sab

Nuclear Nightmares

May 26, 2002
By BILL KELLER

Not If But When Everybody who spends much time thinking
about nuclear terrorism can give you a scenario, something
diabolical and, theoretically, doable. Michael A. Levi, a
researcher at the Federation of American Scientists,
imagines a homemade nuclear explosive device detonated
inside a truck passing through one of the tunnels into
Manhattan. The blast would crater portions of the New York
skyline, barbecue thousands of people instantly, condemn
thousands more to a horrible death from radiation sickness
and -- by virtue of being underground -- would vaporize
many tons of concrete and dirt and river water into an
enduring cloud of lethal fallout. Vladimir Shikalov, a
Russian nuclear physicist who helped clean up after the
1986 Chernobyl accident, envisioned for me an attack
involving highly radioactive cesium-137 loaded into some
kind of homemade spraying device, and a target that sounded
particularly unsettling when proposed across a Moscow
kitchen table -- Disneyland. In this case, the human toll
would be much less ghastly, but the panic that would result
from contaminating the Magic Kingdom with a modest amount
of cesium -- Shikalov held up his teacup to illustrate how
much -- would probably shut the place down for good and
constitute a staggering strike at Americans' sense of
innocence. Shikalov, a nuclear enthusiast who thinks most
people are ridiculously squeamish about radiation, added
that personally he would still be happy to visit Disneyland
after the terrorists struck, although he would pack his own
food and drink and destroy his clothing afterward.

Another Russian, Dmitry Borisov, a former official of his
country's atomic energy ministry, conjured a suicidal
pilot. (Suicidal pilots, for obvious reasons, figure
frequently in these fantasies.) In Borisov's scenario, the
hijacker dive-bombs an Aeroflot jetliner into the Kurchatov
Institute, an atomic research center in a gentrifying
neighborhood of Moscow, which I had just visited the day
before our conversation. The facility contains 26 nuclear
reactors of various sizes and a huge accumulation of
radioactive material. The effect would probably be measured
more in property values than in body bags, but some people
say the same about Chernobyl.

Maybe it is a way to tame a fearsome subject by
Hollywoodizing it, or maybe it is a way to drive home the
dreadful stakes in the arid-sounding business of
nonproliferation, but in several weeks of talking to
specialists here and in Russia about the threats an amateur
evildoer might pose to the homeland, I found an unnerving
abundance of such morbid creativity. I heard a physicist
wonder whether a suicide bomber with a pacemaker would
constitute an effective radiation weapon. (I'm a little
ashamed to say I checked that one, and the answer is no,
since pacemakers powered by plutonium have not been
implanted for the past 20 years.) I have had people
theorize about whether hijackers who took over a nuclear
research laboratory could improvise an actual nuclear
explosion on the spot. (Expert opinions differ, but it's
very unlikely.) I've been instructed how to disperse
plutonium into the ventilation system of an office
building.

The realistic threats settle into two broad categories. The
less likely but far more devastating is an actual nuclear
explosion, a great hole blown in the heart of New York or
Washington, followed by a toxic fog of radiation. This
could be produced by a black-market nuclear warhead
procured from an existing arsenal. Russia is the favorite
hypothetical source, although Pakistan, which has a program
built on shady middlemen and covert operations, should not
be overlooked. Or the explosive could be a homemade device,
lower in yield than a factory nuke but still creating great
carnage.

The second category is a radiological attack, contaminating
a public place with radioactive material by packing it with
conventional explosives in a ''dirty bomb'' by dispersing
it into the air or water or by sabotaging a nuclear
facility. By comparison with the task of creating nuclear
fission, some of these schemes would be almost childishly
simple, although the consequences would be less horrifying:
a panicky evacuation, a gradual increase in cancer rates, a
staggeringly expensive cleanup, possibly the need to
demolish whole neighborhoods. Al Qaeda has claimed to have
access to dirty bombs, which is unverified but entirely
plausible, given that the makings are easily gettable.

Nothing is really new about these perils. The means to
inflict nuclear harm on America have been available to
rogues for a long time. Serious studies of the threat of
nuclear terror date back to the 1970's. American programs
to keep Russian nuclear ingredients from falling into
murderous hands -- one of the subjects high on the agenda
in President Bush's meetings in Moscow this weekend -- were
hatched soon after the Soviet Union disintegrated a decade
ago. When terrorists get around to trying their first
nuclear assault, as you can be sure they will, there will
be plenty of people entitled to say I told you so.

