses want cheap and reliable
sources. “An environmentalist will value sustainability and minimal environmental impact—especially low carbon emissions,” but would need to “smarten” the electricity grid in
order to substitute wind and solar—intermittent sources—for nuclear energy, which can
provide base-load power. “A state utility commissioner, concerned for stability in pricing
and availability of future energy supply, might want to keep the low-cost energy from established nuclear plants in the mix.” Wolfson also discusses whether alternative nuclear technologies “ could “substitute for today’s generation of light-water fission reactors.” Such substitutes could include the potential for fusion or fusion-fission reactors and breeder reactors. But given the technical challenges of commercializing fusion and the “spotty operating
records” of the handful of breeder reactors that have been built over the past several decades, he concludes that light water fission reactors will most likely be the nuclear technology of choice for decades to come. In considering alternative nuclear technologies, the report ends with Daniel Ingersoll examining Generation IV reactors and small, modular reactors. Given adequate government support for these emerging technologies to overcome
financial and institutional challenges, he foresees these nuclear systems as major “components of the future energy portfolio.”
As educators, we believe that this report will serve as both a useful tutorial for the
general public and a guide for policymakers. e insights in the chapters will help further
the necessary public debate as the United States wrestles with the formidable energy challenges in the future.
Charles D. Ferguson
Federation of American Scientists
Frank A. Settle
Professor of Chemistry
Washington and Lee University
February 2012
7 Federation of American Scientists
We are grateful to have had an all-star cast of experts to write the chapters. We chose them
in order to show differing perspectives on the future of nuclear power in the United States.
e views in the chapters are solely the authors and do not necessarily represent the views of
the Federation of American Scientists or Washington and Lee University.
e report benefited from the copyediting and formatting of Monica Amarelo, FAS’s
Director of Communications. We are also thankful for the advice we received from numerous colleagues during earlier phases of our collaboration when Charles Ferguson worked at
the Council on Foreign Relations.
We are very appreciative of the generous financial support from Mr. Harold “Gerry”
Lenfest and the Lenfest Foundation and his dedication to educating the public about
nuclear energy.
Federation of American Scientists 8
The Future of Nuclear Power in the United States
February 2012
About . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Acknowledgements . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
Chapter 1: Nuclear Power in the Global Energy Portfolio . . . . . . . . . . . . . . . . . . . . . . . . . 10
Chapter 2: A Critical Examination of Nuclear Power’s Costs . . . . . . . . . . . . . . . . . . . . . . 32
Chapter 3: Licensing and Regulation of U.S Nuclear Power Plants. . . . . . . .. . . . . . . . . . 46
Chapter 4: Safety of Nuclear Power . . . . . . . . . . . . . .. . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . 57
Chapter 5: Security of U.S. Nuclear Power Plants . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . 73
Chapter 6: Future Uranium Supplies for U.S. Nuclear Reactors . . . . . .. . . . . . . . . . . . . . 86
Chapter 7: Prospects for a Plutonium Economy in the United States . . . . . . . . . . . . . . . 98
Chapter 8: Required Infrastructure for the Future of Nuclear Energy .. . . . . . . . . . . . . 108
Chapter 9: Alternatives to Nuclear Power . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118
Chapter 10: Emerging Nuclear Technologies . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
About the Authors. . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
About the Editors. . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143
9 Federation of American Scientists
Nuclear Power in the Global Energy Portfolio
Chapter 1
by Sharon Squassoni
Nuclear energy has generated commercial electricity for more than half a
century. Although advocates had high hopes for its widespread use, nuclear energy
growth in the last twenty years faltered on lower costs for alternatives like natural gas
and a steep drop in public support aer the reactor accidents at ree Mile Island
and Chernobyl. In the United States, escalating costs and a nascent environmental
movement halted virtually all new construction aer 1978. Some countries, such as
Japan, France and the Republic of Korea, however, embraced nuclear energy enthusiastically.
Today, nuclear power plants produce about 14 percent of global electricity.
Without sustained and aggressive government support, this percentage is expected to
decline to about 10 percent by 2030, according to the International Energy Agency.
At least two factors will make it difficult for nuclear energy to gain a larger market
share – overall electricity demand is projected to double, and older reactors will need
to be retired.
It is this rising electricity demand, along with concerns about improving
energy security and mitigating climate change that led many more countries to consider nuclear energy as a viable option. At least 27 nations since 2005 have declared
they will install nuclear power for the first time and a total of 65 countries have expressed interest to the International Atomic Energy Agency (IAEA). is contrasts
with the thirty countries plus Taiwan that are already operating nuclear power plants.
e rganization for Economic Cooperation and Development’s (OECD) Nuclear
Energy Agency suggested in its 2008 Nuclear Energy Outlook that the world could
be building 54 reactors per year in the coming decades to meet all these challenges.
For many reasons, an expansion of nuclear energy of this magnitude will be
difficult to achieve. e current industrial base for nuclear reactors has supported just
ten reactors coming on-line per year for the past two decades. e nuclear industry is
scaling up its capacity but this could take some time. In some cases, the lack of a price
on carbon dioxide emissions means that new nuclear power plant construction will
Federation of American Scientists 10
The Future of Nuclear Power in the United States
February 2012
remain relatively more expensive than coal, oil or natural gas, although this varies from
country to country, depending on existing resources. Significant shale oil and shale gas
discoveries have made utilities, at least in the United States, less enthusiastic about
nuclear energy as a competitive source of electricity generation.
In response to mitigating climate change, many countries will find that
nuclear power is neither the least-cost nor the quickest approach to reducing carbon
dioxide emissions.1 Until nuclear energy is able to produce hydrogen or process heat,
or until transportation sectors are electrified, nuclear energy’s potential contribution to
reducing carbon dioxide emissions will be somewhat limited.
Perhaps most importantly, the March 2011 accident at Japan’s Fukushima
Daiichi Nuclear Power Plant shook the confidence of the public not just in Japan but
also abroad. e devastating earthquake and tsunami that killed tens of thousands of
people eliminated off-site and backup electricity for four of six reactors and their spent
fuel pools at Fukushima Daiichi. Hydrogen explosions destroyed secondary containments, exposing spent fuel pools, and three of the reactors had partial core meltdowns.
e Japanese government evacuated some of the population immediately. e clean-up
effort at Fukushima will drag on for years and the cost will likely range in the billions of
Other countries with operating nuclear power plants, including the United
States, announced safety reviews, and some halted construction and even operation of
existing power reactors.2 Several countries that had been considering nuclear power may
face a significant challenge in overcoming public mistrust. Still, the long-term impact of
the Fukushima accident on nuclear power in Japan and worldwide is unknowable.
Although many countries regard the possibility of another event combining a
magnitude 9.0 earthquake and tsunami to be very low, the difficulties Japan – a highly
1 Energy Technology Perspectives 2008, International Energy Agency, (Paris: Organization for Economic
Cooperation & Development) 2008, page 65. For example, in the Blue Map Scenario of the International
Energy Agency’s 2008 Energy Technology Perspectives, nuclear energy contributed to 6 percent of total
CO2 emission reductions to 2050, despite an average build rate of 32 reactors per year. Other approaches
contributed more to climate change mitigation: end-use fuel efficiency (24 percent); renewables (21 percent); end-use electricity efficiency (12 percent); end-use fuel switching (11 percent); carbon capture &
storage (CCS) power generation (10 percent); CCS industry transformation (9 percent); power generation efficiency and fuel switching (7 percent).
2 China temporarily halted construction; Germany shut down reactors pending a safety review; and Italy
suspended a national referendum on nuclear power.
11 Federation of American Scientists
Nuclear Power in the Global Energy Portfolio
sophisticated and technologically competent country – experienced because of the lack
of electricity is raising questions about the costs and risks of nuclear power.
Nuclear Energy in the United States: Promises of the Past
e 104 reactors operating in the United States constitute about 25 percent of world
capacity. Commercial nuclear power in the United States was a direct spin-off from the
military’s nuclear programs. General Electric and Westinghouse leveraged their military nuclear contracts with the U.S. Navy and emerged as the two dominant reacto r
vendors not just in the United States but in the world for many years. GE’s introduction of the “turnkey” contract, which offered fully constructed power plants at a fixed
rice, provided significant momentum to construction in the mid-1960s. Westinghouse
followed suit to remain competitive. By 1967, American utilities had ordered more
than 50 power reactors and in the next seven years, they placed an additional 196
orders.3 By 1973, 40 units were operating. ese first- and second-generation reactors
were built primarily by Westinghouse and GE, whose pressurized water (PWR) and
boiling water (BWR) designs, respectively, were adopted worldwide. Two-thirds (69) of
U.S. reactors are PWRs, and the remaining (35) are BWRs. Figure 1 shows the location
of these plants.
3 International Atomic Energy Agency , “50 Years of Nuclear Energy.” General Conference 48 Document.
Vienna, 2004. Available atn3.pdf
Federation of American Scientists 12
The Future of Nuclear Power in the United States
February 2012
Figure 1: Operating Nuclear Power Plants in the United States, 2011
Source: U.S Nuclear Regulatory Commission. Date: September 2008. Available online at
Along with construction of nuclear power plants, the Atomic Energy Commission (AEC) also encouraged spent fuel reprocessing and the development of plutonium breeder reactors, primarily in response to concerns about scarce uranium. In
1966, the AEC granted a license to Nuclear Fuel Services (NFS) to operate a commercial reprocessing plant at West Valley, New York, which reprocessed both defenserelated material and commercial spent fuel until 1972. A temporary shutdown became
permanent and NFS abandoned the plant to the State of New York. Legislation in
1980 committed the federal government to take on 90 percent of the cleanup costs,
which have totaled $2 billion so far.
Two other reprocessing plants under construction never managed to operate:
GE’s Morris, Illinois plant, and Allied-General Nuclear Services’ plant in Barnwell,
13 Federation of American Scientists
Nuclear Power in the Global Energy Portfolio
South Carolina. Declared inoperable in 1974, the GE plant eventually stored spent
fuel; the Barnwell plant was neither complete nor ready for licensing when the Carter
Administration decided in 1977 no longer to support reprocessing and recycling, even
domestically, because of proliferation concerns. By the time the Reagan administration
reversed that decision in 1981, Allied-General decided the Barnwell project was commercially unviable.
Long before ree Mile Island, regulations on nuclear power in the United
States began to tighten. In the early 1970s, critics of the AEC argued that its regulation
was “insufficiently rigorous in several important areas, including radiation protection
standards, reactor safety, plant siting, and environmental protection.”4 A 1974 reorganization of the AEC created the Energy Research and Development Administration
(ERDA, now the Department of Energy) and the Nuclear Regulatory Commission
(NRC). Creation of the Environmental Protection Agency, the Council on Environmental Quality and new requirements for environmental impact statements also had a
significant impact, as did growing public interest in environmental issues. More than
half the challenges to almost 100 construction permits for nuclear power plants between 1962 and 1971 came from environmentalists concerned about the impact of
waste heat from power plants on the local waterways. e creation of the Critical Mass
Energy Program (which reportedly had 200,000 members) by Public Citizen founder
Ralph Nader in 1974 to lobby against nuclear power further increased the pressure.
e changed licensing environment began to affect new reactor orders by
1975. In the early licensing scheme, less than 50 percent of the engineering designs were
generally completed before construction, requiring field engineering and backfitting
based on operating experience in other plants.5 ese designs were released too early to
engineering, procurement and construction. Slowdowns also came from utilities, because high finance costs and falling demand made it very difficult to borrow money to
build plants no longer needed by the original dates. “Cost-plus” construction contracts
also contributed to spiraling costs.6
4 United States Nuclear Regulatory Commission, Our History. Washington, DC. 2009. Available at:
5 Hultman, Nathan E. and Jonathan G. Koomey (2007). “A reactor-level analysis of busbar costs for US
nuclear plants, 1970-2005,” Energy Policy. Volume 35, 2007, page 5638.
6 Maloney, Steven, “Nuclear Construction Risk Drivers – en and Now,” presentation to Platt’s Fih
Annual Nuclear Energy Conference, Bethesda, Maryland. February 13, 2009.
Federation of American Scientists 14
The Future of Nuclear Power in the United States
February 2012
Cost overruns became more transparent and egregious, sometimes ten times
above industry estimates. For the 75 reactors built between 1966 and 1977, cost overruns averaged 207 percent.7 In the end, more than 100 reactor orders were cancelled,
including all those ordered aer 1973. Regulatory hurdles increased in the wake of
ree Mile Island, which may partly account for even greater cost overruns for the 40
plants constructed aer 1979, which averaged 250 percent.8
By 1985, popular magazines such as Fortune and Time had pronounced the
death of nuclear power in the United States; Forbes magazine called it “the largest
managerial disaster in history.” e $2.25 billion municipal bond default of the Washington Public Power Supply System (WPPSS) plants in 1983 certainly contributed to
that popular sentiment. e closing in 1989 of the Shoreham plant – fully constructed
for $5.4 billion and never operated -- was the final nail in the coffin. e initial cost
estimate for Shoreham had been $350 million.Figure 2 shows the distribution of costs
(in 2004 dollars) for 99 U.S. reactors from 1970 to 2005, using a 6 percent discount
7 United States Congressional Budget Office,“Nuclear Energy’s Role in Generating Electricity,” May 2008,
pp. 16-17.
8 Ibid.
15 Federation of American Scientists
Nuclear Power in the Global Energy Portfolio
Source: Hultman, Nathan E., Jonathan G. Koomey and Kammen, Daniel M., “What History Can Teach Us
About the Future Costs of U.S. Nuclear Power,” Environmental Science and Technology, April 1,2007 p. 2,091.
Relaunching Nuclear Energy in the United States
Many of the Bush-era initiatives on nuclear power focused on new plant construction,
but also on returning to a policy of promoting recycling of spent fuel. Although the
Obama administration may have wished to avoid these debates altogether, its decision
to cancel the Yucca Mountain repository in 2009 brought all these difficult issues to the
fore as did the March 2011 accident at the Fukushima-Daiichi reactors, which especially highlighted safety concerns with spent nuclear fuel pools.
Federation of American Scientists 16
The Future of Nuclear Power in the United States
February 2012
New Nuclear Power Plants
With the hiatus in building new plants, the nuclear industry has focused on improving
operating efficiencies and refitting plants. e 1992 Energy Policy Act created a “onestep” licensing procedure for new nuclear reactors, combining construction and operation licenses, and limiting kinds of interventions but did not lead to any new license
applications. Since 2000, however, a panoply of policies, laws, and programs to help
jump-start new nuclear power plant construction has produced applications, if no real
construction yet. A few of the highlights are listed below:
• 2001: the National Energy Policy Development Group recommends supporting “the expansion of nuclear energy in the United States as a major component
of our national energy policy,” including research and development for spent fuel
recycling with the aim of reducing waste streams and enhancing proliferation
• 2002: Nuclear Power 2010 spends $550 million to help jump-start new power
reactor construction. Includes shared costs with industry for regulatory approval
of new reactor sites, applying for licenses and preparing detailed plant designs.
Also includes development of early site permits separate from reactor design
reviews to facilitate licensing process.
• 2001-2009: DOE R&D budget triples for nuclear energy. Programs included
Generation IV program, the Nuclear Hydrogen Initiative Program (NHI), and
the Advanced Fuel Cycle Initiative (AFCI).
• 2005: Energy Policy Act of 2005 includes incentives such as production tax
credits, energy facility loan guarantees, cost-sharing, limited liability and delay
• 2010: Additional loan guarantees announced.
Of all these, the Energy Policy Act (EPACT) of 2005 and the issue of loan
guarantees deserve more description. Under EPACT 2005, a production tax credit
would provide 1.8 cents/kWh during the first eight years of operation of qualified new
nuclear power plants. To put this in context, the average wholesale price of electricity in
17 Federation of American Scientists
Nuclear Power in the Global Energy Portfolio
2005 was 5 cents/kWh.9 e credit has a limit of $7.5 billion, or the first 6,000 MW of
capacity (equivalent to about five plants). Only those projects that have applied for a
combined construction-operating license by December 2008, and that begin construction by January 2014 and operation by 2021 are eligible for the credit.
e U.S. nuclear industry has singled out government loan guarantees as essential because the private market finds loans for nuclear power plants to be too risky, and
U.S. utilities are too small to take on a bigger equity to debt ratio, which would lower
the cost of capital, a key element in the cost of the new plants. Under the loan guarantee program, the U.S. Treasury will guarantee 100 percent of a loan which is limited to
80 percent of the construction costs. is effectively transfers the risk of cost overruns
due to lengthier construction times from project owners to the taxpayer.
Congress appropriated $18.5 billion in loan guarantees for nuclear power
facilities, and President Obama has recommended tripling this to $54 billion. is still
falls far short of the $122 billion in requests. Industry sources suggest DOE will be able
to support no more than 2-4 reactors, given costs of $5 billion to $12 billion per
reactor. e Department of Energy awarded the first loan guarantee to the Vogtle
reactor project in Georgia (over $8 billion) in 2010.
e DOE also committed to sharing design and licensing costs for the “first of
a kind” reactor, with its share estimated at $281 million.10 EPACT also extended
Price-Anderson limits on liability through 2025, capping new plants’ liability in case of
accidents at $10.6 billion. Finally, delay insurance would apply to the first six new
licensed reactors delayed by the regulatory process; some $500 million would be available for each of the first two reactors and $250 million for each of the next four reactors. is was intended to compensate for delays in implementing the new combined
construction and operating license process by the NRC.
Spent Fuel Recycling
As noted above, the Bush administration sought to close the nuclear fuel cycle in the
United States by promoting the development of fast reactors to burn up plutonium and
“recycling” waste for that purpose. e basic idea was to reduce the volume of nuclear
States Congressional Budget Office, “Nuclear Energy’s Role in Generating Electricity,” May 2008,
p. 8. In December 2007 it was 8.9 cents/kWh.
10 Ibid, page 10.
Federation of American Scientists 18
The Future of Nuclear Power in the United States
February 2012
waste by reusing the fuel in fast reactors, which can burn more of the material. e
Global Nuclear Energy Partnership (GNEP), the Advanced Fuel Cycle Initiative, and
other related programs have all sought to implement that goal. us far, the U.S. Congress has taken a “go slow” approach, delaying demonstrations of advanced recycling
technologies until more research can be completed.11 A National Academy of Sciences
report in 2008, which reviewed DOE’s nuclear energy R&D, suggested that DOE reconsider reactor technologies under the Gen IV program that would support both advanced fuel cycles and the production of process heat, instead of pursuing two reactor
technologies – very high temperature reactors and sodium-cooled fast reactors – for
those tasks. It also recommended that DOE continue research on advanced recycling
techniques, rather than move toward a technology demonstration plant. e Obama
administration has advocated research into a modified open fuel cycle in which some
research would be conducted on conditioning spent fuel.
Current Status
e nuclear industry in the United States responded quickly to the incentives package
provided in EPACT 2005. ree designs for pressurized water reactors and two boiling
water reactor designs are now under consideration for this next round of nuclear power
plants. ese include Westinghouse’s Advanced Passive Reactor (AP-1000), AREVA’s
European Pressurized Water Reactor (EPR); and Mitsubishi’s Advanced Pressure Water
Reactor (APWR). In addition, designs for GE/Hitachi/Toshiba’s Advanced Boiling
Water Reactor (ABWR) and GE/Hitachi’s Economic Simplified Boiling Water Reactor (ESBWR) have also been submitted.
ese are described in greater detail in later chapters. It should be noted that
the hoped-for standardization of designs has not happened, that not all the reactors
have yet been certified and that a few of these designs have submitted modifications to
their applications. For example, the designs for both the AP-1000 and the ABWR have
been certified by the NRC, but planned changes will require additional certification
and in the case of the ABWR, design certification renewal. Only one project envisions
building an ABWR. e design certification applications for the other reactors were
11 United States Senate, “Global Nuclear Energy Partnership, Hearing Before e Committee On Energy
And Natural Resources,” First Session, To Receive Testimony On e Global Nuclear Energy Partnership
As It Relates To U.S. Policy On Nuclear Fuel Management. Senate Hearing 110-306, 110th Congress,
November 14, 2007.
19 Federation of American Scientists
Nuclear Power in the Global Energy Portfolio
submitted several years ago: the EPR and APWR in December 2007 and the ESBWR
in 2005. e table below, adapted from Standard & Poor’s, summarizes some of the
differences among these technologies.
Table 2: Comparison of reactor designs currently under consideration in the United States
Source: Swaminathan Venkataraman and Aneesh Prabhu of Standard & Poor’s, “New Build Risks and the NuclearRenaissance.” At
Federation of American Scientists 20
The Future of Nuclear Power in the United States
February 2012
Four considerations affect the attractiveness of these designs for U.S. utilities:
capital cost, time to market, evolutionary versus revolutionary technologies, and active
versus passive safety design features. For the most part, unregulated electricity genera
ors – such as Constellation Energy Group, NRG Energy Inc. and PPL Corp. – have
chosen active safety designs, presumably because they rely on the market for cost recovery and therefore want the most proven technologies. Regulated utilities have so far
valued the lower life-cycle costs of passive designs, choosing the AP-1000 and ESBWR.
Although Exelon Corporation initially chose the ESBWR, it appears that in the wake
of Fukushima, Exelon will focus on building natural gas plants as opposed to nuclear
power plants, citing their overwhelming costs.
New License Applications for Nuclear Power Plants
As of December 2010, 17 licenses for constructing and operating 26 new reactors were
filed with the NRC. By type of reactor, these include: fourteen AP-1000
(Westinghouse-Toshiba) at seven sites; three ESBWR (GE-Hitachi) at three sites; four
EPR (AREVA) at four sites; two ABWR (GE-Hitachi) at a single site; and three
APWR (Mitsubishi) at two sites.
Several of the license applications have been suspended by request of the proect managers, including Entergy for River Bend and Grand Gulf, and Exelon for Victoria County. NRG decided in April 2011 to terminate its involvement in the South
Texas Project, and Constellation Energy decided in late 2010 to pull out of its deal
with Electricite de France for the EPR project at Calvert Cliffs. EDF would like to go
forward, but requires an American partner to do so. e map below shows the locations of sites for new COLs.
21 Federation of American Scientists
Nuclear Power in the Global Energy Portfolio
Figure 3: Location of Projected New Nuclear Reactors
Source: United States Nuclear Regulatory Commission, March 2011, available online.
Enrichment Plants
In the United States, four new enrichment plants are either under construction or
awaiting licenses to begin construction. Until 15 years ago, the Department of Energy
owned and operated gaseous diffusion uranium enrichment plants at Paducah, Ken
ucky, and Portsmouth, Ohio. e 1992 Energy Policy Act privatized DOE’s enrichment capabilities, creating the United States Enrichment Corporation (USEC). e
Federation of American Scientists 22
The Future of Nuclear Power in the United States
February 2012
remaining USEC plant at Paducah enriches uranium for domestic use and for export.
U.S. utilities have relied on downblended Russian highly enriched uranium (HEU) for
about half of their low-enriched uranium (LEU) fuel since 1995. e plant at Paducah,
which is scheduled to shut down in the next few years (between 2010 and 2015), will
be replaced by a six million separative work unit (SWU) capacity plant in New Mexico
(the Louisiana Enrichment Site) and the Advanced Centrifuge Project, a gas centrifuge
plant expected to produce about 3.8 million SWU per year using American technology. 12 e table below shows the new capabilities:
Table 3: Enrichment Projects in the United States
rtium Location Technology Capacity (SWU) Date
Source: Increasing Enrichment Capacity for a Growing Nuclear Industry; presentation by John
M.A. Donelson, February 13, 2009
e Louisiana Enrichment Services (so called because Louisiana was the original proposed site) plant and the American Centrifuge Plant are licensed and under
construction. LES decided to expand its enrichment capacity from 3 million SWU to
5.7 million SWU. USEC has been slow to finish construction on the ACP for several
reasons. Its loan guarantee application was initially declined, but USEC submitted an
updated version in August 2010 for $2 billion. e expected cost of going forward with
the plant is at least $2.8 billion, not counting USEC’s initial investment of $1.8 billion.
Additional costs will include the costs of the loan guarantee, overall project contingency, financing costs and financial assurances.13 USEC has also sought and received
funding from Toshiba and Babcock and Wilcox for construction ($100 million each).
Donelson, John M.A., “Increasing Enrichment Capacity for a Growing Nuclear Industry,” Presentation
to Fih Annual Platt’s Nuclear Energy Conference, Bethesda, Maryland, February 13, 2009.
13 See USEC’s press statement on August 3, 2010, available at
23 Federation of American Scientists
Nuclear Power in the Global Energy Portfolio
AREVA has applied for a license for a gas centrifuge plant at Eagle Rock and expects a
decision in early 2012.
Waste Management
e history of nuclear waste management in the United States reflects a lot of studyand
research, punctuated by a few decisions every few decades. In 1956, a National Academy of Sciences study group concluded that a deep geologic repository was the best
solution to dispose of high-level waste from nuclear reactors. e Nuclear Waste Policy
Act, however, was not passed until 1982. It appears now, almost thirty years later, that
some parts of the law may need revision.
Figure 4: Location of Spent Nuclear Fuel
Source: Nuclear Regulatory Commission website. See
Federation of American Scientists 24
The Future of Nuclear Power in the United States
February 2012
Nuclear power reactors in the United States each year generate about 2,000 metric tons
of fuel. So far, the United States has accumulated about 57,700 metric tons of spent
fuel, which is stored in spent fuel ponds and in dry storage casks at 121 sites in 39
states. According to the 1982 Nuclear Waste Policy Act (NWPA), nuclear power plant
operators are required to pay into the Nuclear Waste Fund (now estimated at $20 billion) in return for DOE waste disposal services – that is, eventual disposal in a geologic
waste repository. at repository, designated as the Yucca Mountain site in 1987, was
supposed to have opened in 1998. Beginning in 1997, nuclear power plant operators
filed 56 lawsuits against the DOE for costs incurred in the absence of shipments to
Yucca Mountain. DOE estimates that its liabilities under the current law will total $11
billion if shipments begin by 2020, and a lot more if they do not. e NWPA did not
provide for another method of disposing waste, such as reprocessing, and the Nuclear
Waste Fund may not be used for anything other than legislated purposes. Further delays
are ahead, since the Obama administration decided to cancel construction funds for
the Yucca Mountain program in early 2009, while continuing the licensing process at
the NRC. is raises the question of whether the funding decision could be reversed in
the future. If so, advocates of Yucca Mountain would still need to address storage capacity and geology issues. e NWPA set an arbitrary limit of 70,000 tons for Yucca
Mountain, but it presupposed a second waste site would be authorized. By 2020, the
level of waste is expected to reach 81,000 metric tons. Including defense waste and
shipments by U.S. reactors through 2066, the expected accumulated waste is estimated
to reach 122,100 metric tons.14 Geologic issues include the risk of transporting radioactive wastes in a porous environment, because fractures in volcanic tuff can transport
water, and the location of the Bow Ridge fault line underneath a facility, rather than a
few hundred feet away.
Public Debate
Nuclear energy had not been debated seriously in the United States for decades
until the Fukushima accident. Support for nuclear energy in the past ten years has
focused on concerns about rising energy prices and dependence on foreign sources
of energy. For example, a February 2008 Pew on-line survey indicated that a majority of Americans believe “developing new sources of energy, rather than protecting the
14 Holt,
Mark, “Nuclear Waste Disposal: Alternatives to Yucca Mountain,” Congressional Research Service
Report, R40202, February 6, 2009.
25 Federation of American Scientists
Nuclear Power in the Global Energy Portfolio
environment, is the more important priority for the country (Pew, 2008).” e Pew
survey reported 48 percent of Americans opposed promoting more nuclear power,
while 44 percent favored doing so. is contrasts with a 2008 UPI/Zogby International Poll showing 67 percent support for nuclear power, and 23 percent opposed to
building new nuclear power plants.15
Gallup polls in the last decade shown in Figure 5 below reveal a steady increase
in favorable support for nuclear energy and a steady decline in negative numbers until
Fukushima. e progression began in 2001 at 46 percent favoring nuclear power and
48 percent opposing nuclear power and climbed to 62 percent favoring nuclear power
in 2010 and 33 percent opposing the use of nuclear power.
Figure 5: Gallup poll on nuclear energy for electricity in the United States, March 2009
Source: Gallup poll, March 20, 2009, available at
Concerns about contamination of the soil and water by radioactivity lay relatively dormant in recent years because of the strong support of the U.S. government for
15 See poll cited at
Federation of American Scientists 26
The Future of Nuclear Power in the United States
February 2012
nuclear power and the portrayal of nuclear energy as “clean, green and secure.” Marketing campaigns by the Nuclear Energy Institute (NEI) portraying nuclear energy as
“clean air” energy and by the NEI-funded the Clean and Safe Energy Coalition were
likely influential.16 On the whole, opponents of nuclear energy generally have had less
money to spend on media campaigns, and their message is less pithy. ey have stressed
that nuclear power is not the solution to climate change and that it is dangerous, polluting, unsafe, and expensive. e accident at Fukushima returned safety and waste
concerns to headline news. Shortly aer the accident, a Gallup poll showed 44 percent
of the public in favor (in contrast to 59 percent the previous year) and 47 percent
opposing nuclear power.17 Figure 6 below shows the results of a Pew Research Center
poll conducted about a week aer Fukushima.18
Figure 6: Pew Research Center Poll on Nuclear Power, March 2011
16 CASEnergy Coalition apparently supports nuclear energy as the only form of clean and safe energy. It is
co-run by former EPA Administrator Christine Todd Whitman and former Greenpeace activist Patrick
17 See
18 Available at
27 Federation of American Scientists
Nuclear Power in the Global Energy Portfolio
Many polls differentiate between support for existing nuclear power plants versus expansion. Oen, there is public support for the continued operation of plants but
perhaps less support for new plants, especially on new sites. Fukushima raised concerns
about existing U.S. reactors, particularly those of the same design as the Japanese
reactors (of which there are 23 in the United States). President Obama called for a
6-month review by the Nuclear Regulatory Commission of the safety of U.S. reactors
and Congress held several hearings in March and April 2011.
Whether Fukushima will have a lasting negative impact on public opinion in
the United States about nuclear energy is unknowable. Much depends on what happens
in Japan, both in terms of cost and environmental consequences, and what happens in
other countries such as Germany, Switzerland, and the UK. Public opinion will also be
swayed by the strength of U.S. government support for nuclear power as a component
of clean energy. While loan guarantees will undoubtedly continue, the enthusiasm of
the Obama Administration could diminish.
Outlook for the Future
Regardless of public opinion, the outlook for nuclear energy in the United States will
be at best, slow progress, possibly bolstered by success in managing and executing the
first five reactors. Lower natural gas prices threaten to derail the current interest in
nuclear power by U.S. utilities, and loan guarantees, while necessary, are not sufficient.
effrey Immelt of General Electric suggested a few years ago that only “five to ten U.S.
nuclear power projects would go ahead unless there was a carbon-pricing framework to
create incentives for utilities to build more.”19 John Rowe of Exelon stated his own
preference for building other electricity generation plants in an interview with Bloomberg news on March 16, 2011.20 For both, building other electricity-generating plants
would continue to be more cost-effective than new nuclear power plants.
A carbon “tax” would need to be higher than $30/ton of carbon dioxide and
19 Crooks, Ed and Francesco Guerrera (2007). “GE chief urges incentives to fuel nuclear switch,” Financial
Times. November 18. Available at
20 See
Federation of American Scientists 28
The Future of Nuclear Power in the United States
February 2012
possibly as high as $100/ton.21 Yet prices in carbon trading in Europe in the first three
years varied from about €30/mt to less than €0.02/mt; in the second round of trading,
allowances have been hovering in the low €20/mt (equivalent to $50/mt) range.22 In
the first half of 2009, the price hovered at 13 Euros/mt. A stable, long-term price for
carbon is far from assured.
A climate bill in the Congress has been stalled in the Senate for more than a
year, even though the House managed to pass a bill in June 2009 (American Clean Energy and Security Act). Although some progress is achievable in the energy appropriate
ions bills, a carbon price is widely believed to be politically too difficult, particularly
with a Republican-dominated House of Representatives that does not consider climate
change as an urgent issue. at said, coal and natural gas will continue to provide the
bulk of electricity generation in the United States. With abundant supplies (particularly arly now with the development of shale natural gas) and lower facility construction costs, both are cost-effective, particularly in deregulated markets. Some restrictions
may occur at the state levels. For example, in February 2009, Michigan Governor Jennifer Granholm called for a near moratorium on new coal-fired power plants, which
would affect eight new coal plants now in the approval process.23 In Florida, concerns
about dependence on natural gas and the ability to pass on construction costs to ratepayers have led to continued focus on nuclear power, despite high cost estimates.24
A wildcard in the mix may turn out to be recommendations from the Blue Ribbon Commission on America’s Nuclear Future, which are scheduled to be provided to the
President and Congress in late 2011 or early 2012. Secretary of Energy Chu appointed
21 Williams, Robert, “Can We Afford to Delay Rapid Nuclear Expansion Until the World is Safe for It?”
Presentation to Bulletin of Atomic Scientists Future of Nuclear Energy Conference. Chicago, Illinois.
November 1-2, 2006. Available at: See also Massachusetts Institute
of Technology (MIT), e Future of Nuclear Power: An Interdisciplinary MIT Study. Cambridge:
MIT, 2003.
22 Ryan, Margaret, “Platt’s White Paper: Profitable Operations and Carbon Costs are Key to Nuclear
Power Enthusiasm,” (NY: McGraw-Hill Publications) May 2008, page 3.
23 Hornbeck, Mark; Charlie Cain; and Gary Heinlein, “Governor pushes greenpower,” Detroit News.
February 4, 2009. Available at
24 Jeffrey Lyash of Progress Energy Florida has predicted that Florida will depend on natural gas for 55
percent of its electricity generation by 2017. Florida has built only natural gas plants since 1984, which are
not considered to be optimal baseload electricity generators because their fuel costs fluctuate significantly.
29 Federation of American Scientists
Nuclear Power in the Global Energy Portfolio
the Commission in early 2010 at the request of President Obama. e Commission,
hich is co-chaired by Lee Hamilton and Brent Scowcro, was tasked to “conduct a
comprehensive review of policies for managing the back end of the nuclear fuel cycle,
including all alternatives for the storage, processing, and disposal of civilian and defense
sed nuclear fuel and nuclear waste. is review should include an evaluation of advanced fuel cycle technologies that would optimize energy recovery, resource utilization, nd the minimization of materials derived from nuclear activities in a manner consistent with U.S. nonproliferation goals.25 Perhaps to influence the Commission or to
oreshadow its decisions since several Commission members were involved with the
tudy, the Massachusetts Institute of Technology made its own recommendations from
its study on the Future of the Nuclear Fuel Cycle in September 2010. e study recommended, among other things, that the United States should:26
• Accelerate the incentives provided to “first movers” – the first 7 to 10 new
plants to make the costs of electricity from new nuclear power plants competitive with that of coal.
• Continue the once-through fuel cycle with light water reactors, focusing on
improved fuel utilization.
• Use centralized storage for spent fuel for up to 100 years.
• Establish a new nuclear waste management organization that is quasigovernment.
• Integrate waste management with design of fuel cycles and pursue vigorous
R&D for innovative reactors and fuel cycle approaches; and
• Pursue fuel-leasing options for nonproliferation purposes.
25 e members of the commission span a range of backgrounds, and include Mark Ayers of the AFLCIO; Vicky Bailey, former DOE Assistant Secretary; Professor Albert Carnesale; former Senator Pete V.
Domenici; Susan Eisenhower; Inc.; former Senator Chuck Hagel; Jonathan Lash, President, World Resources Institute ; Professor Allison Macfarlane; Richard A. Meserve, former Chairman of the NRC; Professor Ernie Moniz; Professor Per Peterson; John Rowe, CEO of Exelon Corporation; and Phil Sharp of
Resources for the Future.
26 See for the text of the full
report.e summary report was released in September 2010 and the full report completed in March 2011.
Federation of American Scientists 30
The Future of Nuclear Power in the United States
February 2012
e MIT study highlighted one of the biggest issues for nuclear power in the United
States – its lack of cost-competitiveness with other means of electricity generation. It is
unlikely, however, that accelerating the current incentives for nuclear power plants will
be enough to motivate significant nuclear power plant construction in the United
States. What is needed is a price on CO2 emissions, which is infinitely more difficult
for the executive branch to engineer.
In the end, U.S. nuclear energy growth can only be achieved with a combination of aggressive government support and a complete revamping of the U.S. nuclear
industry to stress standardization and modularization in construction. Foreign capital
is also likely required. Even then, the challenges are formidable: just to maintain its
share of the electricity market, the nuclear industry would need to build 50 reactors in
the next 20 years. e best approach for the U.S. nuclear industry over the next five
years will be to demonstrate that it can manage each stage of the licensing, construction
and operating processes of the first four reactors competently and efficiently. In short,
U.S. nuclear energy needs to prove it has overcome the problems of the past.
31 Federation of American Scientists
A Critical Examiniation of Nuclear Power’s Costs
Chapter 2
A Critical Examination of Nuclear Power’s Costs
by Stephen Maloney
Since the nuclear industry’s inception more than 50 years ago, its forecasts for
costs have been consistently unreliable. e “first generation” plants, comprising both
prototype reactors and the standard designs of the 1950s-1960s, failed to live up to
promised economics. is trend continued with the construction of Generation II
plants completed in the 1970s, which make up the present nuclear fleet.
First, the total costs were far higher than for coal-generated electricity. In
particular, the capital cost of nuclear plants built through 1980 were, on average, 50
percent higher than comparably-sized coal-fired plants, adjusting for inflation and
including backfits to meet Clean Air Act standards. Second, there were extraordinary
cost escalations over the original low cost promises. Nuclear plant construction costs
escalated approximately 24 percent per calendar year compared to 6 percent annual
escalation for coal plants. ird, the economies of scale expected were not achieved in
the Generation II designs. e scale-up of nuclear plants brought less than half the
economic efficiencies projected.
In addition, over 120 nuclear units, approximately half the reactors ordered,
were never started or cancelled. e total write-offs were more than $15 billion in
nominal dollars. e red ink hit vendors and utilities alike, and cut across geographies,
company structure, company size, reactor design, and experience.
In the late 1970s, the Atomic Industrial Forum (AIF), predecessor to the
Nuclear Energy Institute, identified the main drivers of unmet expectations as growing
understanding of nuclear accident hazards, failure of regulatory standardization
policies, and increased documentation standards to ensure as-built plants actually met
safety standards. e combined effects doubled the quantities of materials, equipment,
and labor needed, and tripled the magnitude of the engineering effort for building a
nuclear power plant. In effect, AIF explained the failure of the cost forecasts in terms of
trying to hit a continuously moving target.
Federation of American Scientists 32
The Future of Nuclear Power in the United States
February 2012
But the AIF assessment suggests that policy changes could make the cost
target easier to hit. at may not be the case. Unlike other generating technologies,
nuclear construction is inherently more complex; complex technologies are prone to
cost underestimation. In addition, project dynamics can be significant. e cost of
materials was assumed to be predictable when, in fact, they were subject to
unprecedented monetary dynamics in the post-Vietnam War era. e cost of capital
was assumed to be low when, in fact, it would rise through the 1970s to extraordinary
levels in response to Federal Reserve Bank policy initiatives to combat asset deflation.
Moreover, utility balance sheets were assumed to be “bankable” when, in fact, severe
liquidity and capital constraints were adversely affecting the utility sector by the late
1960s limiting its ability to add high-cost capacity power plants in a recessionary
market. Furthermore, the demand for power was expected to grow at a premium to
sustained GDP growth when, in fact, demand stalled following the 1973 oil embargo,
the step-change increase in energy prices, and the subsequent recession and
deflationary periods.
Nuclear Construction Overruns before the General Design Criteria
Nuclear construction projects have always over-run their original estimates. e first
generation of U.S. nuclear power plants (pre-General Design Criteria plant (1954 to
about 1967)), and Generation II plants subject to “standardized” requirements of the
General Design Criteria (GDC)) both experienced similar overruns.1
Consider, for example, Consolidated Edison Company’s Indian Point Unit 1.
Announced in October 1954, and built before the Atomic Energy Commission’s
(AEC’s) Regulatory Staff published nuclear licensing and safety design criteria, the
forecasted cost for this thorium-fueled 275 megawatt (MWe) breeder reactor was $55
million. It entered service in 1962 at $110 million. Other prototypical plants
experienced similar overruns.
Vendors attempted to achieve economies of scale and scope through larger
plants and standardized designs. For example, the Oyster Creek plant announced in
December 1963 was nearly twice as large as Indian Point 1, and was the first of the socalled “turnkey projects” – plants designed and built to specifications. Most turnkey
1 See: “General Design Criteria for Nuclear Power Plant Construction Permits,” 10 CFR Part 50, July 11,
33 Federation of American Scientists
A Critical Examiniation of Nuclear Power’s Costs
plants entered service between late 1969 and 1972. Reportedly, Oyster Creek was the
single largest “loss leader” among such projects. By 1966 when the turnkey sales
program ended, General Electric and Westinghouse were reported to have taken nearly
$1 billion in losses constructing the 13 reactors in this class.2
Standardizing Nuclear Regulation: the GDC and its Defects
Regulatory standardization is rooted in risk analysis. As knowledge of nuclear
technology evolved, the hazards became better understood. But plants are built to
standards defined in the past. Costs rose as the “as-built” plants were modified to meet
the growing awareness of the hazards.
e 1954 revision to the Atomic Energy Act included instructions to the
AEC to issue licenses to private companies to build and operate commercial nuclear
power plants, as well as possess fissionable material. To be effective, safety standards
needed to relate to the hazard of a reactor accident. e AEC tasked Brookhaven
National Laboratory to quantify the risk of a reactor accident. is study, “eoretical
Possibilities and Consequences of Major Accidents in Large Nuclear Power
Plants” (WASH-740), was published in March 1957, and set the stage for the
definition and evolution of reactor licensing and safety requirements.
WASH-740 analyzed a so-called “worst case” reactor accident scenario at a
hypothetical 185 MWe reactor located some 35 miles from a major city. e analysis
made a number of assumptions for core damage dynamics and airborne dispersal of
fission products. WASH-740’s hazard findings were stunning: a worst-case accident
could result in 3,400 deaths, 43,000 injuries, and several billion dollars in property
damage. AEC embarked on a research program to better understand reactor safety, and
standardize and strengthen safety standards applied to plants under construction. is
process oen led to backfits.
A detailed summary report of then-current emergency core cooling (ECCS)
technology was published in early 1971 as the “Brockett Report” and highlighted
design inadequacies in a design basis accident. Most backfits were systemic in nature
and the “knock-on” effects rippled through plant designs. For example, ECCS upgrades
meant more water would have to be injected sooner in an accident and at higher
2 H. Stuart Burness, W. David Montgomery, and James P. Quirk, “e Turnkey Era in Nuclear Power,”
Land Economics, Vol. 50, No. 2, May 1980.
Federation of American Scientists 34
The Future of Nuclear Power in the United States
February 2012
pressure requiring larger pumps and valves. Larger pumps would lead to larger pipes,
larger pump motors to drive the pumps, and larger emergency diesels to power these
larger loads. Redesign imposed delays and inefficiencies in the construction process.
At the time, 53 plants were well along in their construction. e redesign of
the ECCS capabilities during mid-construction triggered ripple effects of detailed
engineering, procurement, design, and construction requirements over the next 10
years. ese affected equipment qualification to perform under accident conditions,
seismic protection, pipe rupture in reactor accidents, risk of heavy loads damaging
structures, systems, and components important to reactor safety, flood protection,
tornado protection, fire protection, structural integrity of concrete, reactor
containment penetration integrity, and electrical system independence and protection.
Quality assurance (QA) standards were also introduced into safety regulations
amidst the design and construction boom. Backfitting this program had an especially
significant effect on construction costs. e original QA requirements were contained
in a 1967 policy statement. In 1970, this general statement was amplified with the
publication of Appendix B to 10 CFR 50. At the same time, the American National
Standards Institute (ANSI), the parent standards body, organized committees within
its member professional societies to develop detailed standards for meeting the new
criteria. Twenty-four new quality assurance standards were published amidst the peak
construction period of the 1970s.
Assimilating QA requirements into ongoing engineering and construction
processes, however, was not always easy. An extreme example of QA problems is
evident in the Zimmer plant. Some 10 years aer publication of 10 CFR Part 50
Appendix B and as the plant was nearing completion, Nuclear Regulatory Commission
inspectors cited the plant for problems in its QA program. ese defects surfaced in a
growing number of pipe weld inspections. Tracing the defects to engineering and
design processes, NRC staff found other process deficiencies and eventually concluded
the defects implicated the processes used to procure the plant’s components and
contractors. e more NRC looked, the more systemic the quality concerns became.
As the number of deficiencies grew, the more the inspectors looked. By late 1982, the
deficiency and corrective action lists reached a point whereby NRC issued a stop-work
order on the project so everyone could get a handle on things. By then, the cost
estimates to complete the project had grown to $3.4 billion. e owners cancelled the
project in 1983. e original cost estimate for Zimmer was $230 million.When
35 Federation of American Scientists
A Critical Examiniation of Nuclear Power’s Costs
cancelled, the plant’s sunk costs were said to total about $1.8 billion, almost equal to
the owner’s net worth. To salvage some value from the project, Zimmer was converted
into a natural gas-fired facility.
Economic Considerations
e evolution of reactor designs and regulations undermined the reliability of the
original nuclear cost projections. e construction risk was further amplified by
economic conditions that slowed the demand for electrical power, increased
commodity costs, weakened utility balance sheets, and raised the cost of capital. Many
first and second generation plants were cancelled when it became evident that the total
debt for the plant would approach the owners’ book value (e.g., Zimmer). In other
cases where a utility’s capital adequacy was sufficient to complete a construction
project, capital markets were oen not deep enough to fund the risk.
It is oen forgotten just how poorly positioned utilities were to take on the
risk of building nuclear power plants. Entering the 1970s, the capital adequacy at
electric utilities was well into decline with balance sheets marked by general illiquidity.
Beginning in 1966, current liabilities at utilities exceeded current assets. By 1974, in
the aermath of the oil embargo, the combination of weaker earnings performance and
continued heavy bond financing drove up the spread or “risk premium” on interest rates
paid by electric utilities compared to industrial firms in the same bond-rating category.
By 1975, power demand had dropped as electric rates rose due to higher fuel costs.
With sector profits down some 25 percent and excess generating capacity in the system,
utilities began to trim capital spending centered on the more expensive nuclear
construction programs. By 1979, the credit window for nuclear plants had effectively
closed. Lenders were increasingly cautious about financing utilities.
To summarize, forecasts for the first generation and second generation of
nuclear plants were systemically wrong. But even if they were right, economic
conditions had changed for many companies. e demand curve shied. Cost of
capital increased. By 1975, most utilities lacked the balance sheet needed to build a
licensable plant.
Federation of American Scientists 36
The Future of Nuclear Power in the United States
February 2012
irty years later, with this experience in mind, it is reasonable to ask what, if anything,
has really changed:
Are utilities better able to forecast and manage costs?
Do they have the balance sheet to shoulder that risk?
If not, who should bear the risk?
Do the many billions of dollars of federal loan guarantees make sense from a
taxpayer perspective?
Would taxpayers be better protected with credit deposit fees? How much
should those fees be?
What price on carbon emissions would help even the playing field for nuclear
power plants?
What financing models are effective? Is the United States going about
financing nuclear power plants the wrong way? How can it do a better job?
Will Future Nuclear Power Plants Follow a Similar Cost Trajectory?
e First and Second Generation nuclear plants teach the importance of regulatory
stability. However, regulatory stability alone is not sufficient for reliable cost estimates.
While today the reforms of combined construction and operating license (COL)
application and preapproved design certification are expected to dampen cost
escalations, safety regulations are not static nor can they be. e reformed process is not
all encompassing as some might think. For example, NRC’s design pre-approvals only
apply to engineering documents submitted for review. But most design details are not
presented for such review (e.g., detailed design and field engineering).
Indeed, substantial questions continue to be raised concerning one or another
element of certified designs. e original Design Certification rule approving the
Westinghouse AP1000 design was issued on January 27, 2006. Yet, substantial design
details were not presented for NRC review at that time. According to the NRC, the
largest review effort centers on the expected design changes required to address site
features and other design changes identified aer certification. ose design changes are
37 Federation of American Scientists
A Critical Examiniation of Nuclear Power’s Costs
substantial and include a redesign of the pressurizer, a revision to the seismic analysis to
allow an AP1000 reactor to be constructed on site with rock and soil conditions other
than the hard rock conditions certified in the AP1000 design certification review
(DCR), changes to the instrumentation and control (I&C) systems, a redesign of the
fuel racks, and a revision of the reactor fuel design. Another area requiring attention
will be the review of design acceptance criteria (DAC)-related items, such as the
technical reports on human factors engineering (HFE), the I&C design, and piping.
About a year later, the vendor submitted an application to amend the
AP1000’s DCR and Revision 16 of the AP1000’s design certification design. Revision
16 contains changes proposed in technical reports, some of which have not yet been
reviewed by the NRC staff. By February 2008, two years following certification,
Westinghouse submitted 122 technical reports for NRC review. Although submitted
as part of the Bellefonte Nuclear Power Plant’s COL pre-application phase, these
technical reports apply generically to the remaining COL applications that intend to
reference the AP1000 design. Six months later, additional changes were submitted.
Design Certification does not eliminate the need for detailed engineering design
review, nor does it preclude design revisions.
Learning Curve in Engineering and Construction
Despite reconstituting the regulatory process, nuclear technology is no less complex
today. While a stable regulatory process may reduce the potential for rework caused by
changing requirements, it does not reduce the complexity of power plant design and
construction. Also, regulatory stability does not equip companies with experience in
planning and building a plant. As with previous generations, companies building a
nuclear power plant today have no contemporaneous experience. Every plant built will
be a “first-time” and an entirely new learning curve for each individual and
organization. at learning curve applies as much to estimating commodity prices,
quantities, and schedule as it does to forecasting the processes associated with
procurement and construction.
Although there is no evidence of improvements in cost estimation, there is
contemporary evidence of rising cost estimates similar to what happened in the first
generation. Nuclear vendors in the early 2000s were quoting nuclear electricity
generation’s costs below $1,500/kWe. Within a few years, a utility consortium building
a General Electric advanced reactor design priced two units at $1,611/kWe. A Florida
Federation of American Scientists 38
The Future of Nuclear Power in the United States
February 2012
company subsequently estimated a two-unit Westinghouse project would come in at
$2,444-$3,852/kWe. e utility reported costs for materials, equipment, and labor had
risen more than 50 percent. For all-in costs (i.e., transmission improvements, site
enhancements, land, and risk) the project climbs to $3,108-$4,540/kWe. e company
then dialed in 11 percent carrying charge and cost escalation allowances for a final tally
of $5,780-$8,071/kWe.
e CPS Energy project history is also instructive. In June 2006, a consortium
of companies announced plans to build two more reactors at the South Texas Project
site for an estimated cost of $5.2 billion. NRG, the lead company, made history by
becoming the first company to file an application with the NRC. CPS Energy, a
municipal utility, was one of its partners. In October 2007, CPS Energy’s board
approved $206 million for preliminary design and engineering. In June 2009, NRG
revised the estimate to $10 billion for the two reactors, including finance charges. A
few weeks later, this estimate rose to $13 billion, including finance charges. Later that
year, the estimate reached $18.2 billion, which was reportedly at the break-even point
with natural gas, and the power would not be needed until about 2023. Whereas the
reactors would require upwards of 10 years to build, price-competitive natural gas
could be on-line in three to five years. CPS would reportedly spend about $1 million
per day on the nuclear project, which would not be needed for some 20 years. Moody’s
had downgraded CPS’s outlook to negative. When the municipal exited the project, its
credit rating was lied to stable.
is cost experience is not unique to the United States. Faced with stringent
greenhouse gas (GHG) emissions standards under the Kyoto Protocol, Finland
committed in 2004 to building Olkiluoto, the first Generation-III+ reactor, to enter
production in 2009. Areva, based in France, won the contract to build the first
Evolutionary/European Pressurized Water Reactor (EPR). At $3,000/kWe (2004), the
plant was considered a “loss leader,” similar to the “turnkey plants” of the 1960s. By
2007, project costs escalated 50 percent and construction schedule delayed three years.
Construction cost projections doubled by 2008 mostly due to commodity cost
escalations and weakening dollar to euro exchange rates in the intervening period. At
the time, the plant was over $2 billion over budget.
As the project proceeded, quality assurance issues began to emerge. For
example, late 2008, the Finnish Nuclear Safety Authority (STUK) questioned the
supervision and “safety culture.” STUK reported mandatory welding guidelines were
39 Federation of American Scientists
A Critical Examiniation of Nuclear Power’s Costs
not developed until months aer welding of the reactor began and that a contractor
instructed workers not to report quality problems to inspectors. Other QA concerns
implicated the steel liner of the Olkiluoto reactor containment, and the
remanufactured primary coolant piping.
Quality assurance deficiencies contributed to pushing back the delivery
scheduled for the fourth time in two years out to 2012. In mid-2009, the latest estimate
of construction costs reached EUR5.5 billion, more than twice the price of EUR 2.5
billion originally presented. By the end of 2009, more weld faults led to STUK issuing
a “stop work” order until the issue is resolved. In mid-June 2010, Areva set aside some
EUR400 million ($491 million) for the Olkiluoto 3 construction project leading to an
operating loss for the first half of 2010.
In Canada, the Ontario government suspended its nuclear development
project in mid-2009 when project bids came in around $C26 billion, some three times
higher than what the province expected to pay. In 2007, the Ontario Power Authority
assumed nuclear project costs of $C2,900/kWe installed. e Power Authority had
quoted a break-even with natural gas of around $C3,600/kWe installed. e $C26
billion price tag would have constructed two 1,200-megawatt Advanced CANDU
reactors (~$C10,800/kWe installed). Analysts have attributed the cost increases to
labor, commodity, and vendor risk premiums.
Who bears the risk going forward, and who should bear it?
Risk is traditionally borne by those who benefit from the investment. With major
capital projects, that risk is assumed by those holding the securities and is priced into
credit availability and capital cost for the financial structure employed.
Construction Risk and Evolving Project Finance Structures
Construction risk manifests itself in two ways: (1) uneconomic construction costs that
cannot be recovered, and (2) credit events associated with a range of scenarios from
technical default through bankruptcy. For the first and second generation plants,
ratebase disallowances oen led to claims against the vendor. When construction
projects overrun, both parties are oen damaged. Both parties were at risk.
e first generation and second generation plants were funded by mortgage
bonds financed by utility balance sheets and paid for by ratepayers. As the cost of
Federation of American Scientists 40
The Future of Nuclear Power in the United States
February 2012
capital through the 1970s rose, ratepayers paid the price of the flawed cost estimates.
Today, while the cost of capital is lower at the moment, credit availability is
constrained by other market effects, notably persistent under-performing assets, a
weakened banking system, and rising sovereign risk issues. Recognizing the limitations
of today’s credit markets, some companies would like to follow this model. e
problem is that ratepayers are more likely to be resistant. Balance sheets for utilities are
less able to handle the much larger construction risk of today’s nuclear units.
e Nuclear Energy Institute argues that electric power companies are too
small to generate the financing capability or strength needed to finance nuclear power
projects on their own. NEI considers it essential that project partners (e.g., nuclear
vendors) and limited investment incentives provide the necessary backup to otherwise
limited balance sheets. e federal loan guarantee program helps serve this purpose.
Power projects today oen use limited or non-recourse financing through a vehicle
company (separately incorporated) whose sole purpose is to bring assets online.
Creditors share a portion of the venture’s business risk and capital funding is obtained
strictly for the project itself. In a non-recourse finance model, there is little expectation
that the corporate or government sponsor will fully guarantee the debt. e vehicle
company’s equity might be entirely held by the parent utility. In other cases, equity
might be distributed among several companies in an attempt to diversify risk as well as
take advantage of some unique elements of the partners. A nuclear vendor might
participate to further share risk. To help with local interests and municipal preferences
in some jurisdictions, a municipal or power cooperative might be invited to a minority
position. e problem is that few lenders will be willing to provide non-recourse
financing because their balance sheets are not much stronger than their utility partners.
Securitization is one approach for diversifying the risk and has been advocated
by some investors in recent years. Like the collateralized debt obligations, securitization
combines contractual debts and sells the debt as a bond or security. In the utility sector,
securitizations were originally used to address stranded costs following deregulation. It
has also been used to reimburse utilities for storm restoration costs following two active
hurricane seasons in the U.S. in 2004 and 2005. However, securitization rates are down
today, a victim of failed risk management policies associated with collateralized debt
obligations that led to the collapse of Bear Stearns and Lehman Brothers.
e bottom line is that, without back-end guarantees, even the most
innovative structures and risk sharing among participants are challenged by the weak
41 Federation of American Scientists
A Critical Examiniation of Nuclear Power’s Costs
balance sheets of utilities and vendors alike. at leaves the government as an essential
player in any nuclear renaissance.
Do the many billions of federal loan guarantees make sense om a taxpayer perspective?
On February 16, 2010, $8.3 billion in federal loan guarantees were awarded for two
new reactors to be added to Southern Company’s Vogtle site in Georgia, conditional
until the project is awarded a combined construction and operating license from the
e DOE budget proposal for 2011 requested $36 billion in such loan
guarantees, up from the current authority of $18.5 billion, with the objective of
underwriting the construction risk for 10 nuclear power plants. is loan program is
but a small component of loan guarantees totaling some $1.1 trillion of which some
$77 billion in loan authority is directed at clean energy projects, those that emit
relatively few greenhouse gases.
In contrast to the $787 billion economic stimulus package and the $75 billion
loan modification program, loan guarantees do not automatically lead to a
disbursement of taxpayer money. But guarantees are a subsidy and have the potential
for payout. In the current fiscal crisis, it is unclear just how much guarantee the
government can provide going forward and its willingness to pay should a default
In June 2005, CBO produced a cost estimate for the Senate Committee on
Energy and Natural Resources related to their consideration of revisions to the 2005
Energy Policy Act.3 is estimate covered a variety of loan guarantees under
consideration, including projects involving coal degasification, renewable energy,
ethanol, and nuclear plant construction. CBO observed the subsidy cost of loan
guarantees could vary widely depending on the terms of the contracts and the financial
and technical risk associated with different types of projects. Quoting Standard and
Poor’s, CBO estimates the cumulative default risk for projects rated as speculative
investments can range from about 20 percent to almost 60 percent, depending on a
project’s cash flows and contractual terms. CBO defines the term “subsidy” to mean the
net present value of the anticipated cost of defaults, net of recoveries. A $2 billion loan
3 Congressional Budget Office, “Cost Estimate for S.10, Energy Policy Act of 2005,” June 9, 2005, http://pdocs/64xx/doc6423/s10.pdf
Federation of American Scientists 42
The Future of Nuclear Power in the United States
February 2012
guarantee for a nuclear construction project was estimated to have a 30 percent subsidy
associated with a default event. at is, this would be about $600 million.
Two years later, CBO provided a revised cost estimate to the Senate
Committee on Energy and Natural Resources related to their consideration of the
Energy Savings Act of 2007.4 In its analysis, CBO noted the “significant technical and
market risks” presented challenges and constraints estimating the subsidy making it
“likely that DOE will underestimate than overestimate” cost of insuring against credit
In May 2008, CBO analyzed the effects of Energy Policy Act incentives with
special attention to the production tax credit and a loan guarantee program.5 e tax
credit provides up to $18 in tax relief per megawatt hour of electricity produced at
qualifying power plants during the first eight years of operation. CBO assesses that
generating electricity with nuclear technology would be roughly 35 percent more
expensive than using conventional coal technology and 30 percent more expensive than
using natural gas capacity. CBO concludes that investment in nuclear capacity would
be unlikely in the absence of carbon dioxide charges and Energy Policy Act incentives.
CBO explains the reason for the deficit growth: Medicare and Medicaid continue to
increase by two to three times the rate of everything else in the economy. Unchecked,
these line items will eventually take up every dollar of tax revenues raised, leaving no
money for anything else, including national defense.
e Federal government’s ability to subsidize energy projects of this
magnitude is limited. Actual debt-funded spending currently exceeds receipts, resulting
in monthly debt additions of approximately $34 billion. As this is a 9 percent
compound annual growth rate, such a debt addition rate requires GDP to increase by
about an unrealistic 7 percent annually just to stay at about 100 percent debt/GDP in
2020. Even if U.S. GDP increases by 2.5 percent each year, this will result in a 2020
Debt-to-GDP ratio of 151 percent, comparable to present-day Greece. At such levels,
loan guarantees are as much subject to default risk as any other obligation. In an era
when natural-gas fired capacity can be brought on-line in half the time and at half the
4 Congressional Budget Office, “Cost Estimate of S.1321, Energy Savings Act of 2007,” June 11, 2011,pdocs/82xx/doc8206/s1321.pdf
5 Congressional Budget Office, Nuclear Power’s Role in Generating Electricity, A CBO Study, May 2008,pdocs/91xx/doc9133/05-02-Nuclear.pdf
43 Federation of American Scientists
A Critical Examiniation of Nuclear Power’s Costs
cost of nuclear generation, it is difficult to make a case that the Federal Government
should issue bonds to finance loan guarantees for nuclear construction.
What price on carbon emissions would help even the playing field for nuclear power plants?
Some advocates suggest a significant investment in nuclear power plant capacity can
make an efficient and meaningful contribution to greenhouse gas reduction. As
reported by the U.S. Energy Information Administration (EIA), greenhouse gas
(GHG) emissions are expected to grow about 0.3 percent annually through 2035,
down from 0.8 percent annually (1980-2008). At this rate, pre-recession emissions will
not be reached until 2025. In addition to setting back the need for additional electricity
generation, the recession also set back GHG imperatives. In 2008, power generation
represented 41 percent of the emissions. Coal’s 48 percent market share accounts for 82
percent of power sector CO2 emissions. However, coal-fired generation is currently
being displaced by natural gas capacity, which would emit roughly half the amount of
GHG emissions. While substantial nuclear construction might further cut into the
coal plants’ emissions, it won’t begin to displace coal for some years to decades to come.
To make the economic argument for nuclear requires a reliable forecast of
carbon prices. Just as 1960s-era forecasts of nuclear construction costs were unreliable,
it is just as difficult to forecast the value of carbon emission offsets 10 years from now.
Historic data from the Chicago Climate Exchange demonstrates extreme volatility in
carbon markets with prices ranging by factors of 20 or more. Before its acquisition by
the Intercontinental Climate Exchange (ICE), the April 2010 open interest, or
intensity of trading, in the Chicago Climate Exchange’s futures and options contracts
totaled some 114,064 contracts. By comparison, in April 2010, CME Group, the
leading firm for futures and options trading and risk management, reported a volume of
5,740,439 futures contracts for natural gas physically priced at Henry Hub, the pricing
point for natural gas futures. And even as deep and liquid as the natural gas market is,
it becomes illiquid for deliveries about a year or more out. e high volatility of carbon
markets and low open interest means any forecast of carbon-offset prices for next year is
highly speculative - carbon prices 10-30 years from now are mere guesses.
Federation of American Scientists 44
The Future of Nuclear Power in the United States
February 2012
What financing models are effective? Is the United States going about financing nuclear
power plants the wrong way? How can the United States do a better job?
e ability to finance any long-lived asset requires a viable bond market. Financing
models may apportion the risk in various ways but the cost of this risk capital is
ultimately “baked” into the cost of capital. To do a “better job” requires policies aimed
at fostering a stable, deep, and liquid bond market. Such markets require stable Federal
Reserve policies, low corporate tax and capital gains rates, a strong currency, and stable
growth. ese conditions favor long-term investments in projects because the cash
streams are more predictable and the returns are commensurate with the risk. Under
such conditions, credit markets are subject to fewer artificial constraints, less
concentration (buy or sell side), and less uncertainty regarding sovereign intervention.
e deficit spending policies in place today do very little to create a viable
bond market. Presently, the United States increasingly generates most of its capital
through monetization policies that are ending with the Quantitative Easing program in
the summer of 2011. e Federal Reserve Bank is the largest holder of U.S. debt and
sets interest rates. e ability to finance nuclear power plants, therefore, is ultimately
sensitive to the quality of the balance sheet of the United States. As the U.S. deficit
grows, the supply of U.S. Treasury bonds will grow. As U.S. Treasury supply grows,
corporate debt is either displaced or priced at a premium. In this market in which the
United States commands an ever larger market share of long-term debt, the ability to
finance nuclear power plant construction outside of a Federal loan guarantee will be
increasingly limited.
In summary, policies at the state and federal levels might be engineered for
efficiency and fairness well into the future. But company executive teams decide and
act today based on their perception of risk, their liquidity, and their available risk
capital. eir decisions oen make moot the analysis and speculations of market
conditions in the distant future. With high deficits under current Federal spending
policies, utility balance sheets will continue to suffer and remain poorly positioned to
commit to high cost construction projects that won’t come on line for a decade in the
45 Federation of American Scientists
Licensing and Regulation of U.S Nuclear Power Plants
Chapter 3
Licensing and Regulation of U.S. Nuclear Power
by Albert V. Carr, Jr.
The regulation of the United States nuclear power industry arguably is the most pervasive regulatory system of any in the world. Development, implementation and conduct of
that regulatory environment have followed a difficult and convoluted path over the years. In
each stage of its development the nuclear regulatory and licensing process has reflected substantial tensions among its various constituents. The first was the conflict between the national
security interests of the nuclear weapons program and its absolute governmental control of
nuclear materials and technology and of those private entities that sought to develop the peaceful use of nuclear energy for generation of electric energy.6 Those tensions were to a major extent relieved by the Congress with passage of the Atomic Energy Act of 1954,7 which established a framework for development of the domestic nuclear power industry. The 1954 Act
served as the basis for the Atomic Energy Commission’s (AEC) bifurcated licensing process
that set the framework for the licensing of the 104 commercial nuclear power reactors now
operating in the United States. That licensing process was anything but smooth. Substantial
tensions existed among the AEC and its successor agency, the Nuclear Regulatory Commission (NRC), the applicants/licensees and members of the public. Actions taken thereunder
arguably increased substantially the costs of plants coming into service and resulted in the
dearth of orders for new domestic plants for more than 25 years. As the present generation of
plants came into service the NRC put into place a regulatory environment that governs all
aspects of the day-to-day operation of the plants that it has licensed. Recent legislative and
regulatory changes to the earlier licensing process have hopefully cured many of its defects and
may well lead to an environment in which the domestic nuclear power industry can revive.
George T. Mazuzan and Samuel Walker, Controlling e Atom, e Beginnings of Nuclear Regulation,
1946-1962, University of California Press, (1984), (hereinaer Mazuzan and Walker) at Chapters I-III.
Atomic Energy Act of 1954, 42 USC Sec 2011 et seq.,Public Law 83-703 (68 Stat 919).
Federation of American Scientists 46
The Future of Nuclear Power in the United States
February 2012
e Early Days
It is of course no secret that the initial uses of nuclear energy were in weapons development. However, from the early days following the end of World War II, increasing
numbers of policymakers and scientists sought to find alternate peaceful uses for
nuclear energy.8 ough a number of uses were considered, including nuclear-powered
merchant vessels and aircra, ultimately the major usage of nuclear energy for peaceful
purposes has been in the generation of electricity. e story of how that came to be is
interesting, indeed fascinating, in a number of respects.
To begin, the initial task was to overcome the reluctance of the stewards of the
nuclear weapons program to share nuclear technology and materials with civilian entities. Once those concerns were met, with considerable prodding from the then-nascent
industry, if not alleviated at least accommodated in the Atomic Energy Act of 1954
(e “’54 Act” or the “Act”), the task faced by the agency was formidable, as it had to
put into place an entire licensing and regulatory regime to deal with a completely new
industry — nuclear power plants.9
At the outset, the AEC faced a quandary, as the Act placed squarely on it the
responsibility both for developing the peaceful use of the atom and at the same time
regulating its usage. is potential conflict-of-interest would plague the agency though
it was eventually addressed by setting up a separate office within the agency – the Regulatory division – in a location with its own management and legal offices, particularly as
its licensing actions became more controversial and received greater scrutiny by the
public. is situation continued until 1975, when Congress enacted legislation that
separated the functions, establishing the Nuclear Regulatory Commission as a separate
and independent regulatory agency.10
Licensing the Initial Plants
e ’54 Act put in place a licensing process that while understandable in the context of
the early days of the nuclear power industry, ultimately proved to be an impediment to
its development. Put briefly, the licensing process established by the ’54 Act and its im8
Mazuzan and Walker at Chapters I, III.
9 Chapt III, Pp.60ff.
Energy Reorganization Act of 1974, 42 USC Sec 5801 et seq., Public Law 93-438 (88 Stat. 1233).
47 Federation of American Scientists
Licensing and Regulation of U.S Nuclear Power Plants
plementing regulations11 put in place a two-stage licensing process. First, the applicant
was required to apply for a construction permit, based primarily on a design concept
that would authorize construction of the plant. at application carried with it a mandatory public hearing. e plant was then constructed, with most of the design done
during the construction phase. Following completion of construction the utility then
had to file an application for a license to operate the plant. at application, while not
requiring a hearing, did require the agency to offer an opportunity for a hearing. At the
outset of nuclear power plant licensing by the Commission, the focus understandably
was on the safety aspects of construction and operation of nuclear plants, as that was
the focus of the Commission’s organic statute.
e AEC’s licensing hearings for both Construction Permits and Operating
Licenses were conducted before a three-member Atomic Safety and Licensing Board
(ASLB), made up of two technical members – generally one was a nuclear or mechanical engineer and the other an environmental scientist – and a lawyer, who chaired the
panel. e proceedings were trial-type hearings with the procedural rules based upon
the Federal Rules of Civil Procedure. e procedures placed certain standing and substantive requirements on those who sought to participate as a party. ey further provided for the full range of discovery practices, including depositions, document requests and interrogatories. At hearing, witnesses were presented by applicants, staff and
sometimes interveners. Full cross-examination was allowed, and the Licensing Boards
oen participated in the questioning of witnesses. Following completion of the hearings, the ASLB would issue its decision, which then went through the Commission’s
appeal process.
To implement that process the Commission had established a body known as
the Atomic Safety and Licensing Appeal Board (ASLAB). ASLAB’s task was to monitor
and review proceedings before the ASLBs to protect against legal error that could lead to
reversal on appeal. This process worked reasonably well for the first generation of plants
licensed, such as Duke Power Company’s three-unit Oconee Nuclear Station.12
10 CFR Part 50, Domestic Licensing of Production and Utilization Facilities.
12 Duke’s Oconee Nuclear Station consists of three Babcock&Wilcox Pressurized Water Reactors, each of
850MWe output, originally licensed in the late 1960s and early 1970s. Oconee was the first nuclear plant
in the United States to reach a total electric output of 500 million megawatt-hours.
* Calvert Cliffs Coordinating Committee v. United States Atomic Energy Commission, 449 F.2d 1109
(D.C. Cir. 1971).
Federation of American Scientists 48
The Future of Nuclear Power in the United States
February 2012
It was not long, however, before events began to cause the licensing process to
unravel and with it usher in the long cessation in nuclear power plant orders in the
Unites States. e first of these events came with the passage by Congress in 1969 of
the National Environmental Policy Act (NEPA), which required all Federal agencies to
consider the impacts of their actions on the environment, to include alternatives to the
proposed action, which alternatives included not taking the action proposed. Agency
compliance was to be demonstrated by preparation of an Environmental Impact Statement to guide and document that consideration. In the summer of 1971 the Court Of
Appeals for the District of Columbia issued its Calvert Cliffs decision, which overturned the regulations adopted by the AEC to implement NEPA.[*] at decision embedded environmental evaluations and analyses into the AEC’s licensing processes.
ough there were some initial problems, the requirements of NEPA, as interpreted by
the Calvert Cliffs court, were integrated into the AEC’s licensing activities and the
nuclear power industry seemed to have a bright future.
is was a time in which the demand for electricity was projected to increase
in exponential fashion. In response, electric utilities, which have the absolute obligation to plan, construct and operate their systems to meet present and foreseeable load,
were in search of additional generating capacity to meet that forecasted load. Due at
least in part to federal energy policy, which restricted access to some fuels for electricity
generation,13 a number of those utilities chose the nuclear option. As a result in the early
1970s the AEC was receiving one or more applications for construction permits per month,
and the AEC was processing those applications on a rigid schedule which was amended
only under exceptional circumstances. Under these circumstances power plant licensing
appeared to move forward smoothly.
13 Indeed, as federal policy actively discouraged the use of certain fuels for generation of electricity it can be
argued that federal policy thus encouraged the use of nuclear generation. Prior to the 1970s many utilities,
particularly in the South, were using significant amounts of natural gas to generate electricity. However, in
the late 1960s, shortages of natural gas began to appear as a consequence of federal wellhead natural gas
price controls, and by the winter of 1969-1970 these shortages became acute. e FERC’s predecessor
agency, the Federal Power Commission, required in the early 1970s interstate natural gas pipelines to curtail delivery of gas to electric utilities that would have been used to fuel boilers for electric generation in
order that the gas could be used to serve higher-priority needs – residential and small commercial loads.
For its part, Congress passed the Energy Supply and Coordination Act of 1974 – signed into law by the
president --that, among other things, required the relevant federal officials to prohibit electric generating
stations from using natural gas or petroleum products as boiler fuel and further ordered that existing natural gas and petroleum-fueled generating plants be converted to coal if possible. While these events were
taking place, in 1970 the Clean Air Act was enacted into law and subsequently amended in 1977,
49 Federation of American Scientists
Licensing and Regulation of U.S Nuclear Power Plants
All was not how it seemed, though. A series of events occurred in relatively short
order that imposed unforeseen, and at times insurmountable, stresses on the licensing process. The first of these was well beyond the ability of either the AEC or the nuclear power
industry either to anticipate or cure. In the early 1970s the first Arab Oil Embargo occurred.
This event caused, among other things, a rapid and substantial increase in energy prices
across the board, including in the cost of electricity. This increase in costs led in turn to reduced demand for electricity, which required applicants to stretch out the schedules for
plant construction and operation.14 is extension of construction schedules occurred
at a time of great financial vulnerability as in the 1970s and early 1980s inflation was
rampant and interest rates were at an all-time high. is resulted in substantially increased carrying costs for the monies borrowed to finance plant construction. ese
events were out of the control of the AEC and the applicants.
is was not the case, however, with problems associated with design and construction of the newer generation of plants that were the subject of licensing scrutiny.
Whereas the earlier plants were all sized in the 800-900 megawatts (electric) (MWe)
range, beginning in the early 1970s vendors increased the size of the plants by about a
third, to the 1100 MWe–1250 MWe range. e designs for these new reactors, while
sufficient to support the applications for and issuance of Construction Permits, required substantial further work during plants’ construction. is in itself was enough to
cause construction difficulties, but construction supervision and management issues
exacerbated design and construction difficulties. Even leaving aside considerations of
design, construction of a nuclear power plant is an enormously complicated endeavor.
e work force is substantial, in most instances consisting of a myriad of contractors
which placed uncertain limitations on the use of coal as a boiler fuel. Against this backdrop and during
this timeframe, then, it is not surprising that a large number of nuclear plants were ordered – 128 from
1971 through 1974.
Traditionally a utility may put into its rate base, and recover with a return, costs prudently incurred in
constructing facilities necessary to serve its customers; that is, facilities which in the vernacular are “used
and useful.” Based on the fact that the original forecasts were no longer valid, so the plants would not be
needed when originally projected, decisions to extend construction schedules were understandable; adhering to the original construction schedules would have resulted in capacity that was not “needed” when first
available, thus raising questions as to whether it could be included in the utility’s rate base. See, e.g., 1
A.J.G. Priest, Principals of Public Utility Regulation, 139-190 (1969); Richard J. Pierce, Jr., e Regulatory
Treatment of Mistakes in Retrospect: Cancelled Plants and Excess Capacity, 132 U. Pa. L. Review 497, pp.
511-517 (1984).
Federation of American Scientists 50
The Future of Nuclear Power in the United States
February 2012
and subcontractors.15 A number of the utilities that had filed applications for, and been
granted, construction permits had little if any experience with managing major power
plant construction projects. us, those utilities tended to contract for management as
well as construction services. Many of these design and construction issues were resolved with the NRC during construction, but in many instances required redesign and
additional construction work. ese matters, with others, particularly allegations associated with management issues, oen became issues at the Operating License proceedings. Nuclear construction work is an intensely monitored, highly documented activity. at is, the recordkeeping and paperwork requirements are formidable and need
precise controls to demonstrate to the regulator that all applicable Quality Assurance
standards have been met. Strong and effective management control of these issues was
an absolute necessity, and lack of that could be a contributor to delays in construction,
licensing and operation of proposed nuclear plants.
e problems outlined above with respect to extending design and construction schedules and rapidly increasing costs had a serious adverse effect on the nuclear
power industry. Not only were construction schedules extended, but a number of proposed plants were cancelled, both before and aer issuance of construction permits.16
However there was another factor that also contributed to increased costs and delays.
In March 1979 the accident at ree Mile Island occurred. ough the factors already
discussed were contributing to industry decline, the TMI accident brought the NRC’s
licensing activities to an absolute halt. During the hiatus, the NRC reviewed many design elements at plants in the construction phase and, notwithstanding that design and
installation of plant systems had received NRC approval, required redesign and construction. is was of and in itself incredibly expensive. When added to the alreadyinflated costs of plants in the licensing pipeline, the effect of the increased costs for redesign and added delay was stunning, far exceeding the initial estimates for plant construction.
At peak work times, the number of construction workers on site at a typical nuclear project could
number in the multiple thousands with hundreds of contractors and subcontractors.
In the final analysis, more than 100 proposed nuclear plants were cancelled between 1974 and 1982.
See Pierce at pp. 497-499.
51 Federation of American Scientists
Licensing and Regulation of U.S Nuclear Power Plants
In fact, a number of the problems associated with design and construction
issues, as well as Quality Assurance program sufficiency and documentation came to
the fore when many plants filed applications for operating licenses. Public interest
groups contested those applications. Allegations of management pressure to comply
with construction schedules at the expense of quality assurance measures required attention to both management and to quality assurance documentation. In many instances lack of proper quality assurance documentation combined with constructionrelated issues presented significant licensing challenges. In most cases these issues added
delays and cost to plant completion. In two instances, nuclear plant applicants abandoned their attempts to obtain operating licenses for nearly complete plants and converted them to other fuels for the generation of electricity. One was Consumer Power’s
Midland, Michigan, plant. In 1984 Midland, a two-unit PWR plant, had been under
construction for more than 15 years, was only about 85 percent complete and Consumers had invested almost $4 billion in the project. Faced with sinking and cracking of
buildings due to subsidence on the site, quality assurance issues, and TMI-related design uncertainty, Consumers cancelled the nuclear plant and converted Midland to
what was then the largest gas-fired cogeneration plant in the world. at plant went
into service in 1991. e second was Cincinnati Gas and Electric’s Zimmer plant. In
1983 Zimmer had been under construction for about 10 years and was claimed to be
97 percent complete. Its owners had invested about $1.8 billion in the plant, and because of significant quality assurance issues, the NRC had issued a stop work order in
1982. e plant’s owners received an estimate of an additional $1.5 billion to complete
the plant and obtain the operating license. Zimmer was cancelled in 1983 and converted to a coal plant, at a cost of just over $1 billion, and went into service in 1991.
In the same view, two examples will suffice to demonstrate the magnitude of
the cost overruns. Long Island Lighting and Power’s Shoreham Plant, a single-unit
General Electric plant in the 800 MW(e) range, was initially estimated to cost in the
neighborhood of $217 million when its construction began in 1968. Continued licensing delays and NRC redesign and construction requirements as well as delays in operation caused the final cost for Shoreham to be in the neighborhood of $6 billion by the
time it was ready to operate in 1984.17 e Wolf Creek plant, located in Kansas and
17 ough Shoreham construction was completed and fuel was loaded to permit law-power testing, Shore-
ham never received an operating license and thus never went into commercial operation. Its ownership
ultimately was transferred to the State and the plant was decommissioned.
Federation of American Scientists 52
The Future of Nuclear Power in the United States
February 2012
owned by three Kansas utilities, a single Westinghouse unit of approximately
1200ME(e) capacity, was initially estimated to cost about $525 million when its application for construction permit was filed in 1973. By the time its application for an operating license was filed more than 10 years later its final cost was more than $3 billion.
Neither Wolf Creek nor Shoreham was a particular outlier in terms of final
costs of construction far outpacing the original estimates. Wolf Creek (as did the other
plants completed with significant cost overruns) presented its State regulators with
difficult issues to resolve in placing its investment into rate base so the costs incurred
could be recovered from its customers. If the regulators allowed all the money invested
to build the plant to be placed into rate base, retail customers could see a tripling or
quadrupling of their rates.18 us, state regulators adopted proceedings, generally referred to as “prudence” hearings, to determine how much of the investment should be
disallowed and how much should be recovered from ratepayers. Again, Wolf Creek is
illustrative. Aer the State legislature and the State courts were finished with their
evaluations,19 the Wolf Creek investment was treated as follows: A total of $183 million was disallowed on the grounds that these expenditures were imprudently made.
No recovery of this amount or return on this investment was permitted. Of the remaining investment, no return was then permitted on $944 million as the Kansas Commission concluded that the capacity and energy represented by this investment was excess
to the then-current needs of the system and thus was not “used and useful.” e Commission did determine, however, that as system load grew into the capacity over the
years that the company could file additional rate cases to seek a return on the investment. e Kansas Commission also determined that portions of Wolf Creek capacity
represented “excess economic capacity” and disallowed recovery of $266 million on
this basis.20 Clearly this was not the desired outcome for Wolf Creek or many of the
18 See Kansas Gas and Electric Company v. State Corporation Commission of Kansas, 720 P2d 1063 (Kan.
1986) at pp. 1069-70.
19 Ibid.
20 Briefly, “excess economic capacity”is determined based on an “economic evaluation” of the plant.
theory is that even if excess physical capacity exists, the plant should be evaluated against a hypothetical
competing method of generation – in the case of Wolf Creek a hypothetical coal plant was used – built at
the same time. If the cost of capacity from Wolf Creek was found to be higher than the cost of a similar
block of capacity from the hypothetical coal plant, then the difference represented “excess economic capacity” and should be disallowed. See Ibid, pp. 1084-87.
53 Federation of American Scientists
Licensing and Regulation of U.S Nuclear Power Plants
other nuclear plants – and they were numerous21 — across the country that received
similar treatment. Such disallowances and other variances of recovery of shareholder
dollars invested by utilities hardly served as an incentive for utilities to invest in additional nuclear power plants.22 Taken together, this confluence of events effectively put
an end to the ordering of new nuclear power plants in the United States for a period of
more than 30 years.
A Nuclear Revival
Over the interviewing years, the industry, the Congress, and the NRC have taken a
number of steps to aid in reviving the domestic nuclear power industry. A combination
of enabling legislation and regulatory action has put into place a framework for licensing and regulation of a new generation of nuclear plants that, it is hoped, will lead to a
“nuclear revival.” There were and are a number of drivers behind this effort. One is that
though the early operational experience of the 104 reactors that ultimately were brought
into service – about 20 percent of our present domestic generating capacity – was not
particularly good, over the past two decades their performance has been exceptional. Another is that there is an increasing need for additional base load generating capacity, which
matches well with the capabilities and operating characteristics of nuclear plants. Also,
there are likely to be increasing environmental restraints on carbon emissions; an operating nuclear power plant does not emit greenhouse gases. Moreover, nuclear power plants
hedge against volatility in prices of competing fuels, particularly natural gas. Finally,
additional nuclear power plants will aid in domestic energy security.
An examination of the legal and regulatory changes now in place shows that
they are intended to remove many of the previous uncertainties from the licensing
process to try to spur interest and investment in new domestic nuclear power plants. In
brief, they include a new licensing process before the NRC, federal loan guarantees,
federal “insurance” against licensing delays and credits for certain electricity generated
by nuclear energy.
21 See generally Pierce,132 e University of Pennsylvania Law Review 497 (1984).
22 e issues associated with recovery of investment in nuclear plants placed into service were not the only
matters that vexed financial regulators. Utilities had invested more than $10 billion in nuclear plants that
had been cancelled. ose investments were subjected to a variety of regulatory treatments. Id at pp 498500.
Federation of American Scientists 54
The Future of Nuclear Power in the United States
February 2012
e NRC’s new licensing environment, codified in 10 Code of Federal Regulations (CFR) Part 52, will reduce substantially the risks associated with the first generation of nuclear plants described above. Rather than the NRC’s previous two-stage
licensing process that placed significant regulatory and financial risks on the applicant
and the vendor, the NRC now has in place a process that melds three component parts
into one.23
e concept behind Part 52 provides for preapproved standard reactor designs
and sites. An applicant can take an approved reactor design, add to it a pre-approved
site and then file an application for a Combined License, or COL. e applicant may
include design certification and/or site review in the COL process, at its option. e
COL is a one-step process with a hearing that should move more smoothly than in the
past, while protecting the rights of those involved.24 is new process also allows for
more meaningful and effective public participation. Whereas under the previous process the public participated at the construction permit stage, before much of the information associated with reactor design and operation was available, and/or at the operating license stage, aer the plant had been designed and built, and the substantial financial investment made, the new system allows for public participation in all three
stages at the outset of each, when that participation can be most effective. e application for an Early Site Permit carries with it a mandatory public hearing. e proceeding
for Design Certification is a public process, and public participation is permitted
through a “notice and comment” hearing process. Finally, the COL itself requires a
mandatory public hearing and the public may participate there as well. e sum of the
NRC’s revised licensing process is, of course, that an applicant is not required to make
substantial financial or organizational commitments until (i) it knows it has an alreadyapproved reactor design as a part of its application; (ii) a site that also is approved; and
(iii) a license in hand before it has to make the financial commitment necessary to construct and operate the plant.25
e new licensing process has a number of advantages over the old. Rather
than changing regulatory standards and requirements, there is a much more stable
23 10 CFR Part 52, Licenses, Certifications, and Approvals for Nuclear Power Plants.
24 See 10 CFR Part 2, Appendix L
25 Current projections are that for the new generation of nuclear plants, a two-unit station will cost in the
neighborhood of $14 billion to $18 billion.
55 Federation of American Scientists
Licensing and Regulation of U.S Nuclear Power Plants
regulatory regime including an already-approved reactor design. Moreover, the industry learned much about construction of operating plants, not only from those domestic
plants built in the 1970s and 1980s, but those built overseas in the last 20 years. ose
lessons will be put to good use in current domestic construction. Finally, the present
generation of plants is a much more mature technology than was the case in the earlier
Congress, in the Energy Policy Act of 2005 (EPACT05) included provisions
also designed to encourage investment in nuclear power. First, EPACT05 includes loan
guarantees for investments in low-carbon emitting technologies, and those guarantees,
which can cover up to 80 percent of project cost for technologies, including nuclear, are
a major reason that those moving forward with proposed plants are willing to do so.
EPACT05 also includes provisions for production tax credits for new nuclear
capacity added through 2021. Subject to a limit of $125 million per 1000 MW(e) per
year, and limited to the first 6000 MW(e) added per year. And finally, EPACT05 includes “Standby Support” provisions that are intended to protect license applicants
from delays in the licensing process, such as NRC or litigation delays beyond an applicant’s control, during plant construction up until commercial operation.
e question is whether there is at present a nuclear revival in the United
States. e answer is a cautious “yes.” At present, six applications for Early Site permits
have been filed. Four have been approved and five are still under review. Nine applications for reactor design certification have been filed. Four of these have been approved
and two are still under review. Most tellingly, 20 different applicants have filed with the
NRC 18 COL applications to construct and operate 28 reactors at 18 sites around the
United States. Plant Vogtle, owned by Southern Company and Oglethorpe Electric
Cooperative, is furthest along in the COL process and actually has begun limited construction at the site. All signs appear encouraging at this point.
Federation of American Scientists 56
The Future of Nuclear Power in the United States
February 2012
Chapter 4
Safety of Nuclear Power
by John F. Ahearne
Are nuclear reactors safe? Are they safe enough? What is meant by “safe,” and
how safe is safe enough?
Soon aer the 1942 nuclear experiments at Chicago’s Stagg Field, Enrico
Fermi expressed a concern that remains whenever reactor safety is discussed: “e public may not accept an energy source that is encumbered by vast amounts of radioactivity, and that produces a nuclear explosive, which might fall into hostile hands.”26
e public’s safety concerns were exacerbated by the 1979 TMI and 1986
Chernobyl accidents. e TMI accident destroyed the reactor and did not cause any
physical harm from radioactive releases (which were small). But it traumatized many
thousands. e Chernobyl accident destroyed the reactor and led to a few dozen nearterm deaths of the first responders. It released a large amount of radioactive material,
and that release may cause thousands of long-term deaths.27 e 2011 Fukukshima
Daiichi accident has caused public concern, but the long-term implications are far from
certain as of this writing.
26 Quoted in letter from A. M. Weinberg to John Gibbons, Assistant to the President for Science and
Technology, August 19, 1997.
27 Assessment of the delayed (latent) fatalities associated with the exposure of radioactive material released
by the Chernobyl accident indicates numbers up to 33,000 over the next 70 years assuming a linear nonthreshold effect of radiation….on this basis, natural background radiation would result in 1,500 times as
many deaths…over the same timescale so these additional fatalities, if the occur, would be very difficult to
observe. Comparing Nuclear Accident Risks with ose om Other Energy Sources, Nuclear Energy Agency
No. 6861, OECD 2010.
57 Federation of American Scientists
Safety of Nuclear Power
According to Remy Carle, a senior executive with Electricité de France:
Nuclear safety encompasses all the technical and organizational
measures to be taken to ensure that operation of a nuclear installation
has no harmful consequences for public health and the environment.
Nuclear safety is based on an approach known as ‘defense in depth’
which involves:
accident prevention, from the initial design stage, through
careful sizing of all installations, the taking into account of
possible equipment failures and human error, the taking into
account of external hazards, the implementation of safety
systems, and the quality control of the design and execution
of equipment and work;
continuous monitoring during operation, according to
procedures monitored by national authorities;
implementation of safety systems to maintain the cooling of
nuclear fuel and prevent the release of radioactive products
in the event of abnormal operation; lastly
definition of emergency planning and procedures to deal
with the highly improbably event of a serious accident. 28
As will be described later, new, enhanced safety features are incorporated in
current reactor designs and will continue to be important components of future reactors. e industry has been concerned about what to call these new plants. Should they
be “inherently safe,” “passively safe,” “transparently safe,” or some other term. “Naturally
safe” received the votes of 44 percent. One term did receive more: 49 percent thought
favorably of the term “safer.”29 I doubt this will be the term used by industry, for a reason
described later in this article. According to Alvin Weinberg, a “founding father” of U.S.
nuclear energy programs, ‘inherently safe’ reactors – depends “not on the intervention
28 Remy Carl, Nuclear Power (Presses Universitaires de France, 1994).
29 Post TMI: What Have We Learned, J.F.Ahearne in Nuclear Safety 1989 Conference Proceedings.
Federation of American Scientists
The Future of Nuclear Power in the United States
February 2012
of humans or of electromechanical devises but instead depend on immutable and wellunderstood laws of physics and chemistry.”30
A Fundamental Safety Principle
e International Atomic Energy Agency (IAEA) has produced many documents
addressing reactor operations. A key point from a 1988 IAEA guide is:
Fundamental Principle 3.1.2: e ultimate responsibility for the
safety of a nuclear power plant rests with the operating organization.
is is in no way diluted by the separate activities and responsibilities
of designers, suppliers, constructors and regulators. 31
is became an issue in the aermath of the TMI accident. A federal court
agreed with the NRC that the plant operator had the fundamental responsibility for
the safe operation of the plant.
Methodologies to analyze safety32
A. Overview of the general approach used to achieve high levels of safety in reactor
design and operation.
ere is a broad international consensus within the reactor-safety community
concerning the key elements that are necessary in the design and operation of a nuclear
power reactor to achieve a very high level of safety. While the presence of these key
elements generally should provide a high level of safety, the absence of one or more of
them is always a cause for concern. Here the phrase “a high level of safety” means a very
low probability of an accident that might cause death or injury to offsite populations
30 Inherently Safe Reactors and Second Nuclear Era, A. M. Weinberg and I. Spiewak, Science, 6/20/84
31 Safety Series No. 75 – INSAG-3, IAEA 1988
32 Extracted from Nuclear Energy: Present Technology, Safety, and Future Research Directions: A Status
Report, J. Ahearne, et al. (the primary author for the safety section was R. Budnitz), American Physical
Society Panel on Public Affairs, 2 November 2001.
59 Federation of American Scientists
Safety of Nuclear Power
due to radioactivity, or might cause important contamination of offsite land and property. It also implies that the risk to onsite workers and the risk of damage to the facility
itself are of acceptably low probability because the elements needed to achieve these are
very much congruent with the elements needed to protect offsite populations and
[I]t is important to describe in broad terms the safety-engineering challenge….
[F]or a reactor to be acceptably safe it is necessary to assure under all potential upset
conditions (a) that the nuclear chain reaction can be shut down and maintained in a
shutdown condition [reactivity control]…and (b) that the thermal energy (heat) in the
reactor, both heat present at the onset of the upset and heat generated by the continuing radioactive decay processes in the core, is removed to a safe ultimate heat sink…. If
both of these can be accomplished in an accident, the radioactivity within the reactor
can be contained; if they cannot, it will not be. While other crucial functions, such as
the containment function and the emergency-protective-action function, need to be
accomplished as back-ups in case these vital functions fail, the most important aspects
of preventing harm from the radioactivity are the functions of reactivity control and
heat removal.
Different reactor designs accomplish these vital safety functions in different
ways…. [A] design is preferable – that is, it is generally “safer” or, at least, more “demonstrably safe” – to the extent that each of these functions is accomplished by relying
more on physical principles and passive features and less on active equipment and human intervention. is does not mean that a reactor design relying mainly on active
equipment and human intervention cannot be made acceptably safe, but it does mean
that there is a broadly accepted hierarchy in which designs incorporating physical principles and passive features to accomplish the vital safety features generally are preferred.
A number of quite new reactor designs are now under active development.
Many of them claim as explicit advantages that they rely more on physical principles
and passive features to accomplish the key safety functions. e key elements that are
necessary in the design and operation of a nuclear power reactor to achieve a very high
level of safety are the following:
a) A strong base of both scientific and engineering knowledge to support each
aspect of the reactor-safety program.
Federation of American Scientists
The Future of Nuclear Power in the United States
February 2012
b) A reactor design that accounts for all important potential accident scenarios
by employing systems and operational features that reduce the probability of
each such scenario, or reduce its potential consequences (or both) to acceptable levels; and the ability to analyze that design well enough to provide high
assurance that the above is achieved.
c) A reactor design that uses established codes and standards and incorporates
adequate margins to assure acceptable performance in light of the uncertainties in knowledge.
d) A reactor design that incorporates a defense-in-depth safety philosophy to
maintain multiple barriers, including both physical and procedural barriers as
e) A reactor design that uses a philosophy of redundant and diverse safety systems to assure highly reliable performance during all potential accident scenarios.
f) A reactor design that incorporates technical specifications that conservatively
define, control, and circumscribe a safe operating envelope.
g) An adequate basis, in experiment, theory, and testing, to support the design
specifications and the safety analyses used for safety assurance.
h) e use of quality materials, quality manufacturing of equipment, and quality
construction and maintenance practices.
i) An operating philosophy that embodies a profound respect for the possible
dangers inherent in reactor operations.
j) A staff of qualified operating and maintenance personnel, supported by a management committed to a strong organizational safety culture, and also supported by a strong engineering capability.
61 Federation of American Scientists
Safety of Nuclear Power
k) An ability to analyze the safety achieved by the operation, in terms of both
realistic probabilistic analyses and conservative engineering analyses of the asbuilt-as-operated facility; and an ability to use the information from such
analyses to maintain and enhance safety.
l) Emergency plans that adequately protect offsite populations.
m) An operational safety culture that is both comprehensive and managed properly, and that incorporates an effective self-assessment and corrective-action
n) A system that derives safety insights from operating experience and from
analyses performed both within the reactor organization itself and elsewhere
around the world, and that applies these insights effectively.
o) A strong management organization with both the resources and the motivation to maintain all of the above.
p) An arrangement that has access to a continuing program of nuclear safety research, and that uses the insights derived from that research for safety improvement.
q) An independent regulatory authority that is responsible to the government
and the public for overseeing safety, and for taking corrective or enforcement
actions as necessary.
Over the past thirty years, international operating experience has demonstrated the importance of high-quality engineering of the facility and high-quality human performance…[O]perator qualifications and training must be supplemented by
operating procedures for both normal and abnormal/emergency situations, and by procedures for accident mitigation. All of the above must be embedded in a strong safety
culture, to ensure that each element of the entire safety envelope is maintained. Said
another way, the basic safety values and attitudes of the operating entity, from top to
bottom, can be as important as the basic design — inadequacies in either can lead to a
Federation of American Scientists
The Future of Nuclear Power in the United States
February 2012
degradation of safety.
Many countries deploy nuclear power reactors that achieve very high safety
levels because all of the elements above are present and are maintained. It is in this sense
that the nuclear-engineering community believes that adequate safety levels have been
achieved in these countries. e IAEA has assisted in upgrading the conduct of operations in many of its member countries.
B. Overview of methodologies used to assess the probability of accidents (severe and
In the early days of reactor operation, the methods used to assess how well safety was
achieved were qualitative rather than quantitative because no analytical methodology
existed that could provide quantitative estimates of the risks. is also is true for many
other complex technological endeavors (manned space travel is another example)
where severe accidents are concerns but occur too rarely to provide an evidentiary basis
for estimating the risks.
In the nuclear-power arena, a very capable methodology has evolved that now
provides a strong technical basis for such safety assessments. is is the “probabilistic
risk assessment” (PRA) methodology. e PRA methodology essentially involves writing down in the form of “event trees” each of the important accident sequences (from
initiating event to core damage to radioactive release) that might result in a major accident, and then analyzing the likelihood of each sequence using logic, equipment reliability data, human reliability data, an understanding of the correlations among failures, an understanding of the physical phenomena in each scenario, and a wealth of
design and operational information. Originally, the PRA methods were developed to
deal with accidents initiated by internal equipment faults and human errors. Today, the
methods can also deal well with potential accidents initiated by internal fires, external
phenomena such as earthquakes and tornadoes, and upset conditions occurring during
shutdown conditions as well as at full power.
For analyzing the probabilities and consequences of power-reactor accidents,
the methodology has reached a state of maturity in which it is now routinely used by
both the operating entities and the safety regulators worldwide as a continuing check
on how well each of the major elements of reactor safety is achieved.
63 Federation of American Scientists
Safety of Nuclear Power
However, the PRA methodology cannot provide highly accurate estimates of
the probabilities and consequences of the major accidents of concern: some of the underlying data and models are not known well enough to support such a very accurate
estimate. Hence the uncertainties in the “bottom-line” numbers for the annual coredamage frequency, or the likelihood of a specified large radioactive release of a certain
size and character, are oen as large as plus-or-minus an order of magnitude or more.
erefore, the major use of the PRA methodology is not to produce such
“bottom-line” assessments, important as they are in providing an overall understanding
of the safety levels achieved. Rather, the principal applications of PRA are to enable the
analyst and the safety decision-maker to understand which elements of the overall system contribute how much to safety, and why; and to study the effect on overall safety
of changes in the system (be they undesired changes due to equipment failures or human errors, or planned changes such as scheduled maintenance that may temporarily
compromise part of a safety function).
It is important to emphasize that the methodology of PRA, which was originally developed to assess the overall probabilities and consequences of major undesired
power-reactor accidents, does indeed provide such assessments and that these assessments are of broad use to policy-makers, despite the large numerical uncertainties in the
bottom-line risk numbers. In most countries around the world these bottom-line risk
numbers are judged acceptable by regulatory authorities, providing the context for the
rest of the work that reactor-safety professionals do in maintaining and improving
reactor safety.
Another major use of PRA methods is to assess the effectiveness of the overall
design and operation, by highlighting where additional equipment or modified procedures can enhance safety. Also, PRA can identify where it would be feasible to relax
strict engineering/maintenance standards for equipment that was originally thought to
be “required for safety” using traditional engineering principles, but that in fact contributes little to safety; such relaxations can simplify operations or save on human or
capital resources. PRA helps to establish optimum preventative maintenance programs
by focusing on the risks associated with equipment/system failure. e application of
PRA in maintenance, called ”reliability-centered maintenance,” identified cases where
increased preventative maintenance was needed, as well as cases where relaxed preventative maintenance was appropriate. Another major use of PRA is to allow the regulatory authority to concentrate its own resources on those design or operational aspects
Federation of American Scientists
The Future of Nuclear Power in the United States
February 2012
that contribute most to the safety of a given reactor facility, for example by guiding
regulatory inspectors about “where to look." PRA also can highlight areas where not as
much is known as we would like – thus motivating development of new knowledge,
either knowledge from operating experience or knowledge through advanced research.
It also is useful to recognize that, although PRA methods are mature enough
to be used routinely, there are important areas where additional PRA-methodology
research could be of benefit. ese include our limited understanding of how to analyze and affect the role of safety culture and management as it influences reactor safety;
our incomplete understanding of human performance under stress, including errors of
commission and errors of cognition; our limited understanding of how certain correlations among failures may affect safety; and our need for more realistic models of the
detailed behavior of radioactive materials inside the facility in some accident conditions. Further, the understanding of the effects on human health arising from potential
reactor accidents is severely limited by the incomplete understanding of the doseresponse relationship for radiation doses well below those that produce short-term
clinical effects. Because of knowledge limitations, the reactor-safety-analysis community has always applied dose-response models more suited to radiation protection than
to realistic assessment. All of these areas could benefit from the development of new
knowledge or new analysis tools, or both, which is the purpose of reactor safety
According to NRC historians, overall, the nuclear industry had an excellent reactor
safety record, due undoubtedly to the design and procedural requirements placed on it
as well as the research on safety questions. is track record notwithstanding, with the
increased public scrutiny of both the industry and the AEC in the 1970s, the probability of a major nuclear reactor accident loomed as a distinct threat to many critical
minds. e AEC had Brookhaven National Laboratory conduct a study to provide an
estimate of the upper limit of the consequences that might be involved in a reactor
accident. Report WASH-740 estimated that personal damage might range from no
injuries or fatalities to approximately 3,400 killed and 43,000 injured. e Rasmussen
Report’s or WASH-1400’s objective was to make a realistic estimate of public risk
33 End of APS material
65 Federation of American Scientists
Safety of Nuclear Power
involved in potential “class 9” accidents (accidents that could lead to radiological
consequences outside plant boundaries that would exceed AEC regulations). e basic
conclusion of this complex study suggested that risks to the public from potential nuclear accidents were small compared to other forms of risk in a complex technological
society. e conclusion of the report states, “It will take consideration by a broader
segment of society than that involved in this study to determine what level of nuclear
power plant risks should be acceptable.”34
Presently, according to the Organization for Economic Cooperation and Development (OECD), the theoretically calculated frequency of a severe nuclear power
plant accident followed by a large radioactivity release has [been] reduced by a factor of
1,600 between the original designs of early Generation I reactors and the Generation
III/III+ plants being built today. It is important to note that the “as originally designed” performance of the earlier plants has also been improved by upgrades over subsequent years.35
e Nuclear Energy Institute reported that the nation’s nuclear power plants
in 2009 had one of the safest industrial working environments and fell just shy of setting a
record for reactor efficiency. WANO [the World Association of Nuclear Operators]
found that the nation’s 104 operating nuclear power reactors reached a median unit capability factor – a measure of a plant’s on-line production time – of 91.3 percent.
ere were so few unplanned automatic reactor shutdowns per 7,000 hours of
reactor operation in 2009 that the median value approached zero. Statistics from other
industries through 2008, as compiled by the Bureau of Labor Statistics, show that it is
safer to work at a nuclear power plant than in the manufacturing sector and even the
real estate and financial sectors. 36
Other countries with current programs
U.S. NRC regulations have been used worldwide as a starting point for reactor regulation. For example, in Germany, safety philosophy is based on that of the United States
An Outline History of Nuclear Regulation and Licensing, 1946-1979, G. T. Mazuzan and R.R. Trask,
Historical Office of the Secretary, Nuclear Regulatory Commission, April 1979.
35 Comparing Nuclear Accident Risks, op cit.
36 Nuclear Energy Overview, NEI 6/11-17/2010
Federation of American Scientists
The Future of Nuclear Power in the United States
February 2012
with greater reliance on accident prevention or mitigation concepts.37 In France, for
more than 20 years, there has been an extensive and continuous exchange of information between nearly all safety organizations around the world, either by means of bilateral agreements or under the sponsorship of such international bodies as the International Atomic Energy Agency in Vienna and the Nuclear Energy Agency in Paris.
Furthermore, numerous efforts have been made to promote the harmonization of safety
criteria in use in various countries, and countless working groups have been set up for
that purpose. Consequently international consensus has been achieved on many important safety issues, and gradually the safety philosophies have grown closer.38
New Entries
Many countries without nuclear plants have expressed interest in such plants. ere are
several potential safety problems as these plans progress. An IAEA nuclear safety
review this year gave advice on addressing the issues associated with new entrants:
One striking change from the past is now emerging – namely, the
interest in the pursuit of nuclear power by a large number of countries
that have no previous experience with power reactors. Safety, on the
other hand, is a challenge that every new entrant will necessarily
confront. It is an inevitable and on-going challenge at every nuclear
site. e creation of a culture that enables the achievement of safety
takes persistence, commitment and very hard work and needs to start
at the moment that a decision is made to embark on a nuclear power
program and endure throughout a power plant’s life. It is expensive.
And it involves an attention to detail and a willingness to accept and
learn from intrusive peer review by others. At the opening stage, the
new entrant must establish the legal framework for nuclear activities
and create an independent, competent and well-funded regulator
with an overriding commitment to nuclear safety. e IAEA safety
standards provide crucial general guidance. Practical experience has
37 “e Safety Concept of Nuclear Power Plants in the Federal Republic of Germany,” H.L.Schnurer and
H.G.Seipel, Nuclear Safety, 24, no.6, Nov. – Dec.1983, pp.743-782.
38 e French Approach to Nuclear Power Safety, P. Tanguy, Nuclear Safety, Sept-Oct 1983
67 Federation of American Scientists
Safety of Nuclear Power
shown that the availability of high-quality safety standards is not
enough to guarantee their effective implementation. Vendors must
seek to ensure that a new entrant understands and has the capability
to meet its safety commitments.39
e need for trained personnel was identified in another review: “For nuclear
power to expand in emerging countries will require a massive education and training
program to support nuclear power programs.”40
New Reactors 41
Reactors with the term “advanced” are currently operating and more are planned. All
these are classed as GEN III+. ey include the ABWR (advanced boiling water reactor) built in Japan and the Korean APR-1400, modeled aer the CE System 80+. e
latest is the French EPR being built in Finland and EPR is under construction in Brittany. Japan also is building two large (1540 MWe) advanced pressurized water reactors
(APWR) of Japanese design. e Korean reactor exemplifies a clear trend in Asia,
where countries are designing and building their own reactors. is is seen in Japan,
South Korea, China, and India. For example, in June of 2005, the 540 MWe PHWR
Tarapur-4, a reactor designed and built by the Nuclear Power Corporation of India,
was connected to the grid.
Using a term of the 1990s, these are evolutionary reactors, improved (oen
significantly) modifications of existing reactors. e gas-cooled pebble bed modular
reactor (PBMR) also is related to previous reactors but with operational experience
limited to one German research reactor and a small pebble bed reactor in operation in
China, the 10 MWe HTR-10. China has plans to build a 160 Mwe commercial demonstration pebble bed reactor. e PBMR offers small size, gas cycle efficiency, and
accident-resistant fuel. e early proponent of commercialization has been Eskom, the
39 Letter to IAEA Director General Amano from INSAG Chair Meserve, August 28, 2010.
40 “Report from Como: Expanding Nuclear Power to New States – Defining Needs and Exploring Paths to
Success,” 10-14 June 2008, Como, Italy.
41 Extracted from “Advanced Nuclear Reactors – eir Use in Future Energy Supply,” J.Ahearne, Forum on
Physics and Society, April 2006, vol. 35, No. 2.
Federation of American Scientists
The Future of Nuclear Power in the United States
February 2012
large South African utility.42
Each of the new reactors was designed to be simpler, safer, and have lower cost
than currently operating reactors. e passive safety feature reactors rely on gravity,
natural circulation, and compressed air to provide cooling of both the core and the containment in the case of a severe accident. is permits a reduction in systems that were
designed to force coolant into the system. For example, compared with a typical similar
size reactor, passive safety systems in the AP1000 led to 50 percent fewer valves, 35
percent fewer pumps, 80 percent less pipe, 48 percent less seismic building volume, and
70 percent less cable.43
e IAEA recently summarized some of the new reactor developments:
China has already developed and operated its own domestic mediumsize PWR designs [and] the China National Nuclear Corporation …
has developed the evolutionary China Nuclear Plant (CNP-1000)…
e European pressurized water reactor’s (EPR’s) power level of
1600+ MW(e) has been selected to capture economies of scale.
Electricite de France (EdF) is planning to start construction of an
EPR at Penly beginning in 2012. Two EPR units are also under
construction in China….
[I]n Japan, MHI [Mitsubishi Heavy Industries] has developed the
advanced pressurized water reactor (APWR+)…
In the Republic of Korea, the benefits of standardization and series
construction are being realized with the 1000 MW(e) Korean
Standard Nuclear Plants (KSNPs). Ten KSNPs are in commercial
In the Russian Federation, evolutionary WWER plants have been
designed….WWER-1000 units are currently under construction at
the Kalinin and Volgodonsk sites and WWER-1200 units at the
Novovoronezh-2 and Leningrad-2 sites.
42 In September 2010 the government announced the project had been cancelled.
43 End of advanced reactors article.
69 Federation of American Scientists
Safety of Nuclear Power
Advanced HWR [heavy water reactor] designs are also being
developed in a number of countries. 44
Will these new plants be “safer”? is was addressed in a recent article:
It is worth noting that the NRC does not require the new plants to be
any safer than existing ones. Rather, it only requires the plants to
“provide the same degree of protection” as the current generation of
Although the new U.S. reactors will have some “design enhancements” – digital controls versus analog dials, for example – “at bottom
they are based on familiar and proven technology” says [Nuclear Energy Institute official] Russ Bell.
Will this new generation of reactors be safer than the current nuclear
plants? Ask that of the industry’s Bell and he chose his words carefully, because to imply that the new reactors are “safer” than the old
ones [would] infer that the existing plants are less safe.45
Public Perception and Acceptance
Fermi wondered if the public would accept nuclear plants. e safety of these plants
has been a constant issue for the nuclear industry’s relationship with the public. Polls
show the same trend but are not consistent in the overall message.
Gallup: A majority of Americans have been supportive of the use of nuclear energy in
the United States in recent years, but [the 2009] Gallup Environment Poll found new
highs of support, with 59 percent favoring its use, including 27 percent who strongly
44 “International Status and Prospects of Nuclear Power, Report by the Director General,” GOV/INF/
2010/12-GC (54)/INF/5, September 2, 2010.
45 “e Nuclear Power Resurgence: How Safe Are the New Reactors?” Susan Q. Stranahan, Yale Environ-
ment 360, June 17, 2010.
Federation of American Scientists
The Future of Nuclear Power in the United States
February 2012
favored it. Americans may have not fully embraced the use of nuclear energy because of
concerns about potential health risks from a nuclear meltdown or the nuclear waste
that power plants produce. e poll found that a majority of Americans, 56 percent,
believed that nuclear power plants are safe, but a substantial minority of 42 percent
Bisconti: In March 2010, three out of four Americans said they favor nuclear energy
(74 percent). Also, 70 percent believed that “we should definitely build more nuclear
power plants in the future.”47
MIT: In the five years since the last [MIT] survey in 2002, public preferences have
remained fairly stable, but the percentage of people who want to increase nuclear power
has grown from 28 percent to 35 percent. 48
Post Fukushima: Unsurprisingly, public support went below a majority favoring nuclear
Following the two major reactor accidents of TMI and Chernobyl, designs have been
scrutinized and improved, operating practices improved, and personnel training
stressed. Also, during the ongoing Fukushima accident, the NRC has been conducting
a review of U.S. nuclear plants. is review may spur further design changes and safety
retrofits. Safety remains a mixture of design, construction, maintenance, and what has
been called a “safety culture”: the need for all involved personnel to stress safety in all
their practices. e US NRC has wrestled with “how safe is safe enough” and published
safety goals for operating plants.49 ese were designed to explain safety to the public
but were not made requirements and have been succeeded by a design requirement for
annual core-damage frequency of 10-4, clear to technologists but not transparent to the
46 Jeffrey M. Jones, “Support for Nuclear Energy Inches up to New High,” Gallup poll, March 20, 2009.
47 “Public Support for Nuclear Energy at Record High,” Bisconti Research, March 2010.
48 “Americans warming to nuclear energy – MIT survey,” MIT News
Office, July 23, 2007.
49 “Safety Goals for the Operations of Nuclear Power Plants: Policy Statement,” 51 FR 30028, 8/21/86.
See also “How safe is safe enough?” J.F.Ahearne, Reliability Engineering and System Safety 62, 1998.
71 Federation of American Scientists
Safety of Nuclear Power
public. New designs rely on passive rather than active systems that have not yet been
challenged by an accident. Asia is the location for the large growth in numbers of reactors in the next decade, placing challenges on regulators, education systems, and operator training. What happens in Asia or other parts of the world will likely affect the U.S.
nuclear power fleet.
Federation of American Scientists
The Future of Nuclear Power in the United States
February 2012
Chapter 5
Security of U.S. Nuclear Power Plants
by Harold A. Feiveson
Commercial nuclear power plants in the United States could be targets for
terrorists attempting to release radioactive materials to the environment. According to
the 9/11 Commission, nuclear power plants were among the targets considered in the
original plan for the September 11, 2001 attacks.50 In addition to nuclear power plants,
other nuclear facilities include civilian research reactors, certain naval fuel facilities,
uranium enrichment plants, and fuel fabrication plants. Security at these other nuclear
facilities must also be addressed, but this chapter focuses only on security at the commercial nuclear power plants.51
e reats to Nuclear Power Plants
If terrorist groups could sufficiently damage safety systems to cause a core meltdown,
such an attack could lead to a massive radioactive release to the environment. An
attack on a reactor’s spent fuel pool could also be severe, and it is noteworthy that the
pools are less well protected than the reactor core. Depending on location and the
amount of radioactive material, the release of radioactivity could lead to thousands of
near-term fatalities and still greater numbers of long-term deaths. In the year 2000,
about one-fih of the nuclear sites had more than 100,000 people living in the 10-mile
emergency zone around the sites.52
50 e National Commission on Terrorist Attacks Upon the United States, e 9/11 Commission Report,
July 22, 2004.
51 Security at research reactors were the subject of a recent study by the U.S. Government Accountability
Office, Nuclear Security Action May Be Needed to Reassess the Security of NRC-Licensed Research Reactors,
January 2008.
52 Edwin S. Lyman and David Lochbaum, “Protecting Vital Targets: Nuclear Power Plants,” in James J. F.
Forest, ed., Homeland Security: Protecting America’s Targets, Volume III (Critical Inastructure), 2006, pp.
Federation of American Scientists
Security of U.S. Nuclear Power Plants
If nuclear power is to grow substantially, nuclear facilities – especially the numerous reactors in diverse locations — will have to be made extremely safe from incidents that could release massive quantities of radioactivity to the public. New reactor
designs have improved safety, partly by incorporating features of passive safety, such as
the flooding of the reactor core without active intervention by reactor operators. ese
safety measures have generally been developed and studied with respect to accidents –
not to the deliberate attack on a reactor by a terrorist group, which could in principle
disable or destroy the various multiple protections against accidents.53 However, as explained further below, the Nuclear Regulatory Commission does also require new reactor license applicants to consider security during the design stage.
e terrorist threat is of two general types: commando-like ground-based attacks (including different attack modes), possibly abetted by an insider, on some designated target sets, notably equipment which if disabled could lead to a core meltdown or
dispersal of radioactivity from the spent fuel pool; and external attacks such as an aircra crash into the reactor complex or cyber attacks (which may also be initiated via
Government Oversight of Security at Nuclear Power Plants 54
Overview. e Nuclear Regulatory Commission (NRC) has the primary responsibility
for ensuring that power reactor licensees operate in a secure manner. e NRC is an
independent federal agency responsible for licensing civilian nuclear plants and regulating and overseeing their safe operating and security.55 In discharging this responsibility,
the NRC coordinates its security activities with other federal, state, and local law
enforcement agencies, including the Department of Homeland Security (DHS), the
Federal Bureau of Investigation (FBI), the Federal Aviation Administration (FAA), and
53 Edwin S. Lyman, “Chernobyl on the Hudson? e Health and Economic Impacts of a Terrorist Attack
at the Indian Point Nuclear Power Plant,” Union of Concerned Scientists, September 1994, p. 21.
54 ree excellent recent overviews are Mark Holt and Anthony Andrews, Nuclear Power Plant Security
and Vulnerabilities, CRS Report to Congress, Congressional Research Service, March 18, 2009, and updated, August 23, 2010; and Edwin Lyman, “Security since September 11,” Nuclear Engineering International, March 2010.
55 U.S. Government Accountability Office, “Nuclear Power Plants: Efforts Made to Upgrade Security, but
the Nuclear Regulatory Commission’s Design Basis reat Process Should Be Improved,” GAO-06-388,
March 2006, p. 12.
Federation of American Scientists 74
The Future of Nuclear Power in the United States
February 2012
others. e DHS in particular supports efforts to enhance security outside of the plants
e NRC formulates a design basis threat (DBT) that nuclear plant operators
have to defend against, and issues and enforces regulations to ensure that the nuclear
plants implement appropriate security measures to meet the DBT. e DBT does not
represent the maximum size and capability of a terrorist attack that is possible, but
rather some plausible threat that the plant operators have to consider in devising physical security arrangements. In particular, the DBT and NRC regulations do not require
nuclear power plants to protect against attacks by an “enemy of the United States,”
whether a foreign government or other person. e NRC evidently decides the dividing line between the design-basis threat and the beyond-design-basis threat.56 e
details of the DBT, including the adversary characteristics, are specified in non-public
regulatory guides, so that language in the NRC rule could be compatible with different
numbers of attackers and weapons, and these numbers could be increased without a
change in the overall rule.
e NRC oversight seeks to ensure adequate power plant performance in five key
aspects: access authorization to critical areas of the plant; access control; physical protection
systems, such as fences, cameras, and the like; material control and accounting; and response
to contingency events.57 The last aspect refers to commando-like attacks, and is tested by
the NRC through so-called Force on Force (FOF) inspections, discussed further below.
Design Basis reat. It is widely believed that before the September 11, 2001 attacks,
the DBT consisted of one team of three individuals assisted by a single passive insider
who could provide plant-specific information but not participate in the attack.58 Aer
9/11, the NRC and Congress took actions to strengthen the DBT. In 2002 and 2003,
56 Edwin Lyman, “Security since September 11,” Nuclear Engineering International, March 2010, pp. 16-
57 U.S. Nuclear Regulatory Commission, Report to Congress on the Security Inspection Program for Commer-
cial Power Reactor and Category I Fuel Cycle Facilities: Results and Status Update, Annual Report for Calendar Year 2009, p. 3.
58 Edwin Lyman, “Security since September 11,” Nuclear Engineering International, March 2010, pp. 15-
Federation of American Scientists
Security of U.S. Nuclear Power Plants
the NRC worked to revise the DBT, and it approved a revised DBT in April 2003 “to
represent the largest reasonable threat against which a regulated private guard force
should be expected to defend under existing law.”59 In the Energy Policy Act of 2005,
Congress imposed a statuary requirement on the NRC to revise the DBT based on
assessments of various terrorist threats worldwide, plausible explosive devices, and the
possible use of modern weapons, such as precision-guided munitions. In January 2007,
the NRC approved its final rule amending the DBT effective April 18, 2007. It appears
that the 2007 DBT is similar to that of 2003, though with some changes that have not
been made public. In the 2007 rule, the NRC noted:60
e Nuclear Regulatory Commission (NRC) is amending its regulations that
govern the requirements pertaining to the design basis threats (DBTs). is
final rule makes generically applicable security requirements similar to those
previously imposed by the Commission’s April 29, 2003 DBT Orders, based
upon experience and insights gained by the Commission during implementation, and redefines the level of security requirements necessary to ensure that
the public health and safety and common defense and security are adequately
protected. Pursuant to Section 170E of the Atomic Energy Act (AEA), the
final rule revises the DBT requirements for radiological sabotage, generally
applicable to power reactors and Category I fuel cycle facilities, and for the
or diversion of NRC-licensed Strategic Special Nuclear Material (SSNM),
applicable to Category I fuel cycle facilities.61
Although the specific details have not been made public, the NRC has clarified that the 2003 DBT “expands the assumed capabilities of adversaries to operate as
one or more teams and attack from multiple entry points; assumes that adversaries are
willing to kill or be killed and are knowledgeable about specific target selection; expands the scope of vehicles that licensees must defend against to include water vehicle
and land vehicles beyond four-wheel drive type; revises the threat posed by an insider
59 Federal Register, May 7, 2003.
60 Federal Register, March 19, 2007.
61 Category 1 fuel cycle facilities are those that use or possess formula quantities of strategic special nuclear
material (defined in Title 10 of the Code of Federal Regulations [10 CFR, Section 70.4] as uranium-235,
contained in uranium enriched to 20 percent or more in the U-235 isotope, uranium-233, or plutonium.)
Federation of American Scientists 76
The Future of Nuclear Power in the United States
February 2012
to be more flexible in scope; and adds a new mode of attack from adversaries coordinating a vehicle bomb assault with another external assault.”62
e exact number of attackers in the new DBT is not public. But Time Magazine reported in 2005 that the new number incorporated in 2003 was “less than double
the old figure and a fraction of the size of the [September 11] group.”63 If that report is
accurate, and the final 2007 DBT is close to that of 2003, it would appear that the new
2007 DBT envisions no more than 5 attackers. e new DBT includes a wider range of
weapons available to the attackers than previously considered, but apparently with
some plausible weapons still excluded. Specific adversary attributes are discussed in a
non-public document.
e final DBT rule excluded aircra attacks, an issue which is discussed further below. Cyber security actions were required by the NRC aer 9/11 and subsequently codified through issuance of 10 CFR 73.54 in March 2009. e new regulation
“requires licensees to submit a new cyber security plan and an implementation timeline
for NRC approval. e plan must show how the facility identified (or would identify)
critical digital assets and describe its protective strategy, among other requirements.”64
In January 2010, the NRC published a regulatory guide that provides guidance to licensees on ways to meet the requirements of the regulation. e possibility of cyber
attacks on reactors has been made vivid by speculation that the so-called Stuxnet worm,
which has recently received attention generated by cyber security experts in Germany,
was directed at Iran’s Bushehr reactor, as well as the Natanz centrifuge plant.65
62 Mark Holt and Anthony Andrews, Nuclear Power Plant Security and Vulnerabilities, 2010, p. 3; see 10
CFR 73.1.
63 Mark ompson, “Are ese Towers Safe?” Time, June 20, 2005, pp. 34-48; Lyman, Security since Sep-
tember 11, p. 16.
64 Nuclear Regulatory Commission, “Backgrounder on Cyber Security,” April 2010; Office of the Inspec-
tor General, Nuclear Regulatory Commission, Audit of NRC’s Force-on-Force Inspection Program, July
30, 2009, pp. 1-2; Nuclear Regulatory Commission, Report to Congress on the Security Inspection Program
for Commercial Power Reactor and Category I Fuel Cycle Facilities: Results and Status Update, Annual Report for Calendar Year 2009, p. 4.
65 Mark Clayton, “Stuxnet worm mystery: What’s the cyber weapon aer?,” Christian Science Monitor,
September 24, 2010; Nicolas Falliere, Liam Murchu, and Eric Chien, “W32.Stuxnet Dossier,” Symantec,
November 2010; see more recently, William Broad, John Markoff, and David Sanger, “Israeli Test on
Worm Crucial in Iran Nuclear Delay,” New York Times, January 15, 2011.
Federation of American Scientists
Security of U.S. Nuclear Power Plants
Force on Force (FOF) Inspections and Exercises. e NRC carries out FOF inspections at
all Nuclear Power Plant (NPP) sites at least once every three years. e NPP is given
notice of an upcoming inspection about two to three months in advance. e FOF
inspection, which is typically conducted over the course of 3 weeks, “includes both
tabletop drills and exercises that simulate combat between a mock adversary force and
the licensee’s security force. At an NPP, the adversary force attempts to reach and
simulate damage to key safety systems and components, defined as “target sets” that
protect the reactor’s core or the spent fuel pool, which could potentially cause a radioactive release to the environment. e licensee’s security force, in turn, interposes itself
to prevent the adversaries from reaching target sets and thus causing such a release.”66
An FOF inspection typically includes three FOF exercises over three nights. “Plant
defenders know that a mock attack will take place sometime during a specific period of
several hours, but they do not know what the attack scenario will be.”67 Participants
carry weapons modified to shoot laser bursts, and wear laser sensors to indicate hits.
Other weapons and explosives are also simulated.
Before 9-11, the NRC conducted FOF exercises about once every eight years
at each NPP. Starting in 2004, the NRC has undertaken to conduct FOF inspections at
each plant site once every three years, with tactical security drills, conducted by the licensees, in the intervening years. This implies about 22 FOF inspections each year. In preparation for the FOF exercises, information from the tabletop drills are factored into the
adversary attack strategies. When a complete target set is simulated as destroyed, and the
NRC determines that the licensee’s protective strategy does not assure protection
against the DBT, the NPP has to put in place compensatory measures immediately.
Since October 2004, Wackenhut, the same company that provides security
forces to several nuclear plants, has managed the mock adversary force. e NRC acknowledges that this at a minimum creates a perception of a conflict of interest. However, to guard against such a conflict of interest, the NRC requires that no member of
the mock adversary group may participate in an exercise at his or her home site; and the
NRC emphasizes that it, not the mock adversary group, designs, runs, and evaluates the
66 Nuclear Regulatory Commission, Force-on-Force Security Exercises, Fact Sheet, May 2007; Nuclear Regu-
latory Commission, Report to Congress on the Security Inspection Program for Commercial Power Reactor
and Category I Fuel Cycle Facilities: Results and Status Update, Annual Report for Calendar Year 2009, pp.
67 Mark Holt and Anthony Andrews, Nuclear Power Plant Security and Vulnerabilities, 2010, p. 8.
Federation of American Scientists 78
The Future of Nuclear Power in the United States
February 2012
results of the FOF exercises and that the adversary’s performance is subject to continual
observation by the NRC.68 Department of Defense contractors provide support to the
mock adversary group in tactics planning.
In 2009, the NRC conducted FOF inspections at 22 commercial nuclear
power plants; and by the end of the year, it had completed the second year of the second 3-year cycle of FOF inspections. Cumulatively, from November 2004 through December 2009, the NRC had conducted 112 inspections, 8 of which times a complete
target set was damaged or destroyed.69
Aircra attacks. As noted, the 2007 DBT did not include aircra attacks, and existing
nuclear power plants were not required to undertake active protective measures against
airborne threats. Nevertheless, in a final rule issued on June 12, 2009, the NRC
required applicants for new nuclear plant designs to “perform an assessment of the
effects of the impact of a large, commercial aircra,” and “identify and incorporate design features and functional capabilities to show, with reduced use of operator actions,”
that address such impact. e rule still classifies an aircra attack as “beyond design
basis threat.” Also, in addressing the threat of an aircra attack, the NRC works with
other agencies, including the FAA and the North American Aerospace Defense
Command (NORAD). In the view of critics, all this does not go far enough. For one
reason, new plants will comply with the rule if analysis shows that “either the reactor
core remains cooled or the containment remains intact, and either spent fuel cooling or
spent fuel pool integrity is maintained.” us, it might be that containment is not
breached initially but the core damaged. Also, it is not clear how spent fuel pool cooling can be maintained if spent fuel pool integrity is violated.70
e air crash risk to a power plant is unclear. e Union of Concerned Scientists and other interest groups argue that an air attack could penetrate the containment
68 Nuclear Regulatory Commission, Force-on-Force Security Exercises, Fact Sheet, May 2007, pp. 3-4.
69 Nuclear Regulatory Commission, Report to Congress on the Security Inspection Program for Commercial
Power Reactor and Category I Fuel Cycle Facilities: Results and Status Update, Annual Report for Calendar
Year 2008, and 2009. NRC-2008 reported that for the period November 2004 to December 2008, 4 inspection had resulted in compete target set damage or destruction. NRC-2009 reported 3 such results in
2009, and a cumulative total of 8 from November 2004 to December 2009. It is not clear why the cumulative total was not 4 + 3 = 7.
70 Edwin Lyman, “Promoting Mediocrity: NRC’s Policy for New Facility Security Design,” presented at
the Institute of Nuclear Materials Management 50th Annual Meeting, Tucson, AZ, July 12-16, 2009.
Federation of American Scientists
Security of U.S. Nuclear Power Plants
structure of a nuclear power plant or spent fuel storage facility, causing a core meltdown or spent fuel fire. Nuclear industry people counter that nuclear plants are difficult targets for attack, that penetration of the containment is unlikely even if hit, and
that even if penetration occurred, it would not reach the reactor vessel.
Security Assessment
Without access to classified information, it is not possible to convincingly assess the
security of the reactor sites. Even with such information, for example on the exact character of the DBT, nothing definitive can be concluded. For one reason, no one can
really know how likely is a “beyond a DBT threat.” Nevertheless, a few comments can
be made.
In 2006, the GAO found that “the process NRC used to revise the DBT for
nuclear power plants in April 2003 was generally logical and well defined. … [T]he
NRC threat assessment staff developed and used a comprehensive screening tool to
analyze intelligence information and evaluate particular terrorist capabilities, or ‘adversary characteristics,’ for inclusion in the DBT. … [e NRC Commission] produced a
revised DBT that generally but not always corresponded to the original recommendations of the threat assessment staff.”71
ere were troubling aspects, however, to the GAO findings. With respect to
the size of a car or truck bomb that could be used in the DBT, the Nuclear Energy Institute (NEI), representing the nuclear industry, argued for a less destructive bomb than
initially assumed by the NRC staff on three grounds: “the low probability of a bomb
the size proposed by the NRC; the likelihood that federal authorities or local law enforcement would detect a large vehicle bomb; and the inability of some sites to protect
against the size of the vehicle bomb proposed by the NRC because of insufficient land
for installation of vehicle barrier systems at a necessary distance.”72 is last reason especially appears the most jarring. In the event, the NRC staff did reduce the size of
vehicle bomb assumed.
With respect to weapons, the Nuclear Energy Institute (NEI) argued against
the inclusion of several weapons suggested by the NRC staff, and again it is the reasons
71 U.S. Government Accountability Office, “Nuclear Power Plants: Efforts Made to Upgrade Security, but
the Nuclear Regulatory Commission’s Design Basis reat Process Should Be Improved, GAO-06-388,
March 2006, pp. 5-6.
72 Ibid, p. 20.
Federation of American Scientists 80
The Future of Nuclear Power in the United States
February 2012
forwarded by NEI that appear the most troubling. For example, the NEI wrote that
one particular weapon recommended by the NRC staff “would render the ballistic
shielding used at nuclear power plants obsolete, and that another proposed weapon
would initially cost $1 million to $7 million per site to defend against, with annual
recurring costs of up to $2 million.” Partly in response to the NEI, the NRC staff did
remove some weapons from the DBT, though not one at least which the NEI had objected to – and the NRC Commissioners later voted to remove that particular
weapon.73 e Program on Government Oversight (POGO) has argued that this
weapon likely was the rocket-propelled grenade.74 How much the NRC staff and
Commissioners were swayed by the industry objections is not known. But what is most
puzzling is that the NEI objected to the dra DBT partly on apparently relatively trivial economic grounds. Given that a typical nuclear reactor in the U.S. spends about
$200 million per year in operating costs, the recurring $2 million per year cited by the
NEI looks negligible.
Critics have pointed out that the threat used by the Department of Energy for
its nuclear sites apparently includes three times the number of attackers in the NRC
DBT.75 is may be explained by the more urgent need to protect weapon-usable material at the sites from the or actual use of an improvised nuclear explosive at the sites
by terrorists. Also the of nuclear material would require terrorists to both enter and
leave the protected sites. It would appear that the effective design basis threats used by
the NRC for facilities under its responsibility, which are using so-called Category 1
material (notably plutonium and highly enriched uranium), are not that different from
the threats assumed by the Department of Energy.
e NRC staff told GAO that the NRC did not make changes to the dra
DBT based solely on industry views. But, as noted by the GAO, “the process used to
obtain feedback from the nuclear industry created the opportunity for, and appearance
73 Ibid, pp. 20-21.
74 Danielle Brian, Testimony before the House Subcommittee on National Security, Emerging reats and
International Relations, “Nuclear Security: Has the NRC Strengthened Facility Standards Since 9/11,”
April 4, 2006.
75 Ibid.
e DOE threat assessment is now under some flux. It may be that the 2005 DBT of the Department of Energy, which Danielle Brian of POGO referred to on the number of attackers, was never fully
implemented. e DOE has now replaced its DBT with a “graded security protection” plan: “DOE
Adopts New 'Graded' Terrorist Protection Policy,” George Lobsenz, Energy Daily, August 26, 2008.
Federation of American Scientists
Security of U.S. Nuclear Power Plants
of, industry influence on the threat assessment regarding the characteristics of an attack.“ NRC officials assured the GAO that in considering future changes to the DBT,
“NRC plans to ensure the initial separation of intelligence analysis from interaction
with stakeholders.”76
With respect to the FOF inspections and power-plant security arrangements,
the GAO in its 2006 assessment, “saw a clear connection between the changes in the
DBT and the plants’ recent security enhancements. e plants’ response to the revised
DBT and other NRC orders following the September 11 terrorist attacks has been substantial … Nevertheless, because the plants essentially designed their security to defend
against the DBT outlined by NRC, their capability to defend against an attack is essentially limited to how similar such an attack would be to the DBT.”77 It is troubling that
even with respect to the DBT, there were evidently three serious failures during the
2009 FOF exercises, which led to complete target set damage or destruction.78
Questions on the quality of the security personnel at the nuclear power plants have also
been raised. In 2002, the Project on Government Oversight (POGO), in interviews
with nuclear power plant guards, noted several disturbing concerns raised by the guards
interviewed – that they were under trained, under equipped, and under paid.79 Aer
review, the NRC took various actions to improve security guard performance, including restricting security officer work hours and establishing new security force training
and qualification requirements.80 More recently, further concerns about guard fatigue
were raised. In September 2007, a TV reporter presented the NRC with video
evidence that showed a number of security officers at the Exelon’s Peach Bottom station
in Pennsylvania asleep in the site “ready room” (where security officers not on active
patrol or observation post are stationed, ready to respond if called upon). is led to
76 U.S. Government Accountability Office, “Nuclear Power Plants: Efforts Made to Upgrade Security, but
the Nuclear Regulatory Commission’s Design Basis reat Process Should Be Improved,” GAO-06-388,
March 2006, pp. 21-22.
77 Ibid, p. 42.
78 Nuclear Regulatory Commission, Report to Congress on the Security Inspection Program for Commercial
Power Reactor and Category I Fuel Cycle Facilities: Results and Status Update, Annual Report for Calendar
Year 2009, p. 9.
79 Project on Government Oversight (POGO), Nuclear Power Plant Security: Voices om Inside the Fences,
Revised, October 2, 2002.
80 Mark Holt and Anthony Andrews, Nuclear Power Plant Security and Vulnerabilities, 2010, p. 10.
Federation of American Scientists 82
The Future of Nuclear Power in the United States
February 2012
Exelon severing relations with its security contractor, Wackenhut, and to the NRC
issuing a security bulletin in December 2007 requiring licensees “to gather information
on administrative and management controls and any other actions taken to address
Future Nuclear Power Plants
It may be that a massive release of radioactivity to the environment from a nuclear
power plant, either through a failure of safety measures or through terrorist action, is an
improbable event. But the consequences of such a release would be so devastating that
the question should be asked whether future nuclear designs could render a massive
release of radioactivity still more improbable – even, for example, if a terrorist group
was able to take charge of a reactor site or to contrive to crash an airplane into the
power plant or to attack it with precision-guided weapons.
e question is important because increasing the security at power plant sites,
say by adding more security guards with more weapons, carries its own risks. Scott Sagan has pointed this out in a 2005 article. By illustration, Sagan notes that expanding
the guard force could also increase the insider threat, could lead to social shirking of
responsibilities by the guards, and could give a false sense of assurance, leading to more
reckless behavior.82
e strongest defense to a catastrophic terrorist attack is to make the nuclear
reactor sites as inherently safe as possible. While complete “inherent” safety against all
possible pathways that could release radioactivity to the environment might not be
practically achievable, it is conceivable such a goal could be approached. e goal has
been stated thus by two nuclear experts, then at Oak Ridge National Laboratory:83
81 Nuclear Regulatory Commission, Report to Congress, 2008, p. 6. Steven Mufson, “Video of Sleeping
Guards Shakes Nuclear Industry,” Washington Post, January 4, 2008.
82 Scott Sagan, “e Problem of Redundancy: Why More Nuclear Security Forces May Produce Less Nu-
clear Security,” Risk Analysis, Vol. 24, No. 4, 2004.
83 Charles Forsberg and omas Kress, “Underground Reactor Containments: An Option for the Future?”
2nd International Topical Meeting on Advanced Reactor Safety, paper No. 159, American Nuclear Society,
June 1-4, 1997.
Federation of American Scientists
Security of U.S. Nuclear Power Plants
Nuclear power plants should be designed to ensure that the consequences of
the worst-case accidents [or deliberate attacks] will be of only limited local
concern. e prevention of large-scale release of radionuclides shall not
depend upon plant operation and maintenance practice nor on the relative
reliability of active safety systems and components.
For example, though spent fuel pools are generally not protected by a containment dome and in this sense more vulnerable than the reactor to attacks from the
ground or air, the way in which the pools are managed could greatly affect the risks of
large releases of radioactivity in the event of loss of cooling. us, a study by independent scientists in 2003 showed that with dense packing of the spent fuel, a loss of water
coolant could potentially lead to a propagating zirconium fire (zirconium being the
material commonly used in cladding the fuel rods) and a large radioactive release to the
environment.84 A National Academy of Sciences study, released in 2005, supported this
analysis. Both analyses, however, suggested ways to manage the spent fuel pools (for
example, by more rapid removal of spent fuel to dry-cask storage, or, in the NAS study,
by careful interspersing of hotter and cooler spent fuel), which would make a zirconium fire much less likely.85
With respect to reactor technologies, there may also be ways to achieve a formidable degree of inherent safety. For example, the modular pebble bed reactor in
which the pebbles are not expected to melt even if the coolant is completely and permanently lost, could look attractive on such safety and security grounds, though the
possibility and consequence of graphite fires would have to be further studied. Also
many of the array of small reactors now being designed appear to have features of intrinsic safety.
While it is beyond the scope of this chapter to discuss the feasibility and economics of achieving the goal of inherent safety, in general, it is worth noting one possibility that Forsberg, Kress, and others have highlighted – underground siting of nuclear
reactors. e argument is that underground containments can provide very strong resis-
84 Robert Alvarez et al. “Reducing the Hazards from Stored Spent Power-Reactor Fuel in the United
States,” Science & Global Security, Volume 11, 2003, pp. 1-51, and “Response by the Authors to the NRC
Review of ‘Reducing the Hazards …’,” Science & Global Security, Volume 11, 2003, pp. 213-223.
85National Research Council, Board on Radioactive Waste Management, Safety and Security of Commercial
Spent Fuel Storage, Public Report, National Academies Press, Washington, D.C., 2005.
Federation of American Scientists 84
The Future of Nuclear Power in the United States
February 2012
tance to assaults and accidents. As noted by Forsberg and Kress, “high technology
weapons or some internal accidents can cause existing … containments to fail, but only
very high energy releases can move large inertial masses associated with underground
containments.” It certainly would appear to be the case that underground containment
would protect power plants from aircra crashes and attacks off site by rocket propelled grenades and the like. Whether the underground siting could largely contain
releases from a meltdown of a reactor seems less clear, but that too may be possible.
And naturally, the vulnerability of underground-sited reactors to extreme seismic
events would have to be carefully assessed.
Security at nuclear power plants appears to have improved since 9/11. e Design
Basis reat has been increased some, and the force on force exercises by the NRC,
done once every eight years before 9/11, are now being done every three years.
However, questions remain whether the DBT is yet realistic enough to capture
plausible threats by terrorist groups, and whether the DBT and associated reactor
security operations have been adjusted to accommodate industry concerns with cost.
Whatever the DBT, there will always be the possibility of a beyond-DBT attack on a reactor. is suggests the value of the nuclear industry seeking reactor designs
and operational procedures that are more inherently safe than the current systems.
Federation of American Scientists
Future Uranium Supplies for U.S. Nuclear Reactors
Chapter 6
Future Uranium Supplies for U.S. Nuclear Reactors
by Ivan Oelrich
A possible constraint on future nuclear-electricity production could be limits
in the supply of uranium fuel. Complex calculations of future uranium demand and
supply really boil down to a single central question: Given that nuclear reactors take a
decade to build and can operate for 60 years, should a decision today about building a
nuclear reactor take into account the possibility that its productive lifetime will be cut
short by lack of fuel? e answer today is “no.” Well-characterized reserves of uranium
are large enough to power all existing and currently planned reactors to the ends of
their lives. e history of uranium discovery strongly suggests that the answer will not
change for at least a few decades into the future.
Uranium was discovered in 1789, and for the next one and a half centuries had
only minor practical applications, such as a pottery glaze and a glass coloring agent.86
Consequently, little effort was put into prospecting for ore deposits. Unsurprisingly,
with no one looking, little of the element was found. Uranium deposits were assumed
to be geological oddities, occurring in useful concentrations at very few sites around the
globe. (At the end of World War II, the United States thought it might be able to maintain a nuclear weapons monopoly by cornering the world’s uranium market.) It has
turned out, however, that uranium is fairly common and widespread; even with robust
growth in nuclear power, the United States and the world have many decades’ supply of
uranium available.
e importance of uranium changed dramatically and irrevocably with the
discovery of nuclear fission. In the winter of 1938-9, German scientists found that the
newly discovered neutron could split a uranium atom with an accompanying release of
fantastic amounts of energy. When scientists also found that, when the uranium atom
broke apart, it emitted additional neutrons that could themselves split even more atoms,
86 Tom Zoelllner, Uranium:
War, energy, and the rock that shaped the world, Viking, New York, 2009, p.
Federation of American Scientists 86
The Future of Nuclear Power in the United States
February 2012
the possibility of a chain reaction was obvious. A runaway chain reaction would result
in an explosion while a controlled chain reaction might provide a constant source of
power that could be harnessed. With World War II looming, fission’s application to a
new super-weapon could not be ignored and President Roosevelt established the Manhattan Project to develop nuclear weapons. e first nuclear weapon used in war, exploded over Hiroshima, Japan, was powered by uranium.
e wartime Manhattan Project also included development of the first nuclear
reactors. ey were used to produce plutonium for bombs but their peaceful applications were clear from the beginning. While the United States built the first reactor to
produce electricity, the Experimental Breeder Reactor in Idaho, it was little more than a
stunt using a research reactor to power four large light bulbs (the reactor is now a National Historic Landmark).87 In 1954, the Soviet Union built the first reactor that fed
electricity into a power grid, providing about 5 megawatts (MW) to the city of
Obninsk.88 e British built the first of what would today be recognized as a commercial nuclear reactor, the 50 MW Calder Hall 1, which began operation in 1957.89 e
United States soon followed in December 1957 with the Shippingport reactor. Nuclear
power grew steadily from that point until the Chernobyl accident when new reactor
construction almost stopped for two decades except in some Asian countries. Today,
the world’s approximately 440 commercial nuclear reactors have a electrical generating
capacity of about 373,000 megawatts and produce about 2.6 trillion kilowatt-hours of
energy per year, requiring 68,000 tons of uranium per year.90
Predicting future U.S. uranium availability requires a comparison of demand
and supply. Uranium demand is discussed first. Demand must address both the demand
for nuclear-generated electricity and the efficiency with which uranium is used to produce that electricity. Supply calculation must consider established and estimated reserves of uranium ores and how estimates of resources can expand as greater effort, and
87 “EBR-1 Factsheet,” Idaho National Laboratory,
88 “Nuclear Power in Russia,” World Nuclear Association, updated March 2011,
89 Nuclear Energy Agency,
Forty Years of Uranium Resources, Production and Demand in Perspective,
OECD, Paris, 2006, p. 2.
90 Nuclear Energy Agency/International Atomic Energy Agency, Uranium 2009:
Resources, Production
and Demand, Paris, 2010, p. 59.
87 Federation of American Scientists
Future Uranium Supplies for U.S. Nuclear Reactors
of course money, is devoted to uranium extraction. e chapter ends with a discussion
of uranium enrichment demand and capacity in the coming decades.
Early in the nuclear era, uranium demand was dominated by the need for highly enriched uranium (HEU) for nuclear weapons production but, with the end of the Cold
War, the United States stopped HEU production for nuclear weapons in 1992. us,
today, uranium requirements are essentially determined by commercial nuclear power
plants’ needs and those depend, in turn, on the amount of energy produced by nuclear
reactors and the efficiency with which the reactors use their uranium fuel. (Research
reactors and naval propulsion reactors, while significant in number and sometimes presenting real proliferation risks, are small and have little effect on global uranium consumption.)
Predictions of uranium consumption begin with predictions of reactor operation and production. Nuclear reactors take years to plan and build, so predictions can
be reliably made about installed capacity of nuclear power plants a decade into the future. Reactors, once built, operate for 60 years and perhaps longer, so current reactor
capacity provides a reliable prediction of baseline or minimum installed capacity for a
few decades into the future. In the case of the United States, there are few new reactor
construction projects underway (although many applications for new reactors are under review by the Nuclear Regulatory Commission). us, existing reactors will make
up the U.S. inventory for at least the next 20 years. Uranium use depends, not on installed reactor capacity but on actual electricity production, but reactors are
expensive so utilities try to operate them as close to maximum capacity as possible.
With greater experience, reliability has improved to the point that U.S. reactors now
typically operate at close to 90 percent or greater capacity, thus, maximum capacity is a
good predictor of actual use.
e greatest uncertainty in future reactor capacity lies in the number of reactors that will come online a decade or more into the future. Future construction plans
depend on overall demand for electricity, concerns about atmospheric carbon dioxide,
levels of government support, the cost of capital, and the availability of inexpensive
natural gas; none of these can be reliably predicted. (Moreover, predicting future U.S.
capacity is not enough; uranium is a globally traded commodity, so U.S. operators must
compete for uranium supplies in a global market, which means that global nuclear reac-
Federation of American Scientists 88
The Future of Nuclear Power in the United States
February 2012
tor capacity and consumption affects uranium availability in the United States.)
Aside from one small dip in 1998, the total capacity of U.S. nuclear reactors
has gone up every year since the beginning of the nuclear age. Typically, predictions of
future nuclear power capacity are based on an assumption of some constant growth in
electricity consumption and generation, usually 1.5 to 2.0 percent a year, and further
based on the assumption that nuclear power will continue to provide at least as large a
fraction of overall capacity as today but with the fraction most likely increasing. However, from the beginning of the nuclear age, almost all predictions of future generating
capacity have been too high; thus, current predictions should be viewed with some
caution.91 In the past, overestimates were due to overly optimistic estimates of social
and political acceptance and of the cost of producing nuclear electricity. For predictions made today, the dominating assumption is extrapolation of electricity demand.
e oen cited 2003 MIT study, e Future of Nuclear Power assumed, for example,
that U.S. electricity consumption would almost double over the next 40 years but also
showed that other countries with the equivalent quality of life, as measured by the UN
Human Development Index, use far less electricity.92 Australia uses two thirds as much
electricity per person as Americans while rich European countries, such as Germany,
Britain, and the Netherlands use one half as much electricity per person.93 Simple extrapolation of power demand ignores ever-increasing conservation pressures and potential in the developed world and in the United States, in particular.
Some advocates of nuclear power today talk about a “nuclear renaissance,” resulting from a recovery aer the long post-Chernobyl hiatus in reactor construction
and from increasing concerns about global warming and desires for low-carbon energy
but even this enthusiasm is dampening. For example, the 2003 MIT study predicted,
as a lower bound, that by 2050 the United States would have 386 gigawatts (or GW, a
gigawatt is a billion watts) of installed nuclear capacity, but the MIT study group’s
2009 update noted that no new reactor construction had begun in the United States.
us, the authors saw that even that lower bound projection was overly optimistic.
(Since then, some new construction has begun at the Vogtle site in Georgia.)
91 Forty Years, op. cit., p 27.
92 John Deutch and Ernest Moniz, co-chairs, e Future of Nuclear Power, an interdisciplinary study, Mas-
sachusetts Institute of Technology, Cambridge, Massachusetts, 2003.
93 Ibid., p. 109.
89 Federation of American Scientists
Future Uranium Supplies for U.S. Nuclear Reactors
e global picture is much different from that in the United States alone. At
the end of 2010, there were 441 civilian power reactors operating across the world with
58 more under construction.94 With reactor lifetimes now being extended to 50-60
years, new construction will outpace retirements for at least several years and the number of reactors will certainly increase. e same study cited above that suggests the potential for energy conservation in the United States also shows that there seems to be an
increase in quality of life with electricity use above about two to three thousand kilowatt hours of electricity use per person per year. e majority of the world’s population,
including China and India, is below this level and will want to reach it over the next
half century. erefore, regardless of conservation in the rich countries, global electricity use will continue to increase and the fraction of that produced by nuclear in developing countries can only grow because it is such a small fraction today. is means that
the United States, independent of its own needs, will be competing in a global market
of ever increasing demand.
Today, about 20 percent of total U.S. electric generating capacity comes from
nuclear plants with 101 GW of installed nuclear generating capacity,95 which produced
806 terawatt-hours of electric energy (a terawatt is a trillion watts),96 implying an industry average production of 91 percent of capacity. is energy production in the
United States consumed the equivalent of 16,424 tons of uranium metal (in 2008). e
Department of Energy (DOE) reports that its “high” consumption estimate in 2030 is
about 23,000 tons. Extrapolating to mid-century produces a total requirement between
now and then of 920,000 tons of uranium metal. e “low” estimate of the MIT study
is about twice this rate of growth, which, added to the existing baseline, results in total
requirements by mid-century of 1200 thousand tons of uranium metal.
Global uranium reserves are reported every second year by the International Atomic
Energy Agency (IAEA) in a report called Uranium [the year of publication]: Resources,
Production, and Demand and, because of its cover, referred to informally as the Red
94 See World Nuclear Association,, accessed November
95 Uranium 2009, op. cit., p. 60.
96 Ibid, p. 62.
Federation of American Scientists 90
The Future of Nuclear Power in the United States
February 2012
Book. While the Red Book is considered authoritative, it depends on voluntary reporting by participating nations. Resources are described by level of confidence in the quantity of the deposit, the cost of extraction, and by region.
Until the ore is actually taken out of the ground and processed, there is some
uncertainty about its uranium content. e Red Book distinguishes among various levels of confidence in reserves. e highest confidence deposits are Known Conventional
Resources (KCR), usually comprising deposits in existing or planned mines. Next is Reasonably Assured Resources (RAR), which are deposits that are not fully characterized but
in which there is high confidence. Most reports of uranium “reserves” lump KCR and
RAR together into Identified Resources. Beyond this are more speculative estimates,
down to estimates based, not on actual prospecting data, but on wide-scale geological
surveys that identify geologic formations of a type that have elsewhere contained uranium deposits and are thus assumed to be likely candidates for future resources.
While the United States consumes 17 percent of the world’s uranium production, it has only seven percent of the world’s reserves and cannot, under free market
conditions, be self-sufficient in uranium but this should not be a security concern for
several reasons. A significant fraction of the world’s petroleum is concentrated in the
politically uncertain Middle East; in comparison, uranium is widely distributed around
the globe. Some of the largest reserves are held by America’s closest allies, Canada and
Australia. Other large reserves are widely dispersed, such as in southern Africa and
central Asia, so are unlikely to be disrupted simultaneously. Finally, the cost of uranium
as fuel is, unlike coal or natural gas, a small slice of the total cost of producing electricity; in addition, the energy density of uranium fuel is thousands of times higher than
coal, making the volume of the fuel low. ese two factors together mean that stockpiling even years of supply is economically and strategically feasible.
e world’s largest uranium reserves and the largest producers are shown in
Table 1. Note that the countries with the largest reserves are not always the largest producers. Australia, for example, has the largest proven reserves but did not start to rise as
a big producer of uranium until 1980. (Australia does not have commercial nuclear
plants.) Counting from the beginning of the nuclear age, the United States, Canada,
and the Soviet Union were the largest cumulative producers of uranium. (is was, at
least in part, a reflection of the Cold War and its demand for a huge arsenal of nuclear
weapons.) Since the mid-1970s, uranium became less of a strategic/military commodity and more of a normal commercial commodity and several other countries expanded
91 Federation of American Scientists
Future Uranium Supplies for U.S. Nuclear Reactors
uranium mining operations. While the United States still has large reserves, other
countries have developed ore deposits that are cheaper to extract so the relative current
economic attractiveness of U.S. extraction has declined. e United States’ production
peaked in 1980. Canada, while not having the largest reserves, has some of the richest
ore deposits in the world and the cheapest to extract so has remained the world’s largest
producer until 2008. While Russia still is a major producer, most Soviet production
and reserves were in Kazakhstan. In just the last few years, Kazakhstan has invested
heavily in mining with a resultant surge in production. In 2009, it surpassed Canada as
the world’s largest producer.97
e world’s proven reserves of uranium have been increasing since the 1980s.
Although nuclear reactors are consuming uranium constantly and over two million
tons of uranium have been extracted since 1945,98 reserves are increasing for two reasons: first, continuing prospecting has expanded the number of sites where uranium is
known to be; and, second, technical developments have allowed better extraction of
uranium, for example, by in situ solvent pumping methods and, in the future, bacterial
concentration, in addition to traditional mining operations.
For reserves, absolute numbers are less significant than years of supply. By this
measure, reserves have held roughly constant for the past quarter century with about 45
years of supply identified in proven and probable reserves. is number and its stability
are not a coincidence. Once forty or so years of reserves have been identified, there is
little economic incentive to invest money in finding new sources; consequently, prospecting slows, and the discovery of new deposits slows proportionately until reserves
fall below the forty year threshold. Similarly, as long as high-grade ores that are cheap
to extract are available, there is little incentive to invest in more advanced technology to
extract uranium from poorer quality ores.
e cost of extraction is integral to estimates of reserves. As the market is willing to pay more for extraction, the number of ore deposits that are exploitable increases.
Thus, any estimate of reserves cannot simply be a cited amount but an amount extractable
at a specified cost. (With prices high enough, uranium could theoretically be extracted
97 World Nuclear Association, “Uranium Mining,” updated April 2011,
98 Forty Years, p.13,
Federation of American Scientists 92
The Future of Nuclear Power in the United States
February 2012
from seawater, increasing reserves by factors of hundreds or thousands.)99 Reserves are
today typically reported for cases for which the uranium can be extracted for less than
$130/kg of uranium metal. (e standard 130 number comes about because the
United States once reported reserves in terms of dollars per pound of uranium oxide;
thus $130/ kg metal is equivalent to the $50/lb U3O8 breakpoint.)
Currently identified reserves extractable at $130/kg are 5.4 million tons of
uranium metal but, if ores available at <$260/kg are included, the amount more than
doubles to 11.7 million tons. More speculative resources total an additional ten million
tons.100 e Red Book suggests that data about ongoing mining operations may be considered proprietary so may be under-reported. Reserves of expensive uranium may be
underestimated relative to cheaper reserves because there is no incentive to prospect for
expensive deposits when more than adequate cheaper deposits are known. In the second half of the century, as the cheapest deposits are exhausted, more attention will be
given to what are today marginal ores and these reserves will increase. And some nations do not even report speculative resources; thus, those are almost certainly underestimated. Over the past quarter century, uranium prices have hovered around $50/kg
metal and over the last few years have approached $100/kg, so estimates of resources
available at $130/kg include uranium at prices higher than today’s. Higher uranium
costs are, however, less significant than for other types of electricity-generating fuel because the uranium itself constitutes only about 4 percent of the cost of producing nuclear electricity. (Nuclear costs are dominated by the capital costs of the power plant).
us, if prices actually did reach $130/kg, nuclear electricity costs would only go up by
a percent or two.
In 2010, total uranium production was 53,660tons of metal.101 us, at current extraction rates, existing identified < $130/kg reserves would last 112 years. Current production does not meet current use, however. e United States consumes
about 17 percent of the world’s uranium, but about half of that is provided by a
U.S.-Russian program called Megatons-to-Megawatts, in which 500 tons of highlyenriched uranium from old Soviet nuclear weapons are diluted and consumed in U.S.
99 H. Nobukawa, "Development of a Floating Type System for Uranium Extraction from Sea Water Using
Sea Current and Wave Power," in Proceedings of the 4th International Offshore and Polar Engineering Conference (Osaka, Japan: 10-15 April 1994), pp. 294-300.
100 Uranium 2009, p. 10.
101 “Uranium Mining,” op cit.
93 Federation of American Scientists
Future Uranium Supplies for U.S. Nuclear Reactors
reactors, providing the equivalent of 100,000 tons of natural uranium. is residual
inventory le over from the Cold War will be consumed soon aer the program ends in
2013. Adjusting for the end of this outside source, current known reserves will meet
current demand for only another 92 years. Of course, while future consumption is unknown, it will increase, so current identified reserves will be consumed sometime before
then. Using the Red Book “high” estimates for growth, currently identified reserves will
last 42 years—consistent with the long-term stability of around forty years of reserves—and using its “low” growth estimate they will last 57 years. All of these numbers should be roughly doubled if currently identified and well-characterized <$260/kg
reserves are included.
Most chemical elements have more than one isotope, that is, atoms that are chemically
virtually identical but have slightly different weights or mass. Most of the mass of the
atom is in the nucleus and differences in mass come about because the nuclei have different numbers of neutrons. While the number of neutrons has almost no effect on the
normal chemistry of an atom, it has a profound effect on the nuclear properties. Uranium has two important isotopes, U-235 and U-238, where the number indicates the
total number of protons and neutrons in the nucleus.
It is the lighter U-235 that can be easily split by neutrons and sustain a chain
reaction, powering both nuclear reactors and nuclear weapons. But natural uranium
contains only 0.72 percent U-235; the 99.28 percent remainder is U-238. (Trace
amounts of another isotope, U-234, occur in natural uranium but are not directly important to nuclear power production.) Some reactors, using special materials, such as
heavy water (that is water containing the stable heavy isotope of hydrogen called deuterium), can use natural uranium as fuel. One example is the very successful CANDU, or
CANadian Deuterium Uranium reactor. However, the great majority of the world’s
existing and planned nuclear reactors use and will use normal, or light, water and for
these reactors the fraction of U-235 must be increased above the natural concentrations. is process of increasing the concentration of U-235 is called enrichment. e
amount of enrichment capacity required is closely proportional to the energy produced
by light water reactors.
Because the chemical properties of different isotopes are virtually identical,
some physical process that exploits the small difference in the mass must be used to
Federation of American Scientists 94
The Future of Nuclear Power in the United States
February 2012
separate them. Although uranium is the heaviest natural element, it forms a compound
with fluorine, uranium hexafluoride or UF6, which is a gas at only slightly elevated
temperatures. Coincidentally, fluorine has only one stable isotope, F-19, so any difference in the mass of the UF6 molecule has to be due to differences in the mass of the
uranium. All commercial enrichment methods use uranium in this gaseous form of
UF6. e process of turning uranium metal or oxide into UF6 is called conversion.
e first enrichment process to be used on a large scale was gaseous diffusion,
developed during the Manhattan Project, which exploits the slightly faster diffusion
through a metal mesh of the lighter UF6 containing U-235. e first countries to develop nuclear weapons, the United States, Russia, Britain, France, and China all developed huge gaseous diffusion plants. Each stage achieves only very slight enrichment so
many stages are needed and the gas must be recompressed each time, consuming huge
quantities of electricity. (It is no coincidence that the first American gaseous diffusion
plants were built in the heart of the Tennessee Valley Authority’s hydroelectric production area.)
In theory, centrifuges are more efficient at separating uranium isotopes, requiring only a twentieth or so as much electricity. e idea of using centrifuges predates the
Manhattan Project but, to be effective, the centrifuges must spin at extremely high
speed and technical limitations of the needed high-strength materials and high-speed
bearings made early centrifuges impractical. ese problems were slowly overcome so
that, by the 1960s, centrifuges were far and away the preferred enrichment method.
While some old diffusion plants are still in operation, they are facing retirement and all
currently planned industrial-scale enrichment plants will use centrifuges.
e degree of enrichment of uranium is measured by Separative Work Units,
usually abbreviated SWUs. e separative capacity is the degree of enrichment of a
given amount of material, so total separative “work” is typically measured by a SWU of
enrichment on one kilogram of material, or a kg-SWU. (e “kg” is oen incorrectly
dropped in much writing and reported as simply SWU. Some older British and American texts use pound-SWUs. e output of entire enrichment plants is sometimes reported in ton-SWUs.)
A typical nuclear power plant will have an electrical output of a gigawatt.
With an efficiency of converting heat to electricity of one third, the plant will produce
3 GW of heat and, over the course of a year, consume about 25 tons of uranium fuel
enriched to about 4 percent U-235. Starting with natural uranium, this typical power
95 Federation of American Scientists
Future Uranium Supplies for U.S. Nuclear Reactors
plant requires about 100,000 to 120,000 kg-SWUs per year. (As one fraction of uranium becomes more concentrated in U-235, obviously the remainder is less concentrated; this is called “depleted uranium.” e depleted remainder from the enrichment
process, call the tails, typically contains 0.2-0.25 percent U-235. How much U-235 is
le in the tails depends on a cost calculation determined by a tradeoff between the cost
of separation and the cost of fresh uranium.)
e world’s enrichment capacity is concentrated in just a few countries. Russia
is the largest, with a 25 million kg-SWU/yr capacity. Recall that Russia will, for a few
more years, continue selling off fuel made by dilution of weapons-grade uranium,
which has the same effect as enrichment capacity. France’s Areva/Eurodif can produce
11 million kg-SWU and the Anglo-German-Dutch consortium Urenco can produce
another 11 million.
e enrichment capacity is changing rapidly in the United States. e United
States currently produces about 11 million kg-SWU/yr. Based on a need of 120 thousand
kg-SWU/yr/GW, the United States requires a little over 12 million kg-SWU/yr. The
country now depends on an outdated gaseous diffusion plant in Paducah, Kentucky, a
legacy of the Cold War enrichment program that originally produced highly-enriched
uranium (HEU) for weapons. The Paducah plant is facing retirement and the supply of
diluted Russian nuclear weapon uranium is about to end. To compensate, two modern
gas centrifuge plants are planned. One, to be built by the United States Enrichment Corporation (USEC), will be near Paducah and use advanced centrifuges developed at Oak
Ridge National Laboratory. It will have about four million kg-SWU/yr capacity. The
European consortium, URENCO, is building a centrifuge facility in New Mexico that
will add another six million kg-SWU/yr.102 A pilot plant facility that uses finely-tuned
lasers to separate the isotopes of uranium is under construction in Wilmington, North
Carolina, but the process is considered proprietary and few details are available.
In general, there should be little concern about limitations of enrichment capacity because, by all measures, whether capital requirements, construction time, or
technical, environmental, and political hurdles, the challenges facing nuclear power
plants are far more daunting that those facing enrichment plants. In short, the United
States and the world can apparently install enrichment capacity faster than it can install
reactor capacity; thus, enrichment should be able to keep up with demand.
102 World Nuclear Association, “Uranium Enrichment,” October 2010,
Federation of American Scientists 96
The Future of Nuclear Power in the United States
February 2012
Recall that natural uranium is only 0.7 percent U-235. Only about 70 percent of
that is extracted, the remainder is left in the “tails” because the cost of further extraction
exceeds the cost of starting with fresh natural uranium. If improvements in enrichment
technology, for example, laser enrichment, could substantially reduce the cost of enrichment, then more U-235 would be extracted from any given amount of natural uranium.
This could significantly, if not dramatically, increase the world’s effective supply of U-235.
Predictions of near term uranium demand are quite precise. Growth rates over the
coming decades are, in contrast, highly uncertain. Unknown uranium resources are, obviously, unknown so upper bounds on availability are completely uncertain. But wellcharacterized resources provide a floor on available nuclear fuel, and these resources are
large. Even allowing for robust growth in nuclear-electric power generation and using
fairly conservative assumptions about current and future reserves, decisions about building nuclear reactors should not today be constrained by concerns about fuel availability.
The long-term fuel situation will be constantly reevaluated but, for decades to come,
uranium availability will most likely not be the factor limiting nuclear growth.
Table 1. Major Producers and Known Uranium Reserves
2008 Resources and
Identified Resources at
< $130/kg (1000s tons metal)
(tons metal/yr)
South Africa
United States
97 Federation of American Scientists
Prospects for a Plutonium Economy in the United States
Chapter 7
Prospects for a Plutonium Economy in the United States
by Ivan Oelrich
Plutonium is an element not found in nature, but a typical nuclear power reactor produces hundreds of kilograms of the material each year. Plutonium has the potential to power nuclear reactors by itself, thereby increasing, in theory, the energy extractable from uranium by a factor of about a hundred, extending uranium reserves,
perhaps by a thousand years. Many attempts have been made to exploit this theoretical
potential, but practical limitations of engineering and economics have thwarted widespread application. As long as adequate supplies of cheap uranium are available, as discussed in the previous chapter, which will be at least many decades and probably until
the end of the century, plutonium will have to wait. Plutonium separation also presents
proliferation dangers because it can be used to power nuclear weapons as well as reactors.
Plutonium is one of the most dangerous long-term components of nuclear
waste. Schemes for separating plutonium from waste and treating it have been proposed, but have not proven economical compared to geological disposal or even
century-long storage in dry casks until permanent storage can be arranged.
Plutonium and Commercial Nuclear Power
rough a series of radioactive decays, the uranium-238 (U-238) in the fuel of a reactor
transforms within a few days into plutonium, or Pu-239, which is fissile, and because it
has a half-life of about 24,000 years, it does not appreciably decay further. Plutonium
and other artificial elements beyond the heaviest natural element, uranium, are called
transuranics. Because of its ability to transform into a fissile atom, U-238 is said to be,
not fissile, but fertile.
Federation of American Scientists 98
The Future of Nuclear Power in the United States
February 2012
e production of fissile material from fertile material is called breeding and
reactors specifically designed to produce plutonium or other artificial fissile materials
are breeders. rough this indirect mechanism, the latent energy in U-238 can be extracted. Because all commercial reactors powered by uranium-235 (U-235) have even
more U-238 in their fuel, the sea of neutrons in the reactor will convert some U-238 to
plutonium inside the reactor and some will be consumed right there in the reactor
where it is produced, a process sometimes called in situ or internal breeding. In this
way, a third of so of the energy in a typical reactor is derived from plutonium created
from the U-238, not from the U-235.
When commercial uranium fuel is exhausted, it is approximately one percent
plutonium, with the concentration varying by type of fuel and reactor. e United
States currently has about 62,000 tons of commercial reactor fuel waste and is producing about 2300 tons a year.103 At one percent, that translates into 620 tons of plutonium locked up in used fuel elements. Most of that is the useful isotope Pu-239, which
contains, kilogram for kilogram, about as much energy as U-235 which means that
used reactor fuel actually has a higher fissile energy potential than natural uranium.
Fully exploiting the potential of plutonium production requires, however, that the plutonium be separated from the irradiated nuclear fuel. While different isotopes of an
element are virtually identical chemically, making their separation difficult, uranium
and plutonium are different elements with different chemical properties so they can be
separated by relatively simple chemical means.
Separating Plutonium
e first techniques for producing and separating plutonium were developed in the
Manhattan Project to make plutonium for nuclear weapons. e Manhattan Project
technique, called PUREX, for Plutonium-URanium Extraction,104 still has variants in
use today. Chemically separating the components, including plutonium, from used nuclear fuel is called reprocessing. In the PUREX process, the used fuel is dissolved in nitric
acid and then an organic solvent is used to extract the uranium and plutonium and
103 Nuclear Energy Institute, Nuclear Waste: Amounts and On-Site Storage, undated,, accessed 10 December 2010.
104 M. S. Gerber, A Brief History of the PUREX and UO3 Facilities, U.S. Department of Energy Office of
Environmental Restoration and Waste Management, Novemeber 1993.
99 Federation of American Scientists
Prospects for a Plutonium Economy in the United States
many other heavy, radioactive nuclei. By adjusting the chemical conditions of the solvent, the various components can be isolated by precipitating them one at a time from
the solution.
Plutonium from a commercial nuclear reactor, while not ideal, can be used to
make a nuclear bomb so variations on the PUREX process have been developed that do
not isolate pure plutonium. Some of these schemes involve extracting the plutonium
and uranium together so the concentration of plutonium is not high enough to make a
bomb. e shortcoming of this approach is that it makes the plutonium, albeit impure,
much easier to steal. Spent reactor fuel contains fission products, making it so intensely
radioactive that the International Atomic Energy Agency describes it as “selfprotecting,” that is, potential unauthorized proliferators or terrorists would be killed by
ionizing radiation before they could get very far with their booty. e plutonium/
uranium mixture is far less radioactive and, once it is stolen, it can easily be run back
through the old PUREX process and the pure plutonium extracted. Other schemes
involve intentionally leaving in radioactive materials that are difficult to separate from
plutonium but these typically do not meet the criteria of being self-protecting.105
Most U.S. plutonium was produced as part of the nuclear weapons program at
reactors at Savannah River, South Carolina and Hanford, Washington, and then separated using the PUREX process.106 e only commercial effort to separate plutonium
from fuel took place at a facility at West Valley, New York. It operated from 1966
through 1972 and during that time processed 640 tons of fuel, on average about one
fih its design capacity. e facility caused significant environmental contamination
that is estimated to cost about $4.5 billion to clean up.107
e French still operate a large separation facility at La Hague that uses a
PUREX process.108 e British also have a large facility near Sellafield, called the
105 Jungmin Kang and Frank von Hippel, “Limited Proliferation-Resistance Benefits from Recycling Un-
separated Trasnuranics and Lanthanides from Light-Water Reactor Spent Fuel,” Science and Global Security, 13, pp-169-181, 2005.
106 Plutonium:
e First Fiy Years: United States plutonium production, acquisition, and utilization om
1944 through 1994, United States Department of Energy, February 1996.
U.S. General Accountability Office, GAO-01-314, Nuclear Waste: Agreement Among Agencies Responsible for the West Valley Site Is Critically Needed, May 11, 2001, p. 2.
108 Mycle Schneider and Yves Marignac, Research Report No. 4, International Panel on Fissile Materials,
Spent Nuclear Fuel Reprocessing in France, International Panel on Fissile Materials, April 2008.
Federation of American Scientists 100
The Future of Nuclear Power in the United States
February 2012
ermal Oxide Reprocessing Plant, or THORP, that uses the PUREX process,109 but
the facility was shut down aer a major leak of radioactive material occurred in 2005.
e plant will probably will not be reopened. Russia has a reprocessing facility at Mayak. Japan is struggling to open a large reprocessing facility at Rokkasho-mura. China
and India also want commercial reprocessing facilities. However, the vast majority of
countries that use commercial nuclear power do not reprocess spent fuel, and most of
these countries do not use recycled plutonium fuel.
In some new approaches, called pyroprocessing, plutonium and the rest of the
fuel are melted directly or dissolved in molten salt. Chemical or electrochemical processes then separate the plutonium. Some of these approaches are designed to make it
inherently difficult to separate pure plutonium. Pyroprocessing has, however, not yet
been demonstrated on commercial scales.
Plutonium as Reactor Fuel
e separated plutonium can be recycled through nuclear reactors, fissioned, and the
energy converted into electricity. In the most common commercial power reactors,
neutrons must be slowed down or moderated to increase their reactivity (and because
the neutrons are in thermal equilibrium with the reactor material, these neutrons are
called thermal neutrons). Most reactors use normal or “light” water as both a moderator and coolant (as opposed to “heavy” water made with a heavy isotope of hydrogen).
Light Water Reactors (LWRs) cannot operate with natural uranium because the concentration of U-235 is too low; the concentration must be increased through enrichment to 3-5% U-235 for typical reactors. Plutonium can replace some of the U-235 so
that plutonium-spiked uranium can be used as a fuel much like enriched uranium. e
fuel is used in its ceramic-like oxide form, so the uranium/plutonium fuel is called
MOX, or Mixed OXide, fuel.
Today, France, as mentioned above, reprocesses plutonium from LWRs and
recycles it back through its LWRs, but there are severe limitations to this approach.
e first is cost. When used fuel is removed from a reactor, it is intensely radioactive,
making handling and reprocessing difficult and expensive. A 2005 study estimated reprocessing costs of $2000/kg of spent fuel, which would contain only about ten grams
109 Martin Forwood, Research Report No. 5, International Panel on Fissile Materials, e Legacy of Reproc-
essing in the United Kingdom, International Panel on Fissile Materials, July 2008.
101 Federation of American Scientists
Prospects for a Plutonium Economy in the United States
of plutonium.110 Recent experience with severe delays and cost overruns at Japan’s Rokkasho reprocessing plant suggest that technical advances will not dramatically reduce
costs soon. Fresh uranium and the required enrichment together are significantly cheaper
than extracting plutonium from used fuel and will most likely remain so for at least several decades and probably to the end of the century.
There are also fundamental physical limits on plutonium recycling. As the plutonium is passed back through the LWR, some is fissioned, but some simply absorbs a
neutron without splitting thereby creating heavier isotopes of plutonium; for example,
Pu-239 can be converted to Pu-240, and these isotopes are not readily fissioned in an
LWR; indeed they can absorb neutrons and inhibit the operation of the reactor. Just as in
the case of uranium, different isotopes of plutonium are chemically identical and cannot
be easily separated. In practice, therefore, the plutonium is usually passed through an
LWR only once because of this build up of heavier isotopes.111 The nuclear waste products from MOX fuel must then be pulled from the recycling loop, and they are not significantly different than wastes from pure uranium fuel and must be disposed of in some
manner, almost certainly in a deep geological repository. Reprocessing and recycling the
plutonium from LWR fuel increases by only one sixth the amount of energy extracted
from the uranium.
Special Plutonium Reactors
To fully exploit the energy potential of plutonium requires both reprocessing and a different type of nuclear reactor, one that can fission uranium plus all the isotopes of plutonium, as well as other artificial transuranic elements, such as americium, that are bothersome components of normal nuclear waste. Such a reactor depends on allowing the neutrons to react while they are still at high energy, that is, fast neutrons, hence these are called
fast reactors (as opposed to LWRs, which use slow or thermal neutrons and are called
thermal reactors). Fast neutrons can split most heavy atoms, including the heavier isotopes
of plutonium that an LWR will not. Moreover, the fissioning of plutonium by fast neutrons produces more neutrons than other reactions do, making breeding most effective.
110 Steve Fetter and Frank von Hippel, “Is U.S. Reprocessing Worth the Risk?,” Arms Control Today, Sep-
tember 2005, accessed 10 December 2010.
111 However, France has performed a twice reprocessing on some plutonium but prefers to reserve the
once-recycled material for future fast neutron reactors if and when these become commercially available.
Federation of American Scientists 102
The Future of Nuclear Power in the United States
February 2012
Fast reactors date back to the beginning of the nuclear age. e United States,
Soviet Union, and now Russia, France, Germany, Japan, and Britain have experimented
with fast reactors. (India has plans to develop them.) Global research and development
on fast reactors is conservatively estimated to be at least $50 billion,112 but no fast reactor has been commercially successful.113 Fast reactors are inherently more complex than
LWRs, and designers believe that any eventual fast reactor will be at least a quarter
more expensive than a comparable LWR. Current fast reactors designs are cooled by
liquid sodium, which is inflammable and can explode upon contact with water, raising
additional safety questions.
e failure of fast reactor development explains existing reprocessing efforts.
It is reasonable to ask why, if reprocessing is uneconomic and inappropriate for LWRs,
France does it and Britain did until recently. Both programs go back decades. In the
1950s, Britain, France, and the United States were working toward a commercial reprocessing capability and, in parallel, fast breeder reactors that could fully exploit the
plutonium produced. e French built the Rapsodie reactor, followed by the Phenix,
which was reasonably successful as a demonstration but too small to be commercially
viable. is was followed by the Superphenix, a 1.2 gigawatt reactor that was a technical
and commercial failure. e British had a similar experience; the Dounreay Fast Reactor (DFR) was a reasonably successful small demonstration, but the scaled up version,
the moderately-sized Prototype Fast Reactor (PFR), was far less reliable.
e United States began a demonstration fast reactor at Clinch River, Tennessee,
near Oak Ridge. When the cost exploded several fold, Congress cancelled the program in
1983. But only in the United States was the parallel reprocessing program also cancelled.
Presidents Ford and Carter actually made opposition to reprocessing a government policy,
primarily because of fears that widespread reprocessing would increase the risks of nuclear
weapon proliferation. President Reagan rescinded the ban, allowing commercial reprocessing. But Congress did not reinstate government financial support, and industry showed no
interest in restarting reprocessing.
112 omas B. Cochran, Harold A. Feiveson, Walt Patterson, Gennadi Pshakin, M.V. Ramana, Mycle
Schneider, Tatsujiro Suzuki, and Frank von Hippel, Research Report No. 8, International Panel on Fissile
Materials, Fast Breeder Reactor Programs, History and Status, International Panel on Fissile Materials, February 2010, p. 6.
113 Some Russian engineers and officials may disagree, but the BN type of Russian fast reactor has not been
cost competitive with Russian LWRs.
103 Federation of American Scientists
Prospects for a Plutonium Economy in the United States
Because of persistent optimism about their fast reactor programs, Britain and
France continued with their reprocessing efforts but in the end no fast reactor materialized to exploit the plutonium produced. As a result, the British today have an inventory
of 100 tons of separated plutonium,114and the French hold 80 tons of separated
In spite of British and French experience, the Japanese are now proceeding down
the same path, building a reprocessing facility at Rokkasho (as mentioned earlier) and a
fast neutron reactor, the Monju. Both have faced severe technical problems, schedule delays, and cost overruns. The Monju reactor was shut down in 1995 after a sodium fire and
has not yet restarted normal operation.
Plutonium as Waste
Plutonium is a dangerous component of radioactive waste. When nuclear fuel is fresh
from the reactor, its most intense radiation is emitted by fission products, but these tend
to fade over timescales of hours to decades; plutonium and other transuranics are more
important to the long-term danger over many years to centuries. In addition, one of the
challenges of long-term geological disposal of nuclear waste is the heat released by the
waste, which can, over decades, raise temperatures to dangerous levels. High temperatures
can threaten containment barriers and energize migration of radioactive material. Again,
after a few decades, most of the heat comes from plutonium and other transuranics. Finally, plutonium can be used to make weapons, and reactor waste is a potential source of
that plutonium.
To deal with the dangers of plutonium, the George W. Bush administration proposed, as part of the Global Nuclear Energy Partnership (GNEP), to separate plutonium
and other transuranics and to burn them in fast reactors. These reactors were not breeders. Quite the opposite, they were burner reactors designed to consume plutonium without producing more. Plutonium would be separated from LWR fuel, and for every three
114 International Atomic Enercy Agency, Vienna, INFCIRC/549/Add.8/11,
Communication Received
om the United Kingdom of Great Britain and Northern Ireland Concerning Its Policies Regarding the Management of Plutonium Statement on the Management of Plutonium and of High Enriched Uranium, 2 July
115 International Atomic Enercy Agency, Vienna, INFCIRC/549/Add.5/14, Communication Received om
France Concerning its Policies regarding the Management of Plutonium Statements on the Management of
Plutonium and of High Enriched Uranium, 8 September 2010.
Federation of American Scientists 104
The Future of Nuclear Power in the United States
February 2012
or four LWRs, a burner reactor would use the plutonium as fuel to eliminate it by burning
it. A single pass through the burner would not consume all the plutonium so the burner
reactor fuel itself would have to be recycled several times to completely eliminate the
trasnsuranics. Eventually, only fission products would remain—these are an inevitable
product of nuclear fission—and these could be sent to a geological repository.
These proposals failed because of the projected high cost of separation and,
again, the needed fast reactor was not available. Given the world’s inventory of plutonium, several demonstration fast reactors could operate for decades to prove their feasibility before any additional plutonium would need to be separated from existing stocks of
used fuel.
The GNEP burner proposal was politically attractive, at least in part, because it
seemed to reduce the demand for a geological repository just at the time that the Yucca
Mountain geological repository was meeting ever stiffer political resistance. This may also
explain the support for reprocessing in France, South Korea, and Japan, where political
opposition to permanent disposal is intense. Reprocessing does not eliminate the need for
a geological repository and, without a fast reactor, does not even much reduce the need
for storage volume, but it does put off for a few decades the political decision about
what—and where—a permanent solution should be.
Plutonium for Military Use
In the United States, the production of pure plutonium has been dominated by the
requirements of nuclear weapons. e world’s first nuclear explosion, on July 16, 1945
near Alamogordo, New Mexico, was powered by plutonium. While nuclear weapons
can also use highly-enriched uranium (the Hiroshima bomb was powered by uranium),
plutonium is much preferred. It is a complex material and working with it can be difficult, but plutonium allows far more compact and powerful bombs. It is particularly important, for example, in making powerful thermonuclear bombs that are small enough to
be able to fit several atop a single missile. Nuclear analysts believe that most large
currently-deployed nuclear weapons use plutonium.
In the half century between 1944 and 1994, the United States produced almost
111 tons of separated plutonium metal, almost all of it for the nuclear weapons program,
and current inventories are one hundred tons.116 The nation’s arsenal of nuclear weapons
116 Plutonium:
e First Fiy Years, op. cit., p. 22.
105 Federation of American Scientists
Prospects for a Plutonium Economy in the United States
has declined dramatically since the height of the Cold War, and the United States
stopped production of new weapons-grade plutonium 1994.
During the Cold War, old nuclear weapons were continually being dismantled
but the plutonium was recovered and incorporated back into new weapons. Now, with
the total number of weapons declining, an inventory of unused and unneeded plutonium is building up. Dismantlement continues and thousands more warheads are
awaiting dismantlement (the exact numbers are secret), but perhaps more than 14,000
plutonium “pits,” the cores of nuclear weapons, are being stored at the Pantex nuclear
weapon facility.117 Several alternatives have been considered for disposal of the excess
plutonium, including burying it deep underground.118 Current plans are to mix 34 tons
of the plutonium with uranium fuel in a special facility in Savannah River, South Carolina, and then burn it in commercial nuclear reactors.119 A ton of plutonium can power
a typical commercial reactor for a year. e Russians have agreed to dispose of an equal
amount of plutonium from their retired weapons. Even with an inventory of a few
thousand active and stand-by nuclear weapons, destroying 34 tons will still leave an
inventory of tens of tons of weapons-grade plutonium in storage. Future arms control
agreements may include limits on non-deployed, retired nuclear weapons and inventories of plutonium, requiring more plutonium to be destroyed. People who oppose developing a commercial reprocessing and MOX economy may support destruction of
weapons plutonium but are concerned that this small-scale MOX project will eventually provide the foundation for a larger commercial operation.
Plutonium is inevitably produced in any reactor that has uranium-238 in the fuel,
which today means every commercial reactor. Discussions of plutonium are oen complex because people cannot even agree on what it is: an energy source with almost tremendous potential for generating electricity, a dangerous poison that must be locked
117 Robert S. Norris and Hans M. Kristensen, “U.S. Nuclear Forces, 2009,” Bulletin of the Atomic Scientists,
March/April 2009, vol. 65, no. 2, pp. 59–69.
118 United States Department of Energy, Nonproliferation and Arms Control Assessment of Weapons-Usable
Fissile Material Storage and Excess Plutonium Disposition Alternatives, January 1997, p. 127.
119 National Nuclear Security Administration, Fact Sheet, NNSA’s MOX Fuel Fabrication Facility and U.S.
Plutonium Disposition Program, Aug 3, 2010., accessed 15 December 2010.
Federation of American Scientists 106
The Future of Nuclear Power in the United States
February 2012
away, or the explosive power of a nuclear bomb?
Regardless of plutonium’s potential in the next century, there are some questions that have to be answered today but many other important questions that can and
should be put off until technical developments provide a clearer path ahead. Even with
a robust growth in nuclear power, it seems certain that uranium supplies, at affordable
prices, will be adequate for several decades, so separating plutonium from commercial
reactor fuel to consume it in commercial nuclear power reactors will not be economic.
If greater energy extraction were the goal, then “deep-burn” LWRs that start with more
highly enriched fuel and burn it longer, which breed and burn plutonium in situ are
probably a better approach and do not require reprocessing.
To fully exploit the potential of plutonium requires fast neutron reactors. If
fast breeder reactors are the long term goal, then research on these reactors should continue but this does not mean plutonium separation is needed now. Any reactors that
are built can be fueled for decades by existing inventories of plutonium. If fast reactors
ever prove practical and are built in significant numbers, production of plutonium
could resume and, given the cushion provided by existing stockpiles, always keep ahead
of demand.
Using fast neutron reactors as burners only makes sense if the disposal of
transuranics becomes essentially impossible, perhaps for political if not technical reasons. If, a century hence, breeders form the backbone of the nuclear industry, then the
last thing we should be doing today is burning plutonium.
Given the technical uncertainties, making an irreversible decision today is illadvised and it is unnecessary. While there is universal agreement that some form of
long-term waste repository will be required, wastes may not need to be committed to a
repository right away. Without a long-term repository, the cooling pools at some reactors are filling up, and some utilities have moved waste to above-ground dry casks.
ese are currently licensed by the Nuclear Regulatory Commission for twenty years at
a time, but many analysts believe they should be stable for at least a century or longer.
Pending resolution of questions regarding long-term geological storage or fuel for fast
reactors, the plutonium can sit in the used fuel rods where it is safe from the and cannot be used for weapons.
107 Federation of American Scientists
Required Infrastructure for the Future of Nuclear Energy
Chapter 8
Required Infrastructure for the Future of Nuclear
by Andrew C. Klein
e infrastructure needed for revitalized nuclear energy production in the United
States can be defined along three general lines: industrial/utility infrastructure including the manufacture/construction of new power plants and the supply chain for both
construction and operation; a research and development (R&D) infrastructure including facilities and capabilities in the national laboratories and universities; and the personnel development infrastructure including the university education system and the
training programs in industry.
e industrial infrastructure to support the deployment of nuclear power
plants in the United States grew up with the fledgling nuclear industry in the 1960s
and 1970s as plant concepts were developed, tested and deployed. Power plant vendors
and their suppliers developed the necessary supply chain for large- and small-scale
components and parts while the utility industry built and operated these large industrial facilities and became comfortable with their operation. In the 1970s this supply
chain had become a mature industry and numerous new plants were coming on line in
quick succession. However, triggered primarily by the accident at ree Mile Island
(TMI) Unit 2 in March 1979 and reduced electricity demand throughout the 1980s,
the nuclear power industry went into an extended hiatus of new plant development
and construction. is hiatus in nuclear construction turned much of the U.S. industry
into suppliers of services instead of suppliers of components and parts. Replacement
parts were still available for the nuclear utilities to continue operation, but the domestic
supply lines for major components such as steam generators, reactor vessel heads, etc.
became a much more international supplier network, dominated by French, Japanese,
and South Korean suppliers.
Federation of American Scientists 108
The Future of Nuclear Power in the United States
February 2012
In a similar fashion the nuclear research and development (R&D) infrastructure (both in national laboratories and in industry) grew up from the 1950s through
the 1970s to support the development of new reactor concepts, fuel cycles, and reactor
designs. Many of these included the construction of testing and demonstration facilities, mostly centered on the application of different reactor cooling technologies and
neutron physics. Concerning cooling technologies, the design concepts that were tested
and demonstrated included using “light” water (the most common and lightest form of
hydrogen bonded with oxygen) or “heavy” water (deuterium, a heavier form of hydrogen, bonded with oxygen). Both forms of water provided needed cooling for the reactor core, but had different abilities to slow down, or moderate, neutrons that governed
the occurrence of nuclear reactions inside the core. Other cooling techniques involved
helium and other gases. Concerning neutron physics, in addition to thermal reactors
that used moderated neutrons, design concepts investigated fast neutron reactors that
would use sodium or other liquid metals for removing heat from the core. is was a
very heady and exciting time for the development and testing of different nuclear reactor concepts – for applications ranging from power production, radioisotope production, and deployment of nuclear technologies underwater, on the seas, in the air and in
Also during this time, the educational and training infrastructure grew up with
the nuclear industry as universities added nuclear engineering technology, undergraduate and graduate programs, and utilities developed training programs to supply their
workforce needs. Universities developed the facilities and technical capabilities to educate students and perform research that was useful to the national nuclear energy development program. Radiation measurement laboratories, both teaching and research,
and research reactors became established on university campuses to help educate scientists, engineers and technicians for the nuclear workforce. Many of these facilities were
financed by the federal and state governments to build the personnel development infrastructure for the nuclear industry and to meet their own research and regulatory
responsibilities. By the late 1970s more than 60 universities had nuclear engineering
educational programs and there were more than 50 university-based research reactors
scattered across the United States. Many of the early nuclear engineering educational
programs grew from the efforts of Manhattan Project veterans, the early submarine
developers at Bettis and Knolls Atomic Power Laboratories, and the early power plant
designers at General Electric and Westinghouse. Especially noteworthy was the Oak
109 Federation of American Scientists
Required Infrastructure for the Future of Nuclear Energy
Ridge nuclear power school and the establishment by the U.S. government of the
Atomic Energy Commission (AEC) fellowship program.
However, aer TMI and the reduced demand for electricity that accompanied
the oil crises in the 1970s, there was a significant drop in the numbers of students interested in a nuclear engineering or technology education, a commensurate closure of a
number of high quality academic programs and the closure of one-half of the operating
research reactors during second half of the 1980s and early 1990s. us, by 2005 only
about 25 academic programs and 27 research reactors remained on university campuses. An especially notable decline was in the number of relevant engineering technology programs whose purpose was expressly to prepare students for skilled trades and
technician roles within the nuclear power industry. By the late-1990s only three accredited nuclear engineering technology programs remained in the United States at Excelsior College in New York, University of North Texas in Texas, and ree Rivers Community College in Connecticut.
As the nuclear construction boom ended in the mid-1980s, many nuclear
R&D operations in the national laboratories found it difficult to sustain the necessary
infrastructure and facilities for nuclear energy development. The U.S. government understood that nuclear energy facilities and reactors were expensive to maintain, let alone
refurbish or re-build, and many nuclear experimental facilities were put into maintenance mode.
One important and positive development post-TMI was the establishment of
the Institute for Nuclear Power Operations (INPO) in December 1979. INPO’s mission is to “promote the highest levels of safety and reliability – to promote excellence –
in the operation of commercial nuclear power plants.” ey accomplish this mission by
developing performance objectives, criteria and guidelines for the nuclear power industry and then evaluate every power plant against these objectives. With respect to human capital development, INPO, through its National Academy for Nuclear Training
and the independent National Nuclear Accrediting Board, performs training and support for the nuclear industry and the accreditation of the training programs at every
nuclear power plant in the United States.
Around 2005 a noticeable change in the level of interest in nuclear power took
place. Driven primarily by global warming concerns and the need for increases in electricity generation without the production of greenhouse gases, nuclear energy was
again considered to be a potential electricity supplying technology by both the utility
Federation of American Scientists 110
The Future of Nuclear Power in the United States
February 2012
community and the U.S. government. e government began to publicly acknowledge
the need for a role for nuclear power in the energy mix of the United States, and utility
executives began planning to build new nuclear power plants. Some of this was fueled
by funding levels for federal loan guarantees for the construction of new nuclear power
plants and for research and development that started to become available from the U.S.
Department of Energy (DOE) in the late 2000s. is increased level of interest in new
nuclear power has continued to the present.
At the same time, many more students entered the remaining university nuclear engineering programs, in some cases more than quadrupling the enrollments seen
just a few years earlier. Utilities began working together to resurrect technician training
and skilled trades development programs, mostly local to their operating power plants,
but also collectively in anticipation of the construction of new nuclear power plants.
ese new programs are aimed at producing graduates able to work as technicians, especially in electrical, mechanical, chemical, welding, and in some cases nuclear vocations. ese specialists will be needed to operate the current fleet of nuclear power
plants as well as be capable of building and operating a new generation of nuclear
power plants. New academic programs began to be discussed by universities and DOE,
and the Nuclear Regulatory Commission (NRC) began new university research, development, infrastructure and student support programs that attracted new universities
and students into nuclear energy related academic programs.
Current Status
Currently, there are 104 operating reactors in the United States and an existing, robust
supply chain available to continue the operations of these plants through their licensed
lifetimes. Most of these reactors have either received or will receive a 20-year license
extension to extend their operational lifetime to at least 60 years. Additional license
extensions are also possible beyond 60 years. Research and development activities are
currently being conducted to determine the information and regulatory needs to enable these additional license extensions.
New reactor license applications have been submitted to the NRC with a few
additional units and sites still under consideration. As of 2010, the industry has expressed interest in as many as 26 new reactor units at 17 different sites. However, it is
uncertain how many of these reactors will actually be completed over the next 20 years,
but many of these are quite possible. e highest probability would be for those who
111 Federation of American Scientists
Required Infrastructure for the Future of Nuclear Energy
receive loan guarantee assistance from DOE. Financing these large projects is the greatest impediment to their completion. Successful completion, start-up, and operation
may give other sites and plants the opportunity to consider completion without loan
guarantees, but this is yet to be determined. Additionally, at least three new sites have
been chosen and significant progress has been achieved in the development of new uranium enrichment facilities across the United States. Some of these projects have already
received loan guarantees, which would increase the likelihood of their successful construction and operation.
ese new plants will require a robust supply chain of nuclear manufacturers.
Nuclear power plants are complex undertakings that require hundreds of components
and subcomponents and the construction of numerous new power plants will, according to the Nuclear Energy Institute (NEI), “require a deep and diverse supplier base.”
With the idea of supporting the maintenance of existing nuclear suppliers and the development of new suppliers NEI has developed a “Supply Chain Map of Nuclear Reactor Components.”120 is map is broken down into four main categories: nuclear island, turbine island, balance of plant, and site development and construction and is
aimed at assisting new suppliers to identify where their products fit into the components and subcomponents of new nuclear reactor designs and to enable them to better
understand the quality requirements for these components.
e quality standards for many nuclear power plant components, particularly
those critical to reactor safety, have been well-established. ese standards and requirements ensure that a nuclear facility’s structures, systems, components and controls
can be relied on to be functional and operational under the most rigorous safety conditions. ose components that are subject to these rigorous standards are oen known
as “nuclear-grade” or “safety-related” components. Any manufacturer of these components must have appropriate quality assurance (QA) programs in place in order to ensure that the standards are met.
With respect to the QA programs for new reactors, NRC will review and inspect these programs and their implementation for nuclear steam system suppliers,
architect-engineering firms, suppliers of safety-related and commercial-grade products
and services, all calibration and testing laboratories, and all holders of NRC construction permits, operating licenses, and combined licenses. By conducting inspections, the
120 Nuclear Energy Institute, “Supply Chain Map of Nuclear Reactor Components,”
Federation of American Scientists 112
The Future of Nuclear Power in the United States
February 2012
NRC’s main objective is to determine whether licensees and their contractors
are meeting the agency’s requirements through the implementation of procedures, recordkeeping, inspections, corrective actions, and audits.
An important development in the management of the nuclear R&D infrastructure within the DOE was the designation in 2004 of DOE’s Office of Nuclear
Energy as the Lead Program Secretarial Office (LPSO) for the Idaho National Laboratory (INL) and the identification of the consolidated INL as the lead national laboratory for nuclear energy science and technology development. is has allowed the Office of Nuclear Energy to focus on maintaining and developing the needed R&D capabilities and facilities at INL. Some support also remains for the other national laboratories that have significant capabilities in certain aspects of nuclear energy development,
most notably at Oak Ridge (ORNL), Argonne (ANL), Sandia (SNL), and Los Alamos
(LANL) national laboratories.
Currently, the nuclear industry uses a wide spectrum of high school, community
and technical college as well as university graduates from the B.S. to Ph.D. The nuclear
industry, consisting of the vendors and manufacturers, the architects, designers and engineers, the construction companies, and the utilities and power plant operators, use a
broad spectrum of educated and trained personnel. Government agencies such as DOE
(and its national laboratories), NRC, the Department of Defense, the National Nuclear
Security Administration, the Defense Intelligence Agency, the Defense Threat Reduction
Agency, the Federal Bureau of Investigations, the Food and Drug Administration, the
U.S. Department of Agriculture, the U.S. Department of State, and the Environmental
Protection Agency all employ people who have education and training in nuclear specific
areas. Finally, academia also requires people to fill roles as faculty, staff, and researchers.
e nuclear industry requires many people with a broad spectrum of education and training backgrounds. e range of skilled people needed include engineers
(not only nuclear, but mechanical, structural, chemical, electrical, civil and construction engineers), health physicists, plant managers, lawyers, managers and accountants,
skilled trades-people in mechanical, electrical, maintenance, and construction, nuclear
safeguards and security experts, architects, risk assessment specialists, and others. NEI
has conducted workforce surveys over the past decade and identified the most difficult
employee characteristics to recruit into the nuclear utility industry. ere is little reason to expect that this will be any different for new plants. e people that NEI has
identified as being hardest to find, in order of difficulty, are female candidates, minority
113 Federation of American Scientists
Required Infrastructure for the Future of Nuclear Energy
candidates, nuclear engineers, experienced designers, non-destructive evaluation technicians and health physicists. Specific emphasis has been placed on developing targeted
recruiting in many of these areas through awarding scholarships and fellowships, internships and other methods of supporting students. Also DOE has developed educational modules aimed at middle and high school students to enhance their awareness of
these high paying career opportunities.
e Future
Looking to the future, or more specifically the next 10 to 20 years, there will be a continual need to revitalize the supply chain for the nuclear industry. Since the market for
new nuclear power plants is fully established as a global one, the supply chain will continue to develop with an international character. U.S. utilities and consumers of electricity will benefit from a competitive market for the supply of nuclear reactor systems,
parts, and components. U.S. suppliers of these systems, parts, and components must be
enabled to effectively compete if they are to remain strong participants in this marketplace. Also, as the new reactor concepts become better developed and understood, the
supply chain will need to be developed to ensure that suppliers with appropriate quality
programs are prepared to deliver the needed equipment, people, and services. Continuing and potential future bottlenecks in the supply chain will probably include large
forging capabilities for pressure vessels, steam generators, and other large components
as well as for some of the specialty materials needed by the nuclear industry. In some
cases, such as large forging capabilities, the United States will continue to rely on foreign manufacturers of these important components of nuclear power plants.
e emergence of the technologies for new small- and medium-sized, modular
reactors that respond to the high initial capital cost of building new power reactors is a
welcomed recent development. ese highly creative design concepts, some of which
are new ideas while others are older concepts that have been reinvigorated by the prospects of new market possibilities that may come along through small reactors, are aiming
to develop a new financial paradigm for the nuclear industry by enabling utilities, or even
non-electricity generating owners, to incrementally add small unit sizes, typically ranging
from 10 to 300 megawatts of electricity. The idea is that smaller unit sizes can be added
one at a time, more quickly, and in a modular fashion, rather than requiring the construction of individual large-scale power plants. The developers of these new, smaller concepts
are proceeding toward design certification with the NRC and developing the financial
Federation of American Scientists 114
The Future of Nuclear Power in the United States
February 2012
and business cases that may allow their installation by customers, both domestic and
overseas, who would not have considered nuclear power in large increments because
their financial situations or electric grid capabilities would not enable them to accommodate the introduction of large single-unit power plant additions.
One potential limiting capability will be the development of the people who
are educated and trained to operate these new small reactor systems. e leading concepts being considered are evolutionary developments from current light water based
nuclear reactors and the skills needed to operate these systems may not be far from
those needed to operate current technologies. However, testing facilities will be needed
for these new concepts, in both integral and separate-effects forms, to provide validation and verification of the computer codes used to predict their performance during
both normal and accident conditions.
A few special technologies and materials are important to the new nuclear
energy industry and may need special attention to ensure their availability when they
are needed. Specialty materials, such as zirconium, hafnium, gadolinium, beryllium,
and others, will need suppliers to provide processing, manufacturing, and recycling
technologies that are cost-effective to the manufacturers and utilities building new nuclear power plants. Some, but not all, of these specialty materials have other uses in the
economy but their availability to the nuclear industry needs to be ensured.
Today’s nuclear R&D infrastructure in the nation’s national laboratories is
rather aged. Many of the nuclear R&D facilities across the complex of national laboratories were originally developed in the 1960s and 1970s. However, while they may be
old, many critical facilities have seen reasonable maintenance and upgrades over the
years so that a basic capability remains available. DOE continues to review its infrastructure needs on a regular basis, including updates to the ten-year site plans at each
national laboratory and facility reviews conducted by the National Academies of Science and Engineering, the DOE Nuclear Energy Advisory Committee and others.
ese reports periodically give the government and the public insight into the capabilities and needs of the nuclear energy R&D community and are used by DOE to guide
their annual budget requests to Congress. All of the facilities that researchers might
want may not readily be available, but a basic infrastructure has been maintained for
R&D activities and a process for their maintenance and expansion is available annually
to DOE.
115 Federation of American Scientists
Required Infrastructure for the Future of Nuclear Energy
A few skilled technical areas related to construction of new nuclear power
plants have not been used over the past 20 years in the United States. Since very few
new plants have come on-line, there has been little need for people trained in nuclear
plant construction and plant startup/test engineering. ese highly specialized skills
previously were available while new plant projects were being brought on-line during
the 1970s and 1980s; however, new education and training programs will be needed to
make sure that people are ready when the new plants begin to load fuel and contemplate full operation. Also, should the recycling and reuse of nuclear fuel reach a mature
stage of development over the next 30 years, there will be a significant need for radiochemists and radiochemistry technicians, and the development of education and training programs for recycling facility engineers, technicians and operators.
Competing interests for a top quality workforce will come from various sectors, both inside and outside of the nuclear industry. e electric utility industry, including all means of production and distribution of electricity will look for similarly
educated and trained personnel. e defense, telecommunications, oil and natural gas
industries will also be searching for highly educated and trained workers. However,
utility careers are sometimes viewed by students to be low-technology career paths of
lesser excitement when compared to other high-technology options, and thus the electric utilities must offer competitive compensation packages in order to recruit the best
personnel into the nuclear industry.
One important aspect of the nuclear energy pipeline for both personnel and
equipment is the long design lifetimes for nuclear power plants relative to the length of
time that is typical for any one individual. Current nuclear power plants have initial design and license lifetimes of 40 years. Most, if not nearly all, currently operating nuclear
power plants in the United States will receive a 20-year license extension from the NRC.
Some of these plants may be able to gain an additional 20-year license extension, if current
research and development activities show that they can clearly be operated in a safe manner. The new power plant designs all have initial design lifetimes of 60 years, and conceivably their licensed lifetimes could extend to 80 or 100 years. If five to 10 years are required to construct a plant and then another five to 10 years to decommission it, the
plant’s total product lifetime approaches 110 to 120 years from conception to dismantlement. This is considerably longer than the product lifetime for any other industrial product. Compare this to the roughly 40-year productive career that is typical for most workers. This difference emphasizes the need for continuous education and training of the
Federation of American Scientists 116
The Future of Nuclear Power in the United States
February 2012
nuclear workforce.
Universities that teach nuclear engineering are all “research intensive,” and faculty
are often focused upon getting research grants and contracts and then performing the research. They use graduate students and postdoctoral researchers to accomplish this research,
which typically is on the cutting edge of the nuclear energy industry. The financial structure
of the modern research university puts great emphasis on research funding to support the
activities of the faculty and students which often forces them to focus on finishing the current research while constantly looking for the next grant or contract. Over the years, this
combination of driving forces within the modern research university with a narrow focus
available for research support from government and industry sponsors may yield a nuclear
science and engineering faculty that is insufficiently diverse to teach the broad topical areas
that are needed by the nuclear industry. Students coming through the U.S. academic programs are often exposed to advanced concepts, without obtaining a clear understanding of
the current nuclear power fleet and the types of reactors that are likely to be operational during their careers. Schools that have access to commercial plant simulators and that are able to
incorporate them into their academic programs can provide a significant stimulus for students to become directly interested in working in the current nuclear power fleet.
Finally, there is a significant continuing need for educational and training infrastructure in order to maintain the appropriate technically relevant nuclear workforce. This
infrastructure in the college and university systems includes state-of-the-art classrooms and
teaching laboratories, instrumentation and radiation measurement capabilities, heat transfer,
fluid flow and radiochemistry laboratories, system operation and simulators, and computational capabilities. An increasingly important aspect of today’s education and training infrastructure is the ability to deliver and receive distance education and classes that are available
for students to take at any time of the day or night. While on-site resident education will
remain an important aspect of the university development of scientists and engineers, these
new delivery mechanisms are rapidly and continuously increasing in quality and capability.
In the training realm plant simulators and hands on training for maintenance and installation of plant components such as valves, pumps, piping, fuel handling tools, etc. are needed.
One particular type of simulator that is very useful is a radiation area simulator which enables workers to become accustomed to and familiar with working around radiation, point
and surface radioactivity, and urface contamination, etc. The aim here is to best prepare
workers for working with and around radioactive materials and remove their fear of the unknown.
117 Federation of American Scientists
Alternatives to Nuclear Power
Chapter 9
Alternatives to Nuclear Power
by Richard Wolfson
What sources might substitute for nuclear energy? at depends on who’s asking. If you’re a utility executive, nuclear energy’s chief virtues are low cost and its supply
of baseload power. Nuclear plants, that is, run at full power nearly all the time, supplying loads that are always present. And nuclear plants are large, capable of powering entire cities. So a utility looking to replace a nuclear plant with “plug and play” economic
and grid compatibility is likely to seek large, low-cost, steady, reliable sources—criteria
best met by coal- and gas-fired plants.
If you’re a consumer or business watching your bottom line, you’ll want cheap
replacement power. at, too, is likely to come from fossil-fueled power plants, unless
you’re in a region with ample hydroelectric resources. Natural gas affords the lowest
capital costs, but gas prices fluctuate. Coal prices are steadier, but capital costs of coal
plants are greater. Economic and regulatory conditions might greatly alter this comparison; tightened environmental constraints, a carbon tax, or a requirement for carbon
capture and sequestration could make coal less attractive economically.
An environmentalist will value sustainability and minimal environmental impact—especially low carbon emissions. Wind and solar, accounting for a small but rapidly growing share of the U.S. electrical energy mix, would be leading choices. But these
renewable energy sources, unless linked through “smart grid” technologies covering vast
geographical areas, are too intermittent to qualify as baseload power. So an environmentalist interested in replacing nuclear plants with solar and wind will also advocate a
substantial upgrading and “smartening” of the electrical grid.
A state utility commissioner, concerned for stability in pricing and availability
of future energy supply, might want to keep the low-cost energy from established nuclear plants in the mix. But if nuclear energy must be replaced, a prime utilitycommission choice might be long-term contracts with big hydroelectric producers—if
they’re available for import from nearby states or, as with the vast hydro resources in
Federation of American Scientists 118
The Future of Nuclear Power in the United States
February 2012
Canada, across national borders.
Nuclear energy itself could be a substitute for today’s generation of light-water
fission reactors. Advanced nuclear alternatives include breeder reactors that would
greatly expand the nuclear fuel resource; fast-neutron reactors that could “burn” nuclear waste; fission-fusion hybrids; and, ultimately, pure fusion reactors that could, if
economical, provide nearly limitless energy for billions of years. But, with the exception of a worldwide handful of breeders with spotty operating records, none of these
nuclear alternatives is ready for commercial use.
e answer to the “nuclear substitute” question also depends on scale and timing. If you’re talking about replacing a single nuclear plant, then you have available the
whole range of conventional energy sources, or you can move in new directions. ere’s
power to spare in today’s electric generating industry, so your choices today aren’t necessarily constrained by availability. But if you’re the entire United States, with 20 percent of
its electrical energy from nuclear plants, then a decision on substitute power may face
serious constraints. Timing matters, too: If you’re talking about replacing the entire nuclear power enterprise over several decades, then there’s time to develop new sources, conventional or otherwise. But if you’re looking at nuclear plants that will go offline in the
next few years, then you’ll have to scramble to line up replacement sources—whether
than means contracting with existing suppliers or developing new generating capacity.
Finally, you might prefer an energy strategy that does away altogether with the
need to substitute for nuclear power, by reducing demand to the point where the nuclear contribution is no longer needed. at’s an admirable goal, but with nuclear
plants providing 20 percent of the United States’ electricity, it’s not one that could be
implemented overnight.
Getting Quantitative
Whatever your choice or choices for nuclear substitutes, they’ve got to meet one rigid
criterion: ey must be capable of supplying energy—reliably and almost continuously—at the same rate that the nuclear industry does now. Power is the rate of energy
supply or consumption, so that means the power output of your substitute sources must
at least equal that of the nuclear plants you’re replacing.
In 2010 there were 104 commercial reactors operating in the United States,
down from a peak of 112 in 1990. ese reactors produced just over 20 percent of U.S.
electrical energy—a yearly total of some 800 terawatt-hours (TWh; see Box 1). e
119 Federation of American Scientists
Alternatives to Nuclear Power
104 reactors had a combined electric generating capacity of 101 gigawatts (GWe)—
meaning that if they ran continuously, the electrical energy they produced, in gigawatthours, would be that power multiplied by the number of hours in a year: 885,000 GWh
or 885 TWh. Comparing with their actual 800-TWh annual energy production shows
that the reactors’ average capacity factor was nearly 91 percent—meaning that, on average, each reactor produced its rated power 91 percent of the time. Bottom line: Substituting alternative sources for the entire U.S. nuclear energy enterprise would require
providing electrical energy at the rate of 800 TWh per year. Nearly all suitable nuclearreplacement energy sources are already represented in the United States’ electrical energy mix, shown in Fig. 1, although in the long term additional sources could become
Figure 1 Sources of electrical energy in the U.S. (Source: U.S. Department of Energy, Annual Energy Review 2009,
Table 8.2a, 2009 data).
Power—the rate of using or generating energy—is measured in watts (W).
Electric power plants have typical outputs ranging from a few megawatts (millions of
watts; MW) to several thousand megawatts; 1000 MW is 1 gigawatt (one billion watts;
GW). e United States’ total energy consumption rate is about 3.5 terawatts (TW, or
trillion watts, with 1 TW = 1000 GW). at’s just under one-fourth of the world energy consumption rate of 16 TW, and it amounts to about 10 kW per capita.
Federation of American Scientists 120
The Future of Nuclear Power in the United States
February 2012
Is that 10 kW per day, or per year, or what? Neither; it’s a rate; “per time” is
built in. e average U.S. resident uses energy (all forms, not just electricity) at the rate
of 10 kW, round the clock, day in and day out.
Since power is energy per time, energy can be measured in watts multiplied by
time. One such unit is the kilowatt-hour (kWh), familiar from your electric bill. One
kWh is the energy consumed by a 1000-W stove burner, left on for one hour. It’s also the
energy consumed by the average U.S. resident in 6 minutes (10 kW for one-tenth of an
hour). For larger amounts of energy, we use similar combinations of power multiplied by
time; for the U.S. nuclear industry, for example, yearly energy outputs are conveniently
measured in terawatt-hours (TWh, equivalent to 1012 watt-hours or a billion kWh).
e average efficiency of electrical energy generation in the United States is
between 30 and 40 percent, due to inefficiencies in the fossil-fueled and nuclear plants
that produce nearly all our electricity. at means power plants typically convert to
electricity only 30-40 percent of the energy in their fuels. us, for every kilowatt-hour
of electrical energy a nuclear plant produces, it discharges about two kWh of energy as
waste heat. is inefficiency leads to the distinction between electrical output and
thermal power extracted from fuel. A power plant rated at 1-GWe, for “1 gigawatt electric,” would extract fuel energy at the rate of about 3 GWth (“3 gigawatts thermal”). But
only one of those gigawatts gets converted to electricity.
e Fossil Alternative
In 2009, fossil fuels accounted for 70 percent of electric power generation in the
United States—coal for 45 percent, natural gas for 24 percent, and oil—more valuable
as a transportation fuel—for just under 1 percent (see Fig. 1). Domestic coal is abundant and will be for decades to centuries. Gas supplies are more limited, but should be
ample for several decades. In contrast to oil, where imports account for two-thirds of
the U.S. supply, only about 12 percent of U.S. gas consumption is imported. So, from a
resource standpoint, fossil fuels could provide a viable substitute for nuclear power.
Although some are smaller, most commercial nuclear reactors have power outputs in the 1-GWe range, and oen two or more reactors at a site make a power complex with output of several GWe. e largest coal-burning plants in the United States
have comparable outputs, oen with pairs of generating units each in the 1-GWe range.
In contrast, natural gas plants have typical outputs measured in hundreds of MWe. So it
would take several natural gas generating stations to replace a nuclear plant. But natu-
121 Federation of American Scientists
Alternatives to Nuclear Power
ral gas plants are relatively easy and quick to construct—making natural gas probably
the easiest source to substitute directly for nuclear power.
Fossil fueled power is environmentally problematic. Despite substantial improvements resulting from the Clean Air Act and its amendments, coal plants continue
to pollute the atmosphere, land, and water with toxic substances ranging from sulfur to
mercury to particulates. Combustion of gas is cleaner, making gas preferable from a
pollution standpoint. But the big environmental issue with any fossil fuel is its climatechanging carbon emissions. On this score, gas is also preferable: it produces about half
the carbon dioxide of coal for the same electrical energy production. And gas plants
using combined-cycle technology—incorporating jet-engine-like gas turbines whose
waste heat powers a conventional steam turbine—can achieve efficiencies of 60 percent,
well above the best coal or nuclear plants. No energy source is completely free from
emissions of carbon dioxide or equivalent climate-changing greenhouse gases, and nuclear is no exception. Worst-case estimates for nuclear power involve greenhouse emissions per unit of electrical energy that are 20 percent of those from the best natural gas
plants—and the nuclear emissions could be a lot lower. Substituting natural gas for
nuclear power would, therefore, result in substantially increased greenhouse emissions.
But its availability, economics, ease of implementation, and environmental advantages
over coal make natural gas the obvious choice as a fossil substitute for nuclear power.
Renewable Energy
Renewable sources of electrical energy include hydro, geothermal, biomass, wind, and
solar—although depending on how they’re harvested or extracted, biomass and geothermal may or may not be sustainably renewable. Whether and how well these renewables might substitute for nuclear power depends on the total power capability of each
renewable source, on temporal and geographical availability, on economic factors, and
on the state of each technology and its manufacturing base.
Hydroelectric power
Figure 1 shows that 7 percent of U.S. electricity currently comes from hydropower,
which captures the energy of flowing or falling water. Hydropower was one of the earliest sources of mechanical energy for industry, and became prominent in the generation of electricity in the early 20th century. But in the United States, hydropower is unlikely to substitute for nuclear because our hydroelectric potential is almost entirely
Federation of American Scientists 122
The Future of Nuclear Power in the United States
February 2012
exploited. Imported hydropower is a different story, as ongoing hydro development in
northern Quebec makes power available for sale to the northeastern United States. e
state of Vermont, which gets one-third of its electrical energy from a nuclear plant that
may soon be shut down, is considering replacing its nuclear electricity with purchases of
Canadian hydroelectricity. But replacing all 20 percent of U.S. nuclear-generated electricity with hydropower is probably unrealistic.
Geothermal power
Where vapor-saturated hot rocks lie close to Earth’s surface, geothermal steam can be
used directly to drive a turbine-generator to make electricity. More complex power
plants can extract energy from lower-quality geothermal resources, at greater expense
but with less environmental impact. In principle, holes several miles deep bored into
hot, dry rock anywhere on Earth could yield geothermal steam from water pumped
down the holes, but that technology is hardly developed. A deep-hot-rock project in
Switzerland was halted in 2006 aer it triggered an earthquake. Today about 0.4
percent of U.S. electrical energy is from geothermal sources, a figure that is not likely to
rise substantially. erefore geothermal energy is unlikely to make a significant dent in
the 20 percent of U.S. electricity generated from nuclear fission.
About a third of the “other” category in Fig. 1 represents electricity generated from
burning biomass—ultimately stored solar energy captured by photosynthetic plants.
About two-thirds of biomass electricity comes from wood, usually as chips, burned to
power conventional steam-cycle turbine-generators. e remainder is from combustion
of municipal waste, landfill gases, and other fuels that are ultimately biological in origin. Although liquid biofuels may play increasing roles in transportation, supply limitations and logistical constraints probably rule out much use of biomass for large-scale electric power generation. The largest wood-fired power plants have outputs in the range of
50 MWe, far below the GWe-range outputs of nuclear plants. Local situations excepted,
biomass is unlikely to provide a wholesale substitution for nuclear power.
Some 1 percent of solar energy incident on Earth goes into driving winds—energy that
we can tap directly with wind turbines. A 2010 report by the National Renewable Energy
Laboratory estimates that the United States’ land-based wind resource could support
123 Federation of American Scientists
Alternatives to Nuclear Power
wind installations totaling some 10 TW of peak power. Even accounting for wind’s
relativity low capacity factor—about 36 percent for new U.S wind installations—that
still translates into a potential for about 33,000 TWh of annual electrical energy production. Comparison with nuclear power’s annual production of 800 TWh shows
clearly that wind has the potential to substitute for nuclear power.
at substitution, however, would not be “plug and play.” Nuclear plants can
be sited most anywhere there’s cooling water available—usually close, but not too close,
to the population centers they serve. But the United States’ greatest wind resources are
in the Great Plains, a region of low population density. Supplying the U.S. population
with wind-generated electricity would, therefore, require substantial upgrades to the
long-distance transmission capabilities of the North American electrical grid. And the
intermittent nature of wind makes it difficult to manage electrical distribution when
wind-generated energy is added to the mix.is intermittency is not a significant problem when wind’s percentage of the mix is small, but as it rises above about 20 percent,
problems of managing wind energy become more serious. One solution is to link geographically distant wind sites on the assumption that there will always be wind energy
available somewhere. That, too, requires substantial upgrades to the electrical grid.
Today’s wind farms have peak power outputs in the hundreds of megawatts;
coupled with the 36 percent capacity factor, that makes them considerably smaller than
typical nuclear plants. So something like ten 300-MW wind farms would be required
to replace a 1-GW nuclear plant—and they would occupy some ten times the land area
of the nuclear plant and the mines and processing facilities that supply its fuel. But
unlike land dedicated to nuclear energy, land occupied by wind farms remains suitable
for ranching and other low-density agricultural uses.
Overall, the issues associated with wind-generated electricity are not trivial,
but neither are they insurmountable. Furthermore, wind capacity is growing at more
than 30 percent annually, and wind is approaching cost-competitiveness with fossil fuels and is arguably less expensive than new nuclear installations. A decade ago, wind’s
contribution to the U.S. electricity mix was negligible; today, wind produces some 2
percent of U.S. electricity. Wind clearly offers a realistic if not fully grid compatible
substitute for nuclear power.
Federation of American Scientists 124
The Future of Nuclear Power in the United States
February 2012
Conversion of solar energy into electricity occurs either with solar-thermal systems or
through photovoltaic (PV) panels. Like their fossil and nuclear counterparts, solarthermal systems use heat energy to drive turbine-generators. Semiconductor-based PV
panels convert sunlight energy directly into electricity, with no moving parts; typical
efficiencies of commercial panels are 10 – 20 percent. e largest solar-thermal installations currently operating in the United States are rated at just under 100 MWe, while
the largest U.S. PV installation is 14 MWe. Worldwide, photovoltaic capacity is growing at nearly 40 percent per year, and PV “farms” are approaching 100 MW in peak
output. ere’s no question that the solar resource is adequate: solar energy reaches
Earth’s surface at the rate of about 100,000 TW, nearly 10,000 times the rate at which
humankind uses energy.
Many of the same considerations apply to solar as to wind. ere are issues of
intermittency and remoteness of optimal solar locations. Solar, unlike wind, is not currently economically competitive with conventional power generation—although costs
are dropping rapidly. Figure 1 reflects this economic disparity, with solar producing a
mere 0.02 percent of U.S. electrical energy. On the other hand, solar lends itself better
than wind to so-called distributed generation—for example, individual rooop systems—offering an alternative to a power system based dependent on relatively few
gigawatt-scale power plants. Some solar-thermal plants reduce intermittency through
thermal-energy storage, which can increase capacity factors from 25 percent to around
70 percent. And solar-thermal installations, because they use conventional steam turbines to drive electric generators, can be coupled with non-solar heat sources, usually
natural gas. Such hybrid solar-gas plants offer steady power outputs with reduced fossil
fuel consumption.
Nuclear Alternatives
If the goal in replacing today’s nuclear power plants is to get beyond light-water reactors first developed in the 1950s, then advanced nuclear technologies should be considered as substitutes. Most of the 100+ reactors operating today in the U.S. are nearing
the ends of their original design lifetimes, and while most are receiving 20-year license
extensions from the NRC, they will need replacing if nuclear power is to hold its own
in the U.S. electricity mix. Advanced fission reactors offer advantages in safety, in prolonging the uranium supply, and even in “burning” the waste from current-generation
reactors. Development of fusion reactors would make every gallon of seawater the en-
125 Federation of American Scientists
Alternatives to Nuclear Power
ergy equivalent of 300 gallons of gasoline—providing humankind with energy for billions of years longer than the sun will continue to shine. But none of these advanced
nuclear alternatives is anywhere near technological or economic feasibility, and it will
be decades before any become viable substitutes for today’s light-water fission reactors.
So what’s the best substitute for nuclear power? Again, that depends on your criteria
and timeframe. Fossil-fueled power from coal or gas offers proven technology, relatively
low cost, and fuel availability for decades or longer. Gas plants are quick to construct
and could readily replace aging nuclear plants. But in a world experiencing rapid climate change, fossil-generated energy is a step backward from necessary reductions in
greenhouse-gas emissions. e one large-scale, mature-technology renewable replacement for nuclear power is hydroelectricity—but in the United States there’s little potential for growth in hydro. Geothermal energy is limited to a few geographical regions
and cannot make major inroads into that 800-TWh per year of nuclear electricity.
Biomass, while potentially significant for transportation, is unlikely to see greatly increased use for generation of electricity. Renewable energy from wind and the Sun is
abundant and has minimal—but not zero—environmental impact. Wind is becoming
competitive with conventional energy sources, and growth in the wind industry has
brought wind to a 2-percent share of U.S. electricity generation, a figure that is rising
rapidly. Solar-thermal and photovoltaic technologies are farther behind economically,
but their advantages are similar to those of wind. Both wind and solar challenge the
power grid with their intermittent generation, and increased use of these renewable
energy technologies would require an enhanced and smarter electric grid. Finally, advanced nuclear technologies could replace today’s fission reactors while essentially solving the nuclear waste problem. But they’re decades away.
Comparing nuclear power and its potential replacements depends not only on
technological and immediate economic issues, but also on policy decisions that could
alter the future balance on these issues. Tax incentives for wind and solar encourage
nascent industries, lowering costs and encouraging widespread dissemination. Funding
to develop carbon capture and sequestration could make coal more attractive from an
environmental standpoint. Advanced reactor designs, and especially fusion, require
research collaborations and funding at the international level. Finally, accounting for
negative externalities—especially greenhouse emissions—in the cost of fossil energy
Federation of American Scientists 126
The Future of Nuclear Power in the United States
February 2012
would make renewable alternatives to nuclear power more competitive.
If the United States’ nuclear enterprise is to be replaced, that will most likely
be done with a combination of sources, beginning with fossil fuels and, over time, substituting renewable sources. In the end, there’s no one right choice to replace nuclear power. Geography, economics, resource availability, and related factors may dictate a mix of nuclear
substitutes that vary throughout the country. But before plunging into nuclear replacement,
there’s another question to be asked, one that won’t be answered here: Why replace nuclear
power? There are plenty of good reasons, mostly involving negative aspects of nuclear
power, but they need to be weighed against the negatives associated with substitute energy
sources and with the fact of an established industry that economically generates 20 percent
of the United States’ electricity—and that does so more safely and with less environmental
impact than its most obvious replacement, namely fossil-fueled energy.
127 Federation of American Scientists
Emerging Nuclear Technologies
Chapter 10
Emerging Nuclear Technologies
by Daniel Ingersoll
Commercial nuclear power presently addresses only one of the United States’
energy needs: centralized base-load electricity generation. It is possible to extend the
use of nuclear energy to other energy demands, such as distributed electricity
generation and industrial process heat applications; however, these applications may be
better served by a different plant design and underlying technology than the existing
fleet of water-cooled reactors. Even for electricity production, advanced technologies
and innovative reactor designs can significantly improve the affordability and utility of
nuclear energy for meeting U.S. energy needs. is realization has given rise to a
vigorous advanced reactor development effort in the United States and internationally.
e terminology of reactor “generations” was introduced in the late 1990s in part to
clarify the roles and opportunities for developing advanced nuclear technologies.
Generation I designs represent the early prototypes that were designed and built to gain
familiarity with the technology, and provided the basis for the existing fleet of
commercial light-water reactor (LWR) plants, which are categorized as Generation II.
During the U.S. hiatus of new plant construction in the past three decades, several
Generation III plant designs were developed that incorporated lessons learned from the
previous generation of plants, especially regarding design simplification,
standardization, and increased use of passive safety features. But since U.S. utilities
continued to purchase other energy sources, principally naturally gas, to meet new
capacity requirements, no Generation III plants were ordered and constructed. With a
resurgence in the interest in building new nuclear capacity during the past several years,
updated versions of some of the Generation III designs, referred to as Generation III+
designs, were developed and are now in the process of being fully licensed and ready for
Federation of American Scientists 128
The Future of Nuclear Power in the United States
February 2012
construction. An example of a Generation III+ design is the Westinghouse Advanced
Passive (AP-1000) design. e first four AP-1000 plants are under construction in
China and another four units have been ordered in the United States.
Generation IV Technology
About the same time as the development of the Generation III+ designs for immediate
deployment, several Generation IV concepts were initiated by the research and
development (R&D) community. e Generation IV concepts represent a broad set of
advanced reactor concepts that are intended to dramatically improve the performance
of previous generations, especially with regard to safety, economics, sustainability, and
proliferation resistance. Moreover, they are intended to enable the extension of nuclear
energy to more than base-load electricity production. e Generation IV program was
initiated in the United States in 2000 and was quickly embraced by a consortium of
countries—initially seven and now thirteen countries, including: Argentina, Brazil,
Canada, China, Euratom, France, Japan, Russia, the Republic of Korea, South Africa,
Switzerland, the United Kingdom, and the United States.121 Each country funds its
own participation and actively collaborates with the United States in the development
of advanced reactor concepts and associated technologies.
e Generation IV program began with an extensive review, evaluation and
selection of six specific reactor concepts:
• Very High-Temperature Reactor (VHTR),
• Super-Critical Water-cooled Reactor (SCWR),
• Molten Salt Reactor (MSR),
• Sodium-cooled Fast Reactor (SFR),
• Lead-cooled Fast Reactor (LFR), and
• Gas-cooled Fast Reactor (GFR).
ese six reactor concepts, which actually represent classes of reactor systems
rather than specific designs, can be further grouped into two fundamental categories:
(1) high-temperature reactors principally for process heat applications, and (2)
fast-spectrum reactors principally for fuel cycle applications. While all six reactor
concepts can be used to produce electricity, some with significantly higher conversion
121 “A Technology Roadmap for Generation IV Nuclear Systems,” GIF-002-00, December 2002.
129 Federation of American Scientists
Emerging Nuclear Technologies
efficiencies than traditional light-water reactors, their real strength is in addressing
non-electrical energy demands. While the United States initially studied the viability of
all six Generation IV concepts, it became evident that two systems, the VHTR and the
SFR, were the most mature in their class and were selected for more in-depth
High-Temperature Reactors
e VHTR is largely intended to extend nuclear energy into non-electrical
applications. Electricity generation represents roughly 40 percent of U.S. total energy
consumption, and similarly about 40 percent of U.S. greenhouse gas (GHG)
emissions.122 To significantly reduce our total GHG emissions will require at least a
partial replacement of fossil fuels used for industrial process heat and for transportation
fuels. Reducing fossil fuels for transportation can be accommodated through increased
use of plug-in electric vehicles, but may also require a move to cleaner liquid fuels such
as hydrogen. e de-carbonization of industrial energy consumption will require
adaptation of nuclear energy to process heat applications, many of which require
process temperatures well above what can be delivered from conventional LWRs, which
are limited by the low boiling temperature of water. While conventional water reactors
can provide output temperatures of 300-350°C, many industrial processes require
temperatures >500°C and even as high as 1000°C. As shown in Fig. 1, examples of
high-temperature processes include steam reforming of natural gas, coal gasification,
and thermo-chemical production of hydrogen. To achieve the very high temperatures,
reactor designers must shi to another coolant technology such as gas, super-critical
water, or molten salt, and must also switch to different materials of construction that
can survive sustained operation at elevated temperatures.
122 “Inventory of U.S. Greenhouse Gases Emissions and Sinks: 1990-2007,” EPA-430-R-09-004, April 15,
Federation of American Scientists 130
The Future of Nuclear Power in the United States
February 2012
Fig. 1. Temperature requirements for various industrial processes123
Helium is a natural choice as a coolant for high-temperature reactors because
it is inert, remains in single phase throughout the postulated temperature range of
operation, and has accumulated considerable international experience for nuclear
systems. e United States built two helium-cooled prototype reactors: Peach Bottom
1 and Fort St. Vrain. Typically, graphite provides the neutron moderation function in
helium-cooled reactors and also provides good thermal inertia, which is important for
slowing down the response of the reactor to power transients. Multiple VHTR designs
are being developed, including designs that use fuel in stationary “compacts” that are
replaced on a regular refueling cycle, and designs that use fuel within “pebbles” that
migrate through the system during operation and are removed or reinserted into the
reactor on a continuous basis. Although gas-cooled reactors can be scaled to high
power output, all contemporary designs are constrained to small output (typically
150-300 MWe) to ensure passive removal of the core decay heat. Figure 2 shows a
123 “Next Generation Nuclear Plant Research and Development Program Plan,” Idaho National
Laboratory, January 2005.
131 Federation of American Scientists
Emerging Nuclear Technologies
model of the Modular Helium Reactor (MHR) proposed by General Atomics for the
Next Generation Nuclear Plant (NGNP) project.124
e R&D needed to realize the benefits of the VHTR include: development
and qualification of coated particle fuel, construction materials that can withstand
high-temperature, high-temperature/high-pressure heat exchangers, high-temperature
sensors, and reactor safety analysis methods for gas-cooled systems. Additionally,
alternative coolants are being explored such as liquid fluoride salt, which is largely
based on MSR experience gained in the 1960s and 1970s. Although a comprehensive
R&D effort was initiated within the Generation IV program, the R&D relevant to
helium-cooled reactors is currently funded within the DOE NGNP project, in
addition to a cost-shared public/private partnership to construct a commercial scale
VHTR. e R&D relevant to fluoride-salt-cooled reactors, which is an advanced
option for high-temperature systems, has been resumed recently within the DOE
Advanced Reactors Concepts (ARC) program.
124 “Next Generation Nuclear Plant Conceptual Design Report, INL/EXT-07-12967, Revision 1,
November 2007.
Federation of American Scientists 132
The Future of Nuclear Power in the United States
February 2012
Fig. 2. Model of MHR design for high-temperature process heat applications.
Fast-Spectrum Reactors
e second class of Generation IV reactors addresses the need to manage both nuclear
fuel resources and nuclear waste products, which will become increasingly important as
the use of nuclear energy is expanded. ese functions are best accomplished using
fast-spectrum reactor designs because fast-spectrum reactors have the benefit of
producing more neutrons beyond what is needed to sustain the fission reaction than do
thermal-spectrum reactors. ese excess neutrons can be used for a variety of purposes,
such as producing new fuel from fertile material; in fact they can produce more fuel
than they burn, or they can consume long-lived waste products from used nuclear fuel
133 Federation of American Scientists
Emerging Nuclear Technologies
discharged from other reactors. In order to achieve the excess neutrons, fast-spectrum
reactors must not contain low atomic mass materials such as water or graphite that
moderate the neutrons to lower energy. Typical coolants used in fast-spectrum reactors
include molten sodium, lead or lead-bismuth, and gas. All of these coolants can operate
at temperatures higher than water-cooled reactors, typically in the range of 500-700°C.
Sodium, lead and gas-cooled fast spectrum reactors were studied in the United
States during the early part of the Generation IV program. Of these, sodium appears to
be the most promising, due in part to the more extensive international experience with
sodium-cooled reactors. e United States has built and operated three sodium-cooled
fast spectrum reactors: the Experimental Breeder Reactors I and II and the Fast Flux
Test Facility. Currently, France, Russia, Japan, China, and India have or are pursuing
sodium-cooled reactor technology. A challenge for SFRs is the energetic reaction that
sodium metal has with water, which requires extra precautions, and hence expense, to
avoid sodium-water interactions. is is especially challenging for systems that use the
steam Rankine cycle for power conversion, which necessarily requires water in the
secondary system. Figure 3 shows a model of a sodium-cooled reactor design being
promoted by General Electric-Hitachi (GE-H), designated the Power Reactor
Innovative Small Module (PRISM).125
e promising development of closed supercritical CO2 Brayton cycle power
conversion systems (a high efficiency thermodynamic cycle used in gas turbines) will
help to significantly increase the power conversion efficiency and reduce the
sodium-water interaction challenge in SFRs. Examples of other R&D needs for
successful commercialization of SFRs include material and fuel development and
qualification, and under-sodium viewing technology for inspection and maintenance.
125 C. E. Boardman, et al, “Optimizing the Size of the Super-PRISM Reactor,” Proceedings of the
International Conference on Nuclear Energy (ICONE) 8, Baltimore, MD, 2000.
Federation of American Scientists 134
The Future of Nuclear Power in the United States
February 2012
Fig. 3. Model of GE-H PRISM reactor (le) and plant (right).
Small Modular Reactors
Very recently, a third class of advanced reactors has emerged that straddles between the
Generation III+ designs and the Generation IV concepts. is is the class of small
modular reactors (SMR), which is not a specific reactor technology but rather a special
implementation of a technology. SMRs add affordability and flexibility to the other
benefits of Generation III and IV systems by reducing the total capital cost of a nuclear
plant and enabling more agile plant designs. Although SMR designs for commercial
nuclear power first emerged in the late 1970s, no designs have been certified by the
NRC, and it has been only in the past few years that vendors and utilities have begun
pursuing them actively. Currently, SMRs based on water, helium, sodium, lead, and
fluoride salt coolant technologies are being developed by either the commercial
industry or by the R&D community. e distinguishing characteristics of SMRs are:
their small power output (less than 300 MWe compared to 1100-1600 MWe for the
Generation III+ designs), substantial fabrication within a factory environment and
then transported and installed on-site, and operation typically in parallel with
135 Federation of American Scientists
Emerging Nuclear Technologies
additional reactor modules to constitute a single power plant.
is “bite-size” implementation of nuclear power offers the owner a number of
significant business advantages:
• e power plant can be sized according to the capacity needed by the
customer and potentially expanded later if additional capacity is needed.
• e multiplicity of modules improves power availability since for a
multi-module plant, maintenance and refueling can be performed on a single
module while the other modules continue to operate.
• e lower upfront capital investment and incremental build-out reduces the
owner’s debt profile and improves financing options because the early units can
generate revenue to help finance the construction of later units.
Additionally, a number of technical advantages include:
• Factory fabrication improves quality control, standardization, and schedule
reliability while reducing manufacturing cost.
• Individual modules are simpler and more robust due to extensive use of passive
safety features.
• Below-grade siting of the smaller reactor enhances safety and security.
e first SMRs likely to be ready for deployment are those based on familiar
water-cooled reactor technology. Multiple designs are well underway by both
traditional reactor vendors and new upstart companies. One of the most modular
SMRs is the NuScale design, which has a power of 45 MWe and a reference plant
design consisting of 12 modules (shown in Fig. 4).126
A major challenge for the deployment of SMR designs is the licensing process,
which must account for several design, operational, and financial differences relative to
large plant designs. Many of the regulatory issues have been identified by the NRC and
are being addressed by the NRC, SMR vendors and other stakeholder groups such as
the Nuclear Energy Institute and the American Nuclear Society.127 Also, the new
126 P. Lorenzini, “NuScale Power: Capturing the ‘Economies of Small’,” presentation at the International
Conference on Advances in Nuclear Power Plants 2010, San Diego, CA, June 13-17, 2010.
127 “Potential Policy, Licensing, and Key Technical Issues for Small Modular Reactor Designs,” Nuclear
Regulatory Commission, SECY-10-0034, March 28, 2010.
Federation of American Scientists 136
The Future of Nuclear Power in the United States
February 2012
economic model of “economies of small” to replace the more familiar economies of
scale are yet unproven. For non-LWR-based SMR designs, the R&D needs are similar
to those mentioned above for high-temperature and fast-spectrum reactor designs, for
example: the development and qualification of new materials and fuels, advanced
instrumentation, and innovative components.
Fig. 4. Model of NuScale reactor and containment vessel (le) and plant (right).
e unsurpassed safety and performance record of the current U.S. fleet of nuclear power
plants have ensured that large LWRs will continue to be an important cornerstone of the
U.S. clean energy portfolio. However, the challenging energy goals of the United States
support the need for future generations of reactor designs that can further expand the use
of nuclear energy for electricity production and extend the benefits of nuclear energy to
non-electrical energy-intensive applications. The plant designs and underlying
technologies will need to be different from the current fleet of LWRs, and this will
require a commensurate amount of technology R&D and ultimate demonstration of the
137 Federation of American Scientists
Emerging Nuclear Technologies
new technologies and business models for their successful deployment.
While industry is very engaged in the development and eventual deployment of
advanced reactor systems, the barriers to deployment are significant and will require
government support to overcome both technical and institutional challenges. These
challenges range from relatively modest licensing issues to decade-long fuel and materials
research and development needs. For the range of advanced systems discussed here, the
LWR-based SMR designs are likely to succeed in deployment first due to several factors,
including: (1) minimal technical issues needing resolution, (2) proven safety and
operational characteristics of water-cooled systems resulting in few licensing issues to
industry and NRC, and (3) a rapidly growing customer base of utilities considering SMRs
as affordable, incremental capacity or as carbon-free replacements for older, smaller fossil
plants. Even for these designs, however, government support and resources are needed to
facilitate demonstration of the new engineering, regulatory and business models for
first-mover SMRs. An aggressive public/private partnership to deploy the new designs
could result in the first commercial plants being ready to operate by 2018-2020.
High-temperature and fast-spectrum reactors, while offering important new
functionalities, will likely take longer to achieve commercialization and will require more
extensive government support for research and demonstration. In addition to addressing a
more extensive list of technical and regulatory challenges, introducing high-temperature
reactors to the process heat market requires that the customer base become more familiar
and comfortable with the technology. Good progress has been made with the NGNP
project, which includes a substantial customer engagement, but the process is relatively
new compared to the level of experience of utility customers with water-cooled reactors.
Similarly, fast-spectrum reactors face a comparable number of technical and regulatory
challenges and an uncertain customer base. Deployment of these advanced systems may
be possible by 2025-2030 with appropriate government support and demonstration.
Meaningful collaborations with other countries that are also seeking to develop these
same technologies will be very important to integrate knowledge and minimize
development costs.
In summary, emerging nuclear technologies such as high-temperature,
fast-spectrum, and small modular reactors have the ability to offer clean, affordable and
abundant energy for the United States and should become key components of the future
energy portfolio.
Federation of American Scientists 138
The Future of Nuclear Power in the United States
February 2012
John F. Ahearne
Dr. Ahearne is the Director of the Ethics Program at Sigma Xi, The Scientific Research
Society, a Lecturer in public policy at Duke University, and an Adjunct Scholar at Resources for the Future. His professional interests are reactor safety, energy issues, resource
allocation, and public policy management. He received his B.S. and M.S. degrees from
Cornell University and a Ph.D. in physics from Princeton University. He served in the
U.S. Air Force from 1959 to 1970, resigning as major. He has also served as Deputy and
Principal Deputy Assistant Secretary of Defense (1972-1977), in the White House Energy Office (1977), as Deputy Assistant Secretary of Energy (1977-1978), and as Commissioner and Chairman of the U.S. Nuclear Regulatory Commission (chairman,
1979-1981). He currently serves on the Department of Energy’s Nuclear Energy Research Advisory Committee. He is a fellow of the American Physical Society, the Society
for Risk Analysis, the American Association for the Advancement of Science, and the
American Academy of Arts and Sciences; and a member of the National Academy of Engineering, Sigma Xi, and the American Nuclear Society.
Albert V. Carr, Jr.
Mr. Carr has spent his career in the field of energy law, beginning in 1971 as a Trial Attorney in the Office of General Counsel of the Federal Power Commission, the predecessor agency to the Federal Energy Regulatory Commission. He then took a position as a
Trial Attorney in the regulatory division of the Office of General Counsel of the U. S.
Atomic Energy Commission, the predecessor agency to the U. S. Nuclear Regulatory
Commission. In 1976, he left government service and entered private practice in Washington D. C. In 1981, he joined the Legal Department of Duke Power Company in
Charlotte, NC. While at Duke he represented Duke Power and its subsidiaries in a variety of matters related to Federal regulation of the electric utility industry. From 1981 to
1993, he was responsible for, among other things, regulatory matters for Duke’s seven
nuclear units, including contested licensing proceedings for Duke’s Catawba Nuclear Station. In 1998, he retired from full-time employment as Deputy General Counsel, Duke
Energy Corporation and continued for several years to represent Duke and other clients
in energy regulatory and legislative matters. He has recently participated in several national studies and analyses of, and given presentations on, the potential revival of the domestic nuclear power industry. In the fall of 2000, as an Adjunct Professor of Law at the
Washington & Lee University School of Law, he began teaching a course in Federal Energy Regulation. Most recently, he has become Of Counsel with the international law
firm Duane Morris LLP, working with the firm’s Energy and Construction departments
139 Federation of American Scientists
in developing a nuclear licensing practice. Mr. Carr received his BA in English from the
Virginia Military Institute in 1966 and his JD from Washington & Lee in 1971. A veteran of the United States Marine Corps, he is a member of the Virginia, North Carolina
and District of Columbia Bars.
Harold A. Feiveson
Dr. Feiveson is a Senior Research Scientist and member of Princeton’s Program on Science and Global Security of the Woodrow Wilson School of Public and International
Affairs. He had co-directed the Program from its inception in 1974 until 2007. His
principal research interests are in the fields of nuclear weapons and nuclear energy policy.
He was editor and a principal author of the book, The Nuclear Turning Point: A Blueprint for Deep Cuts and De-alerting of Nuclear Weapons (Brookings 1999). He has taught
regularly in the Woodrow Wilson School for the past 37 years on a range of topics, including on the environment, energy, and nuclear arms control. Before coming to Princeton, he was a member of the Science Bureau of the U.S. Arms Control and Disarmament
Agency from 1963 to 1967. He was one of the founders of the international journal,
Science & Global Security, and editor during the first twenty-one years of the journal, until
2010. He received an M.S. in theoretical physics from the University of California Los
Angeles in 1959 and a Ph.D. in public affairs from Princeton University in 1972.
Daniel Ingersoll
Dr. Daniel Ingersoll is a Senior Program Manager for the Nuclear Technology Programs
Office at Oak Ridge National Laboratory. He recently served as Campaign Director for
the Grid-Appropriate Reactors program within the Global Nuclear Energy Partnership
has participated in several advanced reactor programs, including the Advanced High
Temperature Reactor project, the International Reactor Innovative and Secure project,
and the Space Reactor Technology Program. During his 32 years at ORNL, he led various ORNL research groups and sections dedicated to radiation transport modeling and
reactor physics analysis. He received a B.S. degree in physics from Miami University in
1973 and a Ph.D. in nuclear engineering from the University of Illinois in 1977. He is a
Fellow of the American Nuclear Society and Past Chairman of the ANS Radiation Protection and Shielding Division.
Andrew C. Klein
Dr. Klein is Professor of Nuclear Engineering and Radiation Health Physics at Oregon
State University. Dr. Klein has served as Director of Educational Partnerships at the
Federation of American Scientists 140
The Future of Nuclear Power in the United States
February 2012
Idaho National Laboratory in Idaho Falls, ID, as Head of the Department of Nuclear
Engineering and Radiation Health Physics at Oregon State University, and as Director of
the Oregon State University Radiation Center. Professor Klein received his B.S. in Nuclear Engineering from Pennsylvania State University in 1977. He completed his M.S. in
Nuclear Engineering and his Ph.D., also in Nuclear Engineering, from the University of
Wisconsin, Madison in 1979 and 1983 respectively. His research interests are focused on
nuclear reactor systems design and analysis including applications in space and fusion
energy and nuclear nonproliferation technology. He has been an author on more than 90
technical publications. He is currently a member of the Institute for Nuclear Power Operation’s National Nuclear Accrediting Board, the Associate Editor for North America for
Nuclear Technology and an Advisory Editor for the Annals of Nuclear Energy and was a
member of the U.S. Department of Energy's Nuclear Energy Research Advisory Committee, NASA’s Space Science Advisory Committee, and ABET Inc.’s Engineering Accreditation Commission.
Stephen Maloney
Mr. Maloney is a Partner at Azuolas Risk Advisors. A longtime energy risk analyst in oil,
natural gas, LNG and electric power, he also served as a nuclear risk analyst and
participated in several accident rulemakings, including station blackout, fire protection/
safe shutdown and other severe accidents. He has also consulted on several nuclear power
plant M&A transactions, and investment decisions. He was a former commissioned
officer in the US Navy's nuclear propulsion program and qualified as Chief Engineer for
the C1W reactor. He has an MS Operations Research, BS Physics, and BS Mathematics.
He was licensed as a professional engineer and is a member of the Global Association of
Risk Professionals, the American Mathematical Society and INFORMS.
Ivan Oelrich
An independent defense analyst, Dr. Oelrich was the former Vice President of the Strategic Security Program at the Federation of American Scientists. Previously, he worked at
the Institute for Defense Analyses, Harvard University’s Kennedy School of Government, the Office of Technology Assessment, and the Defense reat Reduction
Agency. He has taught as an adjunct professor at Georgetown University and the Johns
Hopkins University. He received his B.S. from the University of Chicago and a Ph.D.
from Princeton University, both in chemistry. He had a pre-doctoral fellowship at the
Lawrence Livermore National Laboratory and later conducted research and taught in the
Physics Department of the Technical University of Munich in Germany.
141 Federation of American Scientists
Sharon Squassoni
Ms. Squassoni serves as Director and Senior Fellow of the Proliferation Prevention Program at CSIS. Prior to joining CSIS, she was a Senior Associate in the Nuclear Nonproliferation Program at the Carnegie Endowment for International Peace. From 2002-2007,
she advised Congress as a Senior Specialist in weapons of mass destruction at the Congressional Research Service, Library of Congress. Before joining CRS, she worked briefly
as a reporter in the Washington bureau of Newsweek magazine. She also served in the
executive branch of government from 1992 to 2001. Her last position was Director of
Policy Coordination for the Nonproliferation Bureau at the State Department. She also
served as a policy planner for the Political-Military Bureau at State. She began her career
in the government as a nuclear safeguards expert in the Arms Control and Disarmament
Agency. She is the recipient of various service awards and has published widely. She received her B.A. in political science from the State University of New York at Albany, a
Masters in Public Management from the University of Maryland, and a Masters in National Security Strategy from the National War College.
Richard Wolfson
Dr. Wolfson is the Benjamin F. Wissler Professor of Physics at Middlebury College, where
he also teaches Climate Change in Middlebury's Environmental Studies Program. He
completed his undergraduate work at MIT and Swarthmore College, graduating from
Swarthmore with a double major in Physics and Philosophy. He holds a master's degree in
Environmental Studies from the University of Michigan and a Ph.D. in Physics from
Dartmouth. Professor Wolfson's published work encompasses diverse fields such as medical physics, plasma physics, solar energy engineering, electronic circuit design, observational astronomy, theoretical astrophysics, nuclear issues, and climate change. His current
research involves the eruptive behavior of the sun's outer atmosphere, or corona, as well as
terrestrial climate change and the sun–Earth connection. He is the author of several
books, including the college textbooks Physics for Scientists and Engineers, Essential University Physics, and Energy, Environment, and Climate. He is also an interpreter of science
for the nonspecialist, a contributor to Scientific American, and author of the books Nuclear Choices: A Citizen's Guide to Nuclear Technology and Simply Einstein: Relativity Demystified.
Federation of American Scientists 142
The Future of Nuclear Power in the United States
February 2012
Charles D. Ferguson
Dr. Charles D. Ferguson is the President of the Federation of American Scientists (FAS). Prior to FAS, he worked as the Philip D. Reed Senior Fellow for Science and Technology
at the Council on Foreign Relations. Before his work at CFR, he was the Scientist-inResidence in the Monterey Institute’s Center for Nonproliferation Studies, where he cowrote (with William Potter) the book The Four Faces of Nuclear Terrorism (Routledge,
2005). While working at the Monterey Institute, he was the lead author of the report
Commercial Radioactive Sources: Surveying the Security Risks, which was the first indepth, post-9/11 study of the “dirty bomb” threat. This report won the 2003 Robert S.
Landauer Lecture Award from the Health Physics Society. Dr. Ferguson has consulted
with Sandia National Laboratories and the National Nuclear Security Administration on
improving the security of radioactive sources. He has worked as a physical scientist in the
Office of the Senior Coordinator for Nuclear Safety at the U.S. Department of State. He
has recently completed the book Nuclear Energy: What Everyone Needs to Know (Oxford
University Press, May 2011). He graduated with distinction from the United States Naval
Academy, served in the U.S. nuclear Navy, and earned a Ph.D. in physics from Boston
Frank A. Settle
Dr. Frank Settle received a BS from Emory and Henry College and a PhD in chemistry
from the University of Tennessee. Since retiring as professor of chemistry at VMI in
1992, he has been a program officer at the National Science Foundation, a consultant for
the Department of Energy, and is currently a visiting professor at Washington and Lee
University where he directs the Alsos Digital Library for Nuclear Issues. He has taught
courses on nuclear history, nuclear energy, and weapons of mass destruction since 1991. is report reflects the judgments and recommendations of the authors. It does not
necessarily represent the views of the Federation of American Scientists and
Washington and Lee University.
143 Federation of American Scientists
federation of american Scientists and
Washington and Lee University
Federation of American Scientists
1725 DeSales Street, NW, 6th Floor
Washington, DC 20036
TEL 202-546-3300 FAX 202-675-1010
Washington and Lee University
204 West Washington Street
Lexington, VA 24450-2116
TEL 540-458-8400
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF