A Look at the Promise and Problems of Nuclear Energy
Professor Burton Richter - Stanford University
PIME 2006 Conference, Vienna, 15 February 2006
Nuclear energy is undergoing a renaissance, driven by two very
loosely-coupled needs; first, to supply more energy to support
global economic growth, and second, to mitigate global warming
driven by the emission of greenhouse gases from fossil fuel.
With the current mix of fuels, growing the economy increases
emissions and increased emissions lead to climate change
that will eventually harm the economy. Nuclear energy offers
one way out of this cycle.
Many forecasts of energy demand in the 21st century
have been made and all give roughly the same answer. The International
Institute of Applied Systems Analysis, for example, shows in its
mid-growth scenario (figure 1) primary energy demand increasing
by a factor of two by mid-century and by nearly another factor
of two by the end of this century. By the year 2030 the developing
countries are projected to pass the industrialized ones in primary
energy use, and China will pass the United States as the largest
energy consumer. It is worth noting that economic growth in China
and India is already higher than assumed in the mid-growth scenario.
Fig. 1. IIASA Projection of Future Energy Demand
Today, about 80% of primary energy is derived
from fossil fuels. Supply constraints on two out of the three
fossil fuels are already evident. Oil prices have surged and now
are about $60 per barrel. Demand is rising at an average rate
of about 1.5 million barrels per day per year, requiring the output
of another Saudi Arabia every eight years to keep up with increased
While there is a lot of natural gas, there are
transport constraints. Natural gas prices also have risen and
now are at the unprecedented level of $9-$10 per million BTU.
The only fossil fuel in abundant supply is coal.
However, it has serious pollution problems and expensive technological
fixes are required to control environmental problems that have
large-scale economic consequences.
Concern about global warming is increasing and
even the United States government has finally said that there
is a problem. The Intergovernmental Panel on Climate Change (IPCC)
forecasts, in the business-as-usual case, an increase in atmospheric
carbon dioxide to 750 parts per million by the end of the century
with a consequent global temperature rise of 2? to 5? C, less
at the equator and more at the poles. We can surely adapt to this
increase if it is small and occurs smoothly. If, however, it is
large, and accompanied by instabilities in climate, economic and
societal disruptions will be very severe.
It is too late to prevent some global warming,
but limiting the effect requires a move away from carbon-based
fuels. The global-warming issue has caused some prominent environmentalists
to rethink their opposition to nuclear power. The question to
be confronted is which devil would they rather live with, global
warming or nuclear energy? James Lovelock, among others, has come
down on the side of nuclear energy.
There are many who believe that solar or wind
energy would be a better choice than nuclear. However, these are
not now ready for deployment on a large scale. They are costly,
but the real problem is that the sun does not shine nor does the
wind blow all the time. Until the energy storage problem is solved,
solar or wind energy will not make a major contribution to base-load
When economic interests and environmental interests
point in the same direction; things can begin to move in that
direction, in this case toward the deployment of large-scale carbon-free
energy. Nuclear energy is one such source. While it cannot be
the entire solution to the energy supply or climate change problems,
it can be an important part if the public can be assured that
it is safe, that nuclear waste can be disposed of safely, and
that the risk of weapons proliferation is not significantly increased
by a major expansion.
II. Nuclear Power Growth Potential
At present there are about 440 reactors worldwide
supplying 16% of world electricity (NEA Annual Report 2004). About
350 of these are in the OECD nations supplying 24% of their electricity.
The country with the largest share of nuclear electricity is France
at 78%. To an environmentalist, France should be looked at as
a model for the world. Its carbon-dioxide intensity (CO2 per unit
GDP) is the lowest in the world (figure 2). If the entire world
CO2 intensity were as low as France’s, CO2 emissions would
be reduced by a factor of two and global warming would be a much
easier problem to solve.
Fig. 2. CO2 Intensity
Projections for growth in nuclear power are uncertain
because of uncertain costs along with the three potential problems
mentioned earlier, safety, waste disposal, and proliferation risk.
The International Atomic Energy Agency (IAEA) projection (figure
3) of July 2004 for the year 2030 ranges from a high of 592 GWe
to a low of 423 GWe. This represents a net growth of between 16%
and 60% over the next 25 years. A recent MIT study (The Future
of Nuclear Power – an Interdisciplinary Study, July 2003)
projected as much as 1000 GWe by 2050 (an extrapolation of the
IAEA high projection for another 20 years), and an Electricite
de France projection is for about 1300 GWe (private communication).
The more aggressive growth numbers imply completions of about
two 1-GWe power plants per month for the next 45 years.
Fig. 3. Nuclear Power Projection to 2030
The cost of the new reactor being built in Finland
is about Euro 1800 per KWe. Costs will come down with series production
and locations more benign than northern Finland. Reactor manufacturers
claim that the cost of electricity from a new nuclear plant would
be comparable to that from a coal plant after first of a kind
engineering cost has been recovered and after coming down the
learning curve with five or so new plants. Even so, projections
like those above will represent the expenditure of 1-2 trillion
dollars on nuclear plants in the next 50 years. It is not clear
that we will have the trained personnel for the construction,
operation, or regulatory needs of a system that large, so education
and infrastructure are issues that need addressing too.
There’s little new to say on safety. Power
reactors of the Chernobyl type have never been used outside the
old Soviet bloc because of the potential for catastrophic accidents.
Even for reactors of that type, the accident would not have happened
had not the operators, for reasons we will never know, systematically
disabled all of the reactor’s safety systems.
The new generation of light-water reactors has
been designed to be simpler to operate and maintain than the old
generation, and with more passive safety systems.
With a strong regulation and inspection system,
the safety of nuclear systems can be assured. Without one, the
risks grow. No industry can be trusted to regulate itself when
the consequences of a failure extend beyond the bounds of damage
to that industry alone.
IV. Spent Fuel Treatment
In discussing the safe disposition of spent fuel,
I will set aside weapons proliferation concerns for now, and return
to them later. Looking separately at the three main elements of
spent fuel (figure 4), there is little problem with most of it.
The uranium which makes up the bulk of the spent fuel is not radioactive
enough to be of concern. It contains more U-235 than does natural
ore and so could be input for enrichment, or could even be put
back in the mines from which it came.
There is no scientific or engineering difficulty
with fission fragments, the next most abundant component. The
vast majority of them have to be stored for only a few hundred
years. Robust containment that would last the requisite time is
simple to build.
Fig. 4. Components of Spent Reactor Fuel
The problem comes mainly from the last 1% of
the spent fuel which is composed of plutonium and the minor actinides,
neptunium, americium and curium (collectively, the actinides).
For some of the components of this mix, the toxicities are high
and the lifetimes are long. There are two general ways to protect
the public from this material: isolation from the biosphere for
hundreds of thousands of years, or destruction by neutron bombardment.
Long term isolation is the principle behind the
“once through” system as advocated up to now by the
United States for weapons-proliferation-prevention reasons. In
a world with a greatly expanded nuclear power program I do not
believe the once-through system is workable. There are technical
limitations that would require a very large number of repositories,
and there is public doubt that the required extremely long isolation
times can be achieved.
The first technical problem comes from the heat
generated in the first 1500 or so years of storage which limits
the density of material that can be placed in a repository. Limitations
on the allowed temperature rise of the rock of a repository from
this source determine its capacity. The early heat generated from
fission fragments is not difficult to deal with. The decay of
plutonium-241 to americium-241 which then decays to neptunium-237
is the main source of heat during the first 1000 or so years.
The second technical problem is the very long-term
radiation. Here the same plutonium to americium to neptunium decay
chain generates the long-lived component that requires isolation
from the biosphere for hundreds of thousands of years.
For example, if nuclear energy in the United
States were to remain at the present 20% fraction of electricity
supply through the end of this century, the spent fuel in a once
through scenario would need nine repositories of the capacity
of the one proposed at Yucca Mountain. If the number of reactors
in the U.S. increases by mid-century to the 300 GWe projected
in the MIT study, a new Yucca Mountain would have to open every
six or seven years. This would be quite a challenge since we have
not been able to open the first one. In the world of expanded
use of nuclear power, the once-through cycle does not seem workable.
The alternative to once-through is a reprocessing
system that separates the major components, treating each appropriately
and doing something specific to treat the component that produces
the long-term problem. The most developed reprocessing system
is that of France and I will use it as a model. The French make
mixed oxide fuel, MOX, by separating out plutonium from spent
fuel and mixing it with an appropriate amount of uranium from
the same spent fuel. (The extra uranium from the spent fuel not
used for MOX goes to an enrichment facility.) The fission fragments
and minor actinides are embedded in glass (vitrification) for
eventual emplacement in a repository. The glass used appears to
have a lifetime of many hundreds of thousands of years in the
clay of the proposed French repository. The French Parliament
has held a series of hearings early this year and is expected
to soon issue its report on the acceptability of this system.
MOX fuel plus vitrification solves part of the
problem but not all of it. The next question is what to do with
the spent MOX fuel. The plan is to keep it unreprocessed until
fast-spectrum reactors are deployed commercially. These fast-spectrum
reactors burn a mix of plutonium and uranium-238 and can, in principle,
burn all of the minor actinides as well which is not possible
in the present generation of reactors. It is possible to create
a kind of continuous recycling program where the plutonium from
the spent MOX fuel is used to start the fast-spectrum system,
the spent fuel from the fast-spectrum system is reprocessed; all
the plutonium and minor actinides go back into new fuel, and so
forth. In principle, nothing but fission fragments goes to a repository
and these only need to be stored for a few hundred years. The
U.S. has just announced an aggressive R&D program called Global
Nuclear Energy Partnership (GNEP) aimed at destroying the actinides
in fast-spectrum burners (http://gnep.gov).
This sounds good in principle, but there’s
much work to do before putting it into standard, commercial practice.
Clearly a coherent international R&D program is the best way
to move ahead rapidly.
What we have now are two visions for the long-term
solution to the waste problem that are really not that difference
(figure 5). In the cycle of figure 5(a), MOX is burned in LWRs
and the residue is held for later treatment in a FR. In the cycle
of figure 5(b), all of the actinides in LWR spend fuel are separated
and treated in the FR.
Fig. 5(a). Transmutation Schematics with LWR
Fig. 5(b). Without LWR Recycle
In the long term, the two visions will merge
and become one. The current MOX fuel cycle can stabilize the world’s
Pu inventory until the fast systems come along to reduce it, and
to burn the minor actinides. The model of figure 5(a) will evolve
into that of figure 5(b) where the only materials that get to
a repository are fission fragments and the long-lived components
that leak into the fission fragment waste stream from inefficiencies
in the separation process. If that leakage can be kept to below
one percent, the required isolation time is of the order of 1000
years. This is less than the lifetime of the Egyptian pyramids
and we should be able to build at least as well.
V. Proliferation Prevention
Preventing the proliferation of nuclear weapons
is an important goal of the international community. Achieving
this goal becomes more complex in a world with a much expanded
nuclear-energy program involving more countries. Opportunities
for diversion of weapons-usable material exists at both the front
end of the nuclear fuel cycle, the U-235 enrichment stage, and
the back end of the nuclear fuel cycle, the reprocessing and treatment
of spent fuel stage. The more places this work is done, the harder
it is to monitor.
Clandestine weapons development programs have
come from both ends of the fuel cycle. Clandestine enrichment
programs can lead to U235 weapons. Chemical separation techniques
can produce from spent fuel the material needed for plutonium
weapons. For example, concern about Iran’s program relate
to the enrichment phase, while concern about North Korea’s
relate to reprocessing spent fuel.
The level of technical sophistication of the
countries that have developed nuclear weapons outside of the NPT
range from very low to very high, yet all managed to succeed.
The science behind nuclear weapons is well known and the technology
seems to be not that hard to master through internal development
or illicit acquisition. It should be clear to all that the only
way to limit proliferation by nation states it through binding
international agreements that include effective inspection as
a deterrent, and effective sanctions when the deterrent fails.
We in the science and technology (S&T) community
can give the diplomats improved tools that may make the monitoring
that goes with agreements simpler and less overtly intrusive.
These technical safeguards are the heart of the systems used to
identify proliferation efforts at the earliest possible stage.
They must search out theft and diversion of weapons-usable material
as well as identifying clandestine facilities that could be used
to make weapons-usable materials.
The development of advanced technical safeguards
has not received much funding recently. An internationally coordinated
program for their development needs to be implemented, and proliferation
resistance and monitoring technology should be an essential part
of the design of all new reactors, enrichment plants, reprocessing
facility, and fuel fabrication sites.
Some have asserted that reprocessing of spent
fuel leads to less proliferation resistance that the “once
through” fuel cycle. Recent analysis, however, seems to
show that the “once through” fuel cycle is not significantly
more proliferation resistant than reprocessing systems like that
used in France (see, for example, “An Evaluation of Proliferation
Resistant Characteristics of Light Water Reactor Fuels,”
November 2004, available on the DOE’s website (www.nuclear.gov)
under “Advisory Committee Reports”). This is an important
conclusion since one of the objections to the reprocessing schemes
needed to mitigate the spent fuel problem was that it might increase
Recently the IAEA Director General, Dr. ElBaradei,
and United States President, George Bush, have proposed that internationalization
of the nuclear fuel cycle be seriously studied. In an internationalization
scenario there are countries where enrichment and reprocessing
occur. These are the supplier countries. The rest are user countries.
Supplier countries make the nuclear fuel and take back spent fuel
for reprocessing, separating the components into those that are
to be disposed of and those that go back into new fuel.
If such a scheme were to be satisfactorily implemented
there would be enormous benefits to the user countries, particularly
the smaller ones. They would not have to build enrichment facilities
nor would they have to treat or dispose of spent fuel. Neither
is economic on small scales and repository sites with the proper
geology may not be available in small countries. In return for
these benefits, user countries would give up potential access
to weapons-usable material from both the front end and the back
ends of the fuel cycle.
If this is to work, an international regime has
to be created that will give the user nations guaranteed access
to the fuel that they require. This is not going to be easy and
needs a geographically and politically diverse set of supplier
countries to give confidence to user countries that they will
not be cut off from the fuel required for an essential part of
their energy supply.
Reducing the proliferation risk from the back
end of the fuel cycle will be even more complex. It is essential
to do so because we have seen from the example of North Korea
how quickly a country can “break out” from an international
agreement and develop weapons if the material is available. North
Korea withdrew from the Non-Proliferation Treaty at short notice,
expelled the IAEA inspectors, and reprocessed the spent fuel from
their Yongbyon reactor, thus acquiring in a very short time the
plutonium needed for bomb fabrication.
However, the supplier countries that should take
back the spent fuel for treatment are not likely to do so without
a solution to the waste-disposal problem. In a world with a greatly
expanded nuclear power program there will be a huge amount of
spent fuel generated worldwide. The projections mentioned earlier
predict, by mid-century, the deployment of more than a terawatt
(electric) of nuclear capacity producing more than 20,000 tons
of spent fuel per year. This spent fuel contains about 200 tons
of plutonium and minor actinides and 800 tons of fission fragments.
The once-through fuel cycle cannot handle it without requiring
a new Yucca Mountain scale repository opening somewhere in the
world every two or three years.
The U.S. government has recognized this and is
changing its R&D direction to focus on reprocessing spent
fuel and burning the actinides in fast reactors with continuous
recycle. This program, the Global Nuclear Energy Initiative (GNEI)
aims to develop the technology to allow the implementation of
an internationalized fuel cycle as well as to handle its own nuclear
waste. The U.S. long-range program is now aligned with those of
France, Russia, China, Japan, Korea, and India. The possibility
exists for an effective, international control regime.
In this model the supplier-user scenario might
develop as follows. First, every one uses LWRs. Then the supplier
countries begin to install fast-spectrum systems. These would
be used to supply their electricity needs as well as to burn down
the actinides. Eventually, when uranium supplies begin to run
short, the user countries would go over to fast-burner systems,
while the supplier countries would have a combination of breeders
and burners as required.
In summary, nuclear energy is an important component
of a strategy to give the world the energy resources it needs
for economic development while reducing consumption of fossil
fuels with their greenhouse-gas emissions. If this is to happen
on a large scale, advances in both physical S&T and political
S&T will be required.
We on the physical S&T side can produce better
and safer reactors, better ways to dispose of spent fuel, and
better safeguards technology. This can best be done in an international
context to spread the cost and to create an international technical
consensus on what should be done. Countries will be more comfortable
with what comes out of such developments if they are part of them.
While the physical S&T development can best
be done in an international context, the political S&T can
only be done internationally. The IAEA seems to be the best place
to start and the first baby steps have already been taken. I look
forward to larger steps of both kinds in the future. However,
it will be difficult for an organization as large as the IAEA
to create a framework for a new international nuclear enterprise
if too many voices are involved at the start. Discussions might
start off better if a broadly based, but compact, subgroup does
the initial work. If I were setting up such a group, the minimum
membership would include Canada, China, France, India, Japan,
Russia, South Korea, United States, and representatives of the
larger potential user states, Brazil and Indonesia, for example.
I do not think it will be difficult to create mechanisms for the
front end of the fuel cycle. The back end will be the problem
and the most intractable issue is likely to be the final waste