All Sept. 11 did was turn a theoretical possibility into a
felt danger. All it did was supply a credible cast of
characters who hate us so much they would thrill to the
prospect of actually doing it -- and, most important in
rethinking the probabilities, would be happy to die in the
effort. All it did was give our nightmares legs.

And of the many nightmares animated by the attacks, this is
the one with pride of place in our experience and
literature -- and, we know from his own lips, in Osama bin
Laden's aspirations. In February, Tom Ridge, the Bush
administration's homeland security chief, visited The Times
for a conversation, and at the end someone asked, given all
the things he had to worry about -- hijacked airliners,
anthrax in the mail, smallpox, germs in crop-dusters --
what did he worry about most? He cupped his hands
prayerfully and pressed his fingertips to his lips.
''Nuclear,'' he said simply.

My assignment here was to stare at that fear and inventory
the possibilities. How afraid should we be, and what of,
exactly? I'll tell you at the outset, this was not one of
those exercises in which weighing the fears and assigning
them probabilities laid them to rest. I'm not evacuating
Manhattan, but neither am I sleeping quite as soundly. As I
was writing this early one Saturday in April, the floor
began to rumble and my desk lamp wobbled precariously.
Although I grew up on the San Andreas Fault, the fact that
New York was experiencing an earthquake was only my second
thought.

The best reason for thinking it won't happen is that it
hasn't happened yet, and that is terrible logic. The
problem is not so much that we are not doing enough to
prevent a terrorist from turning our atomic knowledge
against us (although we are not). The problem is that there
may be no such thing as ''enough.''

25,000 Warheads, and It Only Takes One My few actual
encounters with the Russian nuclear arsenal are all
associated with Thomas Cochran. Cochran, a physicist with a
Tennessee lilt and a sense of showmanship, is the director
of nuclear issues for the Natural Resources Defense
Council, which promotes environmental protection and arms
control. In 1989, when glasnost was in flower, Cochran
persuaded the Soviet Union to open some of its most secret
nuclear venues to a roadshow of American scientists and
congressmen and invited along a couple of reporters. We
visited a Soviet missile cruiser bobbing in the Black Sea
and drank vodka with physicists and engineers in the secret
city where the Soviets first produced plutonium for
weapons.

Not long ago Cochran took me cruising through the Russian
nuclear stockpile again, this time digitally. The days of
glasnost theatrics are past, and this is now the only way
an outsider can get close to the places where Russians
store and deploy their nuclear weapons. On his office
computer in Washington, Cochran has installed a detailed
United States military map of Russia and superimposed upon
it high-resolution satellite photographs. We spent part of
a morning mouse-clicking from missile-launch site to
submarine base, zooming in like voyeurs and contemplating
the possibility that a terrorist could figure out how to
steal a nuclear warhead from one of these places.

''Here are the bunkers,'' Cochran said, enlarging an area
the size of a football stadium holding a half-dozen
elongated igloos. We were hovering over a site called
Zhukovka, in western Russia. We were pleased to see it did
not look ripe for a hijacking.

''You see the bunkers are fenced, and then the whole thing
is fenced again,'' Cochran said. ''Just outside you can see
barracks and a rifle range for the guards. These would be
troops of the 12th Main Directorate. Somebody's not going
to walk off the street and get a Russian weapon out of this
particular storage area.''

In the popular culture, nuclear terror begins with the
theft of a nuclear weapon. Why build one when so many are
lying around for the taking? And stealing tends to make
better drama than engineering. Thus the stolen nuke has
been a staple in the literature at least since 1961, when
Ian Fleming published ''Thunderball,'' in which the
malevolent Spectre (the Special Executive for
Counterintelligence, Terrorism, Revenge and Extortion, a
strictly mercenary and more technologically sophisticated
precursor to al Qaeda) pilfers a pair of atom bombs from a
crashed NATO aircraft. In the movie version of Tom Clancy's
thriller ''The Sum of All Fears,'' due in theaters this
week, neo-Nazis get their hands on a mislaid Israeli nuke,
and viewers will get to see Baltimore blasted to oblivion.

Eight countries are known to have nuclear weapons -- the
United States, Russia, China, Great Britain, France, India,
Pakistan and Israel. David Albright, a nuclear-weapons
expert and president of the Institute for Science and
International Security, points out that Pakistan's program
in particular was built almost entirely through black
markets and industrial espionage, aimed at circumventing
Western export controls. Defeating the discipline of
nuclear nonproliferation is ingrained in the culture.
Disaffected individuals in Pakistan (which, remember, was
intimate with the Taliban) would have no trouble finding
the illicit channels or the rationalization for diverting
materials, expertise -- even, conceivably, a warhead.

But the mall of horrors is Russia, because it currently
maintains something like 15,000 of the world's (very
roughly) 25,000 nuclear warheads, ranging in destructive
power from about 500 kilotons, which could kill a million
people, down to the one-kiloton land mines that would be
enough to make much of Manhattan uninhabitable. Russia is a
country with sloppy accounting, a disgruntled military, an
audacious black market and indigenous terrorists.

There is anecdotal reason to worry. Gen. Igor Valynkin,
commander of the 12th Main Directorate of the Russian
Ministry of Defense, the Russian military sector in charge
of all nuclear weapons outside the Navy, said recently that
twice in the past year terrorist groups were caught casing
Russian weapons-storage facilities. But it's hard to know
how seriously to take this. When I made the rounds of
nuclear experts in Russia earlier this year, many were
skeptical of these near-miss anecdotes, saying the security
forces tend to exaggerate such incidents to dramatize their
own prowess (the culprits are always caught) and enhance
their budgets. On the whole, Russian and American military
experts sound not very alarmed about the vulnerability of
Russia's nuclear warheads. They say Russia takes these
weapons quite seriously, accounts for them rigorously and
guards them carefully. There is no confirmed case of a
warhead being lost. Strategic warheads, including the 4,000
or so that President Bush and President Vladimir Putin have
agreed to retire from service, tend to be stored in
hard-to-reach places, fenced and heavily guarded, and their
whereabouts are not advertised. The people who guard them
are better paid and more closely vetted than most Russian
soldiers.

Eugene E. Habiger, the four-star general who was in charge
of American strategic weapons until 1998 and then ran
nuclear antiterror programs for the Energy Department,
visited several Russian weapons facilities in 1996 and
1997. He may be the only American who has actually entered
a Russian bunker and inspected a warhead in situ. Habiger
said he found the overall level of security comparable to
American sites, although the Russians depend more on people
than on technology to protect their nukes.

The image of armed terrorist commandos storming a nuclear
bunker is cinematic, but it's far more plausible to think
of an inside job. No observer of the unraveling Russian
military has much trouble imagining that a group of
military officers, disenchanted by the humiliation of
serving a spent superpower, embittered by the wretched
conditions in which they spend much of their military lives
or merely greedy, might find a way to divert a warhead to a
terrorist for the right price. (The Chechen warlord Shamil
Basayev, infamous for such ruthless exploits as taking an
entire hospital hostage, once hinted that he had an
opportunity to buy a nuclear warhead from the stockpile.)
The anecdotal evidence of desperation in the military is
plentiful and disquieting. Every year the Russian press
provides stories like that of the 19-year-old sailor who
went on a rampage aboard an Akula-class nuclear submarine,
killing eight people and threatening to blow up the boat
and its nuclear reactor; or the five soldiers at Russia's
nuclear-weapons test site who killed a guard, took a
hostage and tried to hijack an aircraft, or the officers
who reportedly stole five assault helicopters, with their
weapons pods, and tried to sell them to North Korea.

The Clinton administration found the danger of disgruntled
nuclear caretakers worrisome enough that it considered
building better housing for some officers in the nuclear
rocket corps. Congress, noting that the United States does
not build housing for its own officers, rejected the idea
out of hand.

If a terrorist did get his hands on a nuclear warhead, he
would still face the problem of setting it off. American
warheads are rigged with multiple PAL's ( ''permissive
action links'') -- codes and self-disabling devices
designed to frustrate an unauthorized person from
triggering the explosion. General Habiger says that when he
examined Russian strategic weapons he found the level of
protection comparable to our own. ''You'd have to literally
break the weapon apart to get into the gut,'' he told me.
''I would submit that a more likely scenario is that
there'd be an attempt to get hold of a warhead and not
explode the warhead but extract the plutonium or highly
enriched uranium.'' In other words, it's easier to take the
fuel and build an entire weapon from scratch than it is to
make one of these things go off.

Then again, Habiger is not an expert in physics or weapons
design. Then again, the Russians would seem to have no
obvious reason for misleading him about something that
important. Then again, how many times have computer hackers
hacked their way into encrypted computers we were assured
were impregnable? Then again, how many computer hackers
does al Qaeda have? This subject drives you in circles.

The most troublesome gap in the generally reassuring
assessment of Russian weapons security is those tactical
nuclear warheads -- smaller, short-range weapons like
torpedoes, depth charges, artillery shells, mines. Although
their smaller size and greater number makes them ideal
candidates for theft, they have gotten far less attention
simply because, unlike all of our long-range weapons, they
happen not to be the subject of any formal treaty. The
first President Bush reached an informal understanding with
President Gorbachev and then with President Yeltsin that
both sides would gather and destroy thousands of tactical
nukes. But the agreement included no inventories of the
stockpiles, no outside monitoring, no verification of any
kind. It was one of those trust-me deals that, in the
hindsight of Sept. 11, amount to an enormous black hole in
our security.

Did I say earlier there are about 15,000 Russian warheads?
That number includes, alongside the scrupulously counted
strategic warheads in bombers, missiles and submarines, the
commonly used estimate of 8,000 tactical warheads. But that
figure is at best an educated guess. Other educated guesses
of the tactical nukes in Russia go as low as 4,000 and as
high as 30,000. We just don't know. We don't even know if
the Russians know, since they are famous for doing things
off the books. ''They'll tell you they've never lost a
weapon,'' said Kenneth Luongo, director of a private
antiproliferation group called the Russian-American Nuclear
Security Advisory Council. ''The fact is, they don't know.
And when you're talking about warhead counting, you don't
want to miss even one.''

And where are they? Some are stored in reinforced concrete
bunkers like the one at Zhukovka. Others are deployed.
(When the submarine Kursk sank with its 118 crewmen in
August 2000, the Americans' immediate fear was for its
nuclear armaments. The standard load out for a submarine of
that class includes a couple of nuclear torpedoes and
possibly some nuclear depth charges.) Still others are
supposed to be in the process of being dismantled under
terms of various formal and informal arms-control
agreements. Some are in transit. In short, we don't really
know.

The other worrying thing about tactical nukes is that their
anti-use devices are believed to be less sophisticated,
because the weapons were designed to be employed in the
battlefield. Some of the older systems are thought to have
no permissive action links at all, so that setting one off
would be about as complicated as hot-wiring a car.

Efforts to learn more about the state of tactical
stockpiles have been frustrated by reluctance on both sides
to let visitors in. Viktor Mikhailov, who ran the Russian
Ministry of Atomic Energy until 1998 with a famous scorn
for America's nonproliferation concerns, still insists that
the United States programs to protect Russian nuclear
weapons and material mask a secret agenda of
intelligence-gathering. Americans, in turn, sometimes balk
at reciprocal access, on the grounds that we are the ones
paying the bills for all these safety upgrades, said the
former Senator Sam Nunn, co-author of the main American
program for securing Russian nukes, called Nunn-Lugar.

''We have to decide if we want the Russians to be
transparent -- I'd call it cradle-to-grave transparency
with nuclear material and inventories and so forth,'' Nunn
told me. ''Then we have to open up more ourselves. This is
a big psychological breakthrough we're talking about here,
both for them and for us.''

The Garage Bomb One of the more interesting facts about the
atom bomb dropped on Hiroshima is that it had never been
tested. All of those spectral images of nuclear coronas
brightening the desert of New Mexico -- those were to
perfect the more complicated plutonium device that was
dropped on Nagasaki. ''Little Boy,'' the Hiroshima bomb,
was a rudimentary gunlike device that shot one projectile
of highly enriched uranium into another, creating a
critical mass that exploded. The mechanics were so simple
that few doubted it would work, so the first experiment was
in the sky over Japan.

The closest thing to a consensus I heard among those who
study nuclear terror was this: building a nuclear bomb is
easier than you think, probably easier than stealing one.
In the rejuvenated effort to prevent a terrorist from
striking a nuclear blow, this is where most of the
attention and money are focused.

A nuclear explosion of any kind ''is not a sort of
high-probability thing,'' said a White House official who
follows the subject closely. ''But getting your hands on
enough fissile material to build an improvised nuclear
device, to my mind, is the least improbable of them all,
and particularly if that material is highly enriched
uranium in metallic form. Then I'm really worried. That's
the one.''

To build a nuclear explosive you need material capable of
explosive nuclear fission, you need expertise, you need
some equipment, and you need a way to deliver it.

Delivering it to the target is, by most reckoning, the
simplest part. People in the field generally scoff at the
mythologized suitcase bomb; instead they talk of a ''conex
bomb,'' using the name of those shack-size steel containers
that bring most cargo into the United States. Two thousand
containers enter America every hour, on trucks and trains
and especially on ships sailing into more than 300 American
ports. Fewer than 2 percent are cracked open for
inspection, and the great majority never pass through an
X-ray machine. Containers delivered to upriver ports like
St. Louis or Chicago pass many miles of potential targets
before they even reach customs.

''How do you protect against that?'' mused Habiger, the
former chief of our nuclear arsenal. ''You can't. That's
scary. That's very, very scary. You set one of those off in
Philadelphia, in New York City, San Francisco, Los Angeles,
and you're going to kill tens of thousands of people, if
not more.'' Habiger's view is ''It's not a matter of if;
it's a matter of when'' -- which may explain why he now
lives in San Antonio.

The Homeland Security office has installed a plan to
refocus inspections, making sure the 2 percent of
containers that get inspected are those without a clear,
verified itinerary. Detectors will be put into place at
ports and other checkpoints. This is good, but it hardly
represents an ironclad defense. The detection devices are a
long way from being reliable. (Inconveniently, the most
feared bomb component, uranium, is one of the hardest
radioactive substances to detect because it does not emit a
lot of radiation prior to fission.) The best way to stop
nuclear terror, therefore, is to keep the weapons out of
terrorist hands in the first place.

The basic know-how of atom-bomb-building is half a century
old, and adequate recipes have cropped up in physics term
papers and high school science projects. The simplest
design entails taking a lump of highly enriched uranium,
about the size of a cantaloupe, and firing it down a big
gun barrel into a second lump. Theodore Taylor, the nuclear
physicist who designed both the smallest and the largest
American nuclear-fission warheads before becoming a
remorseful opponent of all things nuclear, told me he
recently looked up ''atomic bomb'' in the World Book
Encyclopedia in the upstate New York nursing home where he
now lives, and he found enough basic information to get a
careful reader started. ''It's accessible all over the
place,'' he said. ''I don't mean just the basic principles.
The sizes, specifications, things that work.''

Most of the people who talk about the ease of assembling a
nuclear weapon, of course, have never actually built one.
The most authoritative assessment I found was a paper,
''Can Terrorists Build Nuclear Weapons?'' written in 1986
by five experienced nuke-makers from the Los Alamos weapons
laboratory. I was relieved to learn that fabricating a
nuclear weapon is not something a lone madman -- even a
lone genius -- is likely to pull off in his hobby room. The
paper explained that it would require a team with knowledge
of ''the physical, chemical and metallurgical properties of
the various materials to be used, as well as
characteristics affecting their fabrication; neutronic
properties; radiation effects, both nuclear and biological;
technology concerning high explosives and/or chemical
propellants; some hydrodynamics; electrical circuitry; and
others.'' Many of these skills are more difficult to
acquire than, say, the ability to aim a jumbo jet.

The schemers would also need specialized equipment to form
the uranium, which is usually in powdered form, into metal,
to cast it and machine it to fit the device. That effort
would entail months of preparation, increasing the risk of
detection, and it would require elaborate safeguards to
prevent a mishap that, as the paper dryly put it, would
''bring the operation to a close.''

Still, the experts concluded, the answer to the question
posed in the title, while qualified, was ''Yes, they can.''

David Albright, who worked as a United Nations weapons
inspector in Iraq, says Saddam Hussein's unsuccessful crash
program to build a nuclear weapon in 1990 illustrates how a
single bad decision can mean a huge setback. Iraq had
extracted highly enriched uranium from research-reactor
fuel and had, maybe, barely enough for a bomb. But the
manager in charge of casting the metal was so afraid the
stuff would spill or get contaminated that he decided to
melt it in tiny batches. As a result, so much of the
uranium was wasted that he ended up with too little for a
bomb.

''You need good managers and organizational people to put
the elements together,'' Albright said. ''If you do a
straight-line extrapolation, terrorists will all get
nuclear weapons. But they make mistakes.''

On the other hand, many experts underestimate the prospect
of a do-it-yourself bomb because they are thinking too
professionally. All of our experience with these weapons is
that the people who make them (states, in other words) want
them to be safe, reliable, predictable and efficient.
Weapons for the American arsenal are designed to survive a
trip around the globe in a missile, to be accident-proof,
to produce a precisely specified blast.

But there are many corners you can cut if you are content
with a big, ugly, inefficient device that would make a
spectacular impression. If your bomb doesn't need to fit in
a suitcase (and why should it?) or to endure the stress of
a missile launch; if you don't care whether the explosive
power realizes its full potential; if you're willing to
accept some risk that the thing might go off at the wrong
time or might not go off at all, then the job of building
it is immeasurably simplified.

''As you get smarter, you realize you can get by with
less,'' Albright said. ''You can do it in facilities that
look like barns, garages, with simple machine tools. You
can do it with 10 to 15 people, not all Ph.D.'s, but some
engineers, technicians. Our judgment is that a gun-type
device is well within the capability of a terrorist
organization.''

All the technological challenges are greatly simplified if
terrorists are in league with a country -- a place with an
infrastructure. A state is much better suited to hire
expertise (like dispirited scientists from decommissioned
nuclear installations in the old Soviet Union) or to send
its own scientists for M.I.T. degrees.

Thus Tom Cochran said his greatest fear is what you might
call a bespoke nuke -- terrorists stealing a quantity of
weapons-grade uranium and taking it to Iraq or Iran or
Libya, letting the scientists and engineers there fashion
it into an elementary weapon and then taking it away for a
delivery that would have no return address.

That leaves one big obstacle to the terrorist nuke-maker:
the fissile material itself.

To be reasonably sure of a nuclear explosion, allowing for
some material being lost in the manufacturing process, you
need roughly 50 kilograms -- 110 pounds -- of highly
enriched uranium. (For a weapon, more than 90 percent of
the material should consist of the very unstable
uranium-235 isotope.) Tom Cochran, the master of visual
aids, has 15 pounds of depleted uranium that he keeps in a
Coke can; an eight-pack would be plenty to build a bomb.

The world is awash in the stuff. Frank von Hippel, a
Princeton physicist and arms-control advocate, has
calculated that between 1,300 and 2,100 metric tons of
weapons-grade uranium exists -- at the low end, enough for
26,000 rough-hewed bombs. The largest stockpile is in
Russia, which Senator Joseph Biden calls ''the candy store
of candy stores.''

Until a decade ago, Russian officials say, no one worried
much about the safety of this material. Viktor Mikhailov,
who ran the atomic energy ministry and now presides over an
affiliated research institute, concedes there were glaring
lapses.

''The safety of nuclear materials was always on our minds,
but the focus was on intruders,'' he said. ''The system had
never taken account of the possibility that these carefully
screened people in the nuclear sphere could themselves
represent a danger. The system was not designed to prevent
a danger from within.''

Then came the collapse of the Soviet Union and, in the
early 90's, a few frightening cases of nuclear materials
popping up on the black market.

If you add up all the reported attempts to sell highly
enriched uranium or plutonium, even including those that
have the scent of security-agency hype and those where the
material was of uncertain quality, the total amount of
material still falls short of what a bomb-maker would need
to construct a single explosive.

But Yuri G. Volodin, the chief of safeguards at
Gosatomnadzor, the Russian nuclear regulatory agency, told
me his inspectors still discover one or two instances of
attempted theft a year, along with dozens of violations of
the regulations for storing and securing nuclear material.
And as he readily concedes: ''These are the detected cases.
We can't talk about the cases we don't know.'' Alexander
Pikayev, a former aide to the Defense Committee of the
Russian Duma, said: ''The vast majority of installations
now have fences. But you know Russians. If you walk along
the perimeter, you can see a hole in the fence, because the
employees want to come and go freely.''

The bulk of American investment in nuclear safety goes to
lock the stuff up at the source. That is clearly the right
priority. Other programs are devoted to blending down the
highly enriched uranium to a diluted product unsuitable for
weapons but good as reactor fuel. The Nuclear Threat
Initiative, financed by Ted Turner and led by Nunn, is
studying ways to double the rate of this diluting process.

Still, after 10 years of American subsidies, only 41
percent of Russia's weapon-usable material has been
secured, according to the United States Department of
Energy. Russian officials said they can't even be sure how
much exists, in part because the managers of nuclear
facilities, like everyone else in the Soviet industrial
complex, learned to cook their books. So the barn door is
still pretty seriously ajar. We don't know whether any
horses have gotten out.

And it is not the only barn. William C. Potter, director of
the Center for Nonproliferation Studies at the Monterey
Institute of International Studies and an expert in nuclear
security in the former Soviet states, said the American
focus on Russia has neglected other locations that could be
tempting targets for a terrorist seeking bomb-making
material. There is, for example, a bomb's worth of
weapons-grade uranium at a site in Belarus, a country with
an erratic president and an anti-American orientation.
There is enough weapons-grade uranium for a bomb or two in
Kharkiv, in Ukraine. Outside of Belgrade, in a research
reactor at Vinca, sits sufficient material for a bomb --
and there it sat while NATO was bombarding the area.

''We need to avoid the notion that because the most
material is in Russia, that's where we should direct all of
our effort,'' Potter said. ''It's like assuming the bank
robber will target Fort Knox because that's where the most
gold is. The bank robber goes where the gold is most
accessible.''

Weapons of Mass Disruption The first and, so far, only
consummated act of nuclear terrorism took place in Moscow
in 1995, and it was scarcely memorable. Chechen rebels
obtained a canister of cesium, possibly from a hospital
they had commandeered a few months before. They hid it in a
Moscow park famed for its weekend flea market and called
the press. No one was hurt. Authorities treated the
incident discreetly, and a surge of panic quickly passed.

The story came up in virtually every conversation I had in
Russia about nuclear terror, usually to illustrate that
even without splitting atoms and making mushroom clouds a
terrorist could use radioactivity -- and the fear of it --
as a potent weapon.

The idea that you could make a fantastic weapon out of
radioactive material without actually producing a nuclear
bang has been around since the infancy of nuclear weaponry.
During World War II, American scientists in the Manhattan
Project worried that the Germans would rain radioactive
material on our troops storming the beaches on D-Day.
Robert S. Norris, the biographer of the Manhattan Project
director, Gen. Leslie R. Groves, told me that the United
States took this threat seriously enough to outfit some of
the D-Day soldiers with Geiger counters.

No country today includes radiological weapons in its
armories. But radiation's limitations as a military tool --
its tendency to drift afield with unplanned consequences,
its long-term rather than short-term lethality -- would not
necessarily count against it in the mind of a terrorist. If
your aim is to instill fear, radiation is anthrax-plus. And
unlike the fabrication of a nuclear explosive, this is
terror within the means of a soloist.

That is why, if you polled the universe of people paid to
worry about weapons of mass destruction (W.M.D., in the
jargon), you would find a general agreement that this is
probably the first thing we'll see. ''If there is a W.M.D.
attack in the next year, it's likely to be a radiological
attack,'' said Rose Gottemoeller, who handled Russian
nuclear safety in the Clinton administration and now
follows the subject for the Carnegie Endowment. The
radioactive heart of a dirty bomb could be spent fuel from
a nuclear reactor or isotopes separated out in the process
of refining nuclear fuel. These materials are many times
more abundant and much, much less protected than the
high-grade stuff suitable for bombs. Since Sept.11, Russian
officials have begun lobbying hard to expand the program of
American aid to include protection of these lower-grade
materials, and the Bush administration has earmarked a few
million dollars to study the problem. But the fact is that
radioactive material suitable for terrorist attacks is so
widely available that there is little hope of controlling
it all.

The guts of a dirty bomb could be cobalt-60, which is
readily available in hospitals for use in radiation therapy
and in food processing to kill the bacteria in fruits and
vegetables. It could be cesium-137, commonly used in
medical gauges and radiotherapy machines. It could be
americium, an isotope that behaves a lot like plutonium and
is used in smoke detectors and in oil prospecting. It could
be plutonium, which exists in many research laboratories in
America. If you trust the security of those American labs,
pause and reflect that the investigation into the great
anthrax scare seems to be focused on disaffected American
scientists.

Back in 1974, Theodore Taylor and Mason Willrich, in a book
on the dangers of nuclear theft, examined things a
terrorist might do if he got his hands on 100 grams of
plutonium -- a thimble-size amount. They calculated that a
killer who dissolved it, made an aerosol and introduced it
into the ventilation system of an office building could
deliver a lethal dose to the entire floor area of a large
skyscraper. But plutonium dispersed outdoors in the open
air, they estimated, would be far less effective. It would
blow away in a gentle wind.

The Federation of American Scientists recently mapped out
for a Congressional hearing the consequences of various
homemade dirty bombs detonated in New York or Washington.
For example, a bomb made with a single footlong pencil of
cobalt from a food irradiation plant and just 10 pounds of
TNT and detonated at Union Square in a light wind would
send a plume of radiation drifting across three states.
Much of Manhattan would be as contaminated as the
permanently closed area around the Chernobyl nuclear plant.
Anyone living in Manhattan would have at least a 1-in-100
chance of dying from cancer caused by the radiation. An
area reaching deep into the Hudson Valley would, under
current Environmental Protection Agency standards, have to
be decontaminated or destroyed.

Frank von Hippel, the Princeton physicist, has reviewed the
data, and he pointed out that this is a bit less alarming
than it sounds. ''Your probability of dying of cancer in
your lifetime is already about 20 percent,'' he said.
''This would increase it to 20.1 percent. Would you abandon
a city for that? I doubt it.''

Indeed, some large portion of our fear of radiation is
irrational. And yet the fact that it's all in your mind is
little consolation if it's also in the minds of a large,
panicky population. If the actual effect of a radiation
bomb is that people clog the bridges out of town, swarm the
hospitals and refuse to return to live and work in a
contaminated place, then the impact is a good deal more
than psychological. To this day, there is bitter debate
about the actual health toll from the Chernobyl nuclear
accident. There are researchers who claim that the people
who evacuated are actually in worse health over all from
the trauma of relocation, than those who stayed put and
marinated in the residual radiation. But the fact is, large
swaths of developed land around the Chernobyl site still
lie abandoned, much of it bulldozed down to the subsoil.
The Hart Senate Office Building was closed for three months
by what was, in hindsight, our society's inclination to err
on the side of alarm.

There are measures the government can take to diminish the
dangers of a radiological weapon, and many of them are
getting more serious consideration. The Bush administration
has taken a lively new interest in radiation-detection
devices that might catch dirty-bomb materials in transit. A
White House official told me the administration's judgment
is that protecting the raw materials of radiological terror
is worth doing, but not at the expense of more catastrophic
threats.

''It's all over,'' he said. ''It's not a winning
proposition to say you can just lock all that up. And then,
a bomb is pretty darn easy to make. You don't have to be a
rocket scientist to figure about fertilizer and diesel
fuel.'' A big fertilizer bomb of the type Timothy McVeigh
used to kill 168 people in Oklahoma City, spiced with a
dose of cobalt or cesium, would not tax the skills of a
determined terrorist.

''It's likely to happen, I think, in our lifetime,'' the
official said. ''And it'll be like Oklahoma City plus the
Hart Office Building. Which is real bad, but it ain't the
World Trade Center.''

The Peril of Power Plants Every eight years or so the
security guards at each of the country's 103 nuclear power
stations and at national weapons labs can expect to be
attacked by federal agents armed with laser-tag rifles.
These mock terror exercises are played according to
elaborate rules, called the ''design basis threat,'' that
in the view of skeptics favor the defense. The attack teams
can include no more than three commandos. The largest
vehicle they are permitted is an S.U.V. They are allowed to
have an accomplice inside the plant, but only one. They are
not allowed to improvise. (The mock assailants at one
Department of Energy lab were ruled out of order because
they commandeered a wheelbarrow to cart off a load of dummy
plutonium.) The mock attacks are actually announced in
advance. Even playing by these rules, the attackers manage
with some regularity to penetrate to the heart of a nuclear
plant and damage the core. Representative Edward J. Markey,
a Massachusetts Democrat and something of a scourge of the
nuclear power industry, has recently identified a number of
shortcomings in the safeguards, including, apparently, lax
standards for clearing workers hired at power plants.

One of the most glaring lapses, which nuclear regulators
concede and have promised to fix, is that the design basis
threat does not contemplate the possibility of a hijacker
commandeering an airplane and diving it into a reactor. In
fact, the protections currently in place don't consider the
possibility that the terrorist might be willing, even
eager, to die in the act. The government assumes the
culprits would be caught while trying to get away.

A nuclear power plant is essentially a great inferno of
decaying radioactive material, kept under control by
coolant. Turning this device into a terrorist weapon would
require cutting off the coolant so the atomic furnace rages
out of control and, equally important, getting the
radioactive matter to disperse by an explosion or fire. (At
Three Mile Island, the coolant was cut off and the reactor
core melted down, generating vast quantities of radiation.
But the thick walls of the containment building kept the
contaminant from being released, so no one died.)

One way to accomplish both goals might be to fly a large
jetliner into the fortified building that holds the
reactor. Some experts say a jet engine would stand a good
chance of bursting the containment vessel, and the sheer
force of the crash might disable the cooling system --
rupturing the pipes and cutting off electricity that pumps
the water through the core. Before nearby residents had
begun to evacuate, you could have a meltdown that would
spew a volcano of radioactive isