The Economist  [Printer-friendly version]
September 6, 2007

ATOMIC RENAISSANCE

America's nuclear industry is about to embark on its biggest expansion
in more than a generation. This will influence energy policy in the
rest of the world

OVER the next few months America's Nuclear Regulatory Commission (NRC)
expects to receive 12 applications to build new nuclear-power reactors
at seven different sites. It is preparing to see plans for another 15
at 11 more locations next year. These will be the first full
applications to build new nuclear plants in America for 30 years. If
they are all successful, the number of reactors in the country will
increase by roughly a third. The output of nuclear electricity would
grow even more sharply -- the new reactors would be more powerful than
older ones. The new enthusiasm for building reactors means America's
long-awaited "nuclear renaissance" is about to become reality.

Whether it is a leap forwards or a step backwards remains to be seen.
Since the 1970s, far from being "too cheap to meter" -- as it
proponents once blithely claimed -- nuclear power has proved too
expensive to matter. The problem is finance: nuclear plants cost a lot
to build but are relatively cheap to run, unlike gas-fired ones, for
which the reverse is true. So to be profitable they must be built
quickly, to minimise the period when no revenue is coming in and
interest payments are piling up on construction loans. Yet America's
previous generation of nuclear plants was plagued by safety scares,
design revisions and time-consuming regulatory procedures, which
resulted in ruinously protracted construction.

America's most recent nuclear plant, at Watts Bar in Tennessee,
started operations in 1996. But it took 23 years to complete at a cost
of $6.9 billion; a second reactor at the site has been under
construction, on and off, since 1973. Another plant, at Shoreham in
New York, was completed and tested, but never allowed to start
commercial operations because of local opposition. By the time it was
decommissioned, in 1994 -- 21 years after construction had begun --
the costs had exploded from $70m to $6 billion. The local utility was
able to pass most of this bill on to its customers. Not all energy
firms have been so lucky: in 1988 Public Service Company of New
Hampshire became the first American utility to go bust since the
Depression, thanks largely to the fallout from a much-delayed nuclear
project.

Even when they were switched on, nuclear-power stations did not fulfil
their promise. They were supposed to run almost constantly, but proved
much less reliable. In the early 1970s, for example, the average
nuclear plant produced power for under half the time. Since most
utilities had planned to run them flat out to generate enough revenue
to repay their debts, this poor performance led to further financial
troubles. And, as anti-nuclear activists complain, all this happened
despite the government's generous subsidies to help cover the costs of
developing new designs and building prototypes.

As for safety... What is worse, nuclear power has a spotty safety
record. There have never been any catastrophic releases of radiation
in Western countries. But one did occur in 1986 at Chernobyl, in what
was then the Soviet Union and is now Ukraine. America came perilously
close to such a disaster in 1979, when a reactor at Three Mile Island
in Pennsylvania overheated and began to melt down. There have been
lesser safety scares and scandals in many countries, including
Britain, Germany and Sweden. In August an earthquake resulted in
several small leaks of radioactive material from a nuclear reactor in
Japan.

The next generation of nuclear plants is said to be very different.
Firms which make them, such as America's General Electric and
Westinghouse, and foreign manufacturers like France's AREVA, insist
that such episodes will soon be a thing of the past. Their latest
designs, they maintain, are simpler and safer than existing nuclear
plants. That should make it easier to obtain operating permits, allow
them to be built faster and be cheaper to run -- and so much less
risky financially. Meanwhile, contractors are said to be getting
better at building them, the NRC better at regulating them and
utilities better at running them. Although nuclear power's boosters
welcome a smorgasbord of new subsidies that Congress has approved to
nourish the industry, they say that in the long run even this will not
be necessary because the industry will be able to move forward under
its own nuclear-generated steam.

Consolidating reactors America's utilities have certainly warmed to
their existing nuclear- power plants now that they are running them
more efficiently. In the 1970s, says Colette Lewiner, of Capgemini, a
consultancy, even small municipally owned firms ordered nuclear
reactors, imagining they would be no more complicated to operate than
their existing power stations, except in so far as workers would need
uranium to shovel into the furnace instead of coal. But they found
that they had neither the expertise to maintain their new investments,
nor the scale to absorb all the extra regulatory costs, nor the clout
to secure fuel and parts at competitive prices. Many ended up putting
their nuclear plants up for sale.

That allowed bigger firms to acquire reactors on the cheap, and thus
to achieve economies of scale and to capitalise on their experience.
These nuclear specialists have been able to speed up the refuelling
process, keep shutdowns for maintenance to a minimum and so keep the
reactors going more of the time. Last year the average nuclear reactor
in America was in use 90% of the time. Better still, utilities have
found ways to improve the non-nuclear parts of the power station, such
as the steam turbines. These so-called "uprates" have increased
America's nuclear capacity by almost 5,000MW since 1977, the
equivalent of about five new nuclear reactors, according to the
Nuclear Energy Institute, an industry group. At the same time, the NRC
has agreed to extend the working life of about half of America's
nuclear plants for an extra 20 years.

All this has turned nuclear-power plants into virtual mints -- as long
as the bill for construction has been paid down or written off. In
most of America, the wholesale power price is closely linked to the
price of natural gas, since gas-fired plants tend to provide the extra
power required at times of peak demand. So the price of power has
risen along with that of gas over the past few years, whereas the
operating costs of nuclear plants have remained relatively stable.
According to the Energy Information Administration, a government
agency, the average wholesale power price in 2005 was 5 cents per
kilowatt-hour (kWh); the Nuclear Energy Institute, an industry group,
reckons that the average operating cost of America's nuclear plants
was 1.7 cents per kWh that year. So their margins were almost 200%.

No wonder that utilities are rushing to the NRC with their plans for
new reactors. But to get any of them off the ground they must not only
persuade the NRC of the safety of their designs, but also convince
potential creditors that there will be no repeat of the financial
meltdowns of the 1970s and 1980s. They point to three reasons for
optimism: changing conditions in the energy business, a streamlining
of the NRC's process for obtaining permits and an overhaul of
construction techniques.

Until recently coal-fired plants seemed to be safer investments. But
nowadays most utilities expect -- and in some cases are calling for --
Congress to limit emissions of greenhouse gases in the near future to
temper climate change. Coal-fired plants, which have a working life of
40 years or more, spew out globe-warming pollution, whereas nuclear
ones produce almost no greenhouse gases at all. So coal is now subject
to a massive "regulatory risk" of its own. Utilities are piling into
green-generation technologies, such as wind turbines and solar panels.
But for a constant source of clean power, they have few choices other
than nuclear.

Meanwhile, to avoid the fiascos of the past the NRC has simplified its
procedures. It used to require utilities to obtain two different
licences, the first to build a nuclear plant and after that a second
licence to start it up. Both applications involved lengthy reviews,
which culminated in interminable public hearings. New reactors, like
the one at Shoreham, could be finished at great expense and yet never
secure an operating licence. So the NRC is combining the two stages:
utilities can now apply for a single "combined construction and
operating licence". Construction need not start -- and for most bits
of the plant is not allowed to -- until the licence is issued.

To speed things up even more, the NRC is allowing firms selling
nuclear reactors to get designs cleared in advance. That way, when a
utility applies to build a reactor of an approved design, the NRC will
only need to review the modifications that are unique to its site.
Westinghouse has already got its AP-1000 model cleared; the NRC is in
the process of certifying GE's latest design, called the ESBWR, and
AREVA is about to submit an application for its new offering, the EPR.

By the same token, utilities can now ask the NRC to approve a location
as suitable for a nuclear-power station before they go to the trouble
and expense of applying for a combined licence. Four firms have asked
for these "early site permits" and two have already received them.
Another short cut involves submitting the environmental part of a
combined licence before the part that deals with the design. UniStar,
a joint venture between America's Constellation and Electricite de
France (EDF), filed that sort of paperwork in July for a new reactor
in Maryland.

The NRC has also made a point of asking utilities about their nuclear
plans before any applications arrive. This is so it can be sure it
will have enough staff to handle them -- which is how it knows how
many new plants are in the works. It is hiring about 200 new staff
every year, and since most of the utilities contemplating nuclear
plants are in the South, has set up a field office in Georgia to co-
ordinate with them directly. It is even planning to suggest to
Congress possible amendments to the relevant laws to reduce the hassle
and uncertainty of licensing even more.

The process will still be time-consuming: the NRC reckons it will need
two and a half years to review each application and a further year to
conduct hearings on its conclusions. Certification of new reactor
designs might take as long as four years: AREVA says its application
for the EPR runs to 17,000 pages and fills a small bookcase.
Nonetheless, the NRC aims to issue its first new licences at some
point in 2011.

Obstreperous local authorities could still put a spanner in the works.
It was opposition from county and state officials, for example, that
finally did in the Shoreham plant. Although they have no explicit
authority to block a new reactor, local officials can withhold permits
to use the water from a river for cooling, for example, or refuse to
co-operate on emergency planning. But utilities are hoping to avoid
such pitfalls by locating their new reactors only in welcoming
jurisdictions -- preferably next door to existing ones. Locals in such
places know that expanding existing nuclear facilities will bring more
jobs and produce more tax revenue. Moreover, they have grown
accustomed to having nuclear reactors nearby and do not find the idea
particularly frightening. As Dale Klein, chairman of the NRC, puts it,
the staff and management of nuclear plants, and local residents, all
go to church together.

Fast builders Utilities are also confident that they can build new
reactors more quickly than before. Many have already placed orders for
the parts that take a long time to build. They have also brought in
partners that have completed nuclear projects on time and on budget in
other countries. Westinghouse, for example, points to the exemplary
record of its parent company, Toshiba, in Japan. Similarly, GE has
teamed up with Hitachi, another respected Japanese nuclear contractor.
AREVA, meanwhile, looks to the series of successful plants it has
built in conjunction with EDF in France. All three vendors say they
plan to save time and money by using as many identical parts as
possible for the different nuclear plants they build in America --
unlike the bespoke designs of the past. All this should reduce the
time required for construction to four years, they say, which would
allow the first new reactors to enter service in 2015 or 2016.

But bankers are still sceptical. They are worried that when the new
designs and the NRC's new procedures are put to the test, hidden flaws
will emerge. After all, the first of AREVA's EPR designs is under
construction in Finland and is two years behind schedule and
dramatically over budget. To avoid such nasty surprises, NRG Energy, a
power-generation company that is applying to build two new reactors in
Texas, has opted for one of GE's older and already-proven designs,
even though GE insists that its ESBWR will be cheaper to build and to
run. Other utilities are planning to build nuclear plants in the
regulated markets of the South, in the hope that the regulators will
allow them to pass any cost over-runs on to their customers.

Even so, says David Crane, NRG's boss, banks are simply not prepared
to lend money to build nuclear plants in America without some extra
surety. The Energy Policy Act, which Congress approved in 2005, is
supposed to provide that. It offers four different types of subsidies
for new reactors. First, it grants up to $2 billion in insurance
against regulatory delays and lawsuits to the first six reactors to
receive licences and start construction. Second, it extends an older
law limiting a utility's liability to $10 billion in the event of a
nuclear accident. Third, it provides a tax credit of 1.8 cents per kWh
for the first 6,000MW generated by new plants. Fourth, and most
importantly, it offers guarantees for an indeterminate amount of loans
to fund new nuclear reactors and other types of power plant using
"innovative" technology.

The scope of these loan guarantees is the subject of great
controversy. Some politicians fear that the costs of the programme
might balloon; others complain that the Department of Energy, which
will administer it, is too stingy. Meanwhile, some financial experts
argue that the rules, as drafted, would not allow issuing banks to
repackage and sell on the loans in question, making them less
attractive. There is also some debate about what proportion of a
nuclear plant's debt should be covered: the act says up to 80% of the
costs of construction, but that might be sufficient to cover the full
amount borrowed, leaving banks with no risk at all. Until all this is
settled, utility bosses insist, new nuclear plants will not get built.

Clearing up afterwards The fate of America's nuclear waste, which the
government has vowed to sequester for a million years, is another
unresolved issue. In theory, the Department of Energy is in charge of
looking after it all. It requires utilities to set aside a tenth of a
cent for each kilowatt- hour of nuclear power they generate to help
defray the costs of transporting nuclear waste to a safe repository
and storing it there permanently. The only hitch is that no such
repository yet exists.

Most countries with nuclear power have determined that the safest way
to store their waste is underground, deep in the bedrock in air- and
watertight containers. But no one has actually built such a facility.
America has got as far as selecting a site for one, at Yucca Mountain,
a ridge in the middle of a former nuclear-testing ground in Nevada.
The Department of Energy is planning to submit an application to the
NRC next year to build a repository there. The NRC, in turn, thinks
the application will take about three years to review. Officials say
the facility will be open for business in 2017.

But Harry Reid, a senator from Nevada, has vowed to derail the scheme.
As it is, Congress has been cutting funding for the Yucca Mountain
project, which was first proposed in 1978 and has since been the
subject of several lawsuits. Now that Mr Reid has become Senate
majority leader the odds of the repository ever getting built have
diminished.

Meanwhile, nuclear waste continues to pile up in ponds and containers
at nuclear plants around the country. The NRC monitors these and
claims that they are safe for the foreseeable future. But Mr Klein,
its chairman, tactfully hints that it would be prudent for the
government to find a more permanent solution, especially since it is
encouraging a dramatic expansion of nuclear power.

Nonetheless, Mr Klein believes that the expansion of nuclear energy is
now in motion and is unlikely to be slowed down by concerns about what
to do with the waste. The only thing that could stop a nuclear
renaissance now, he suggests, is a serious accident at an existing
plant. Unfortunately, it would not be the first.

Copyright The Economist Newspaper Limited 2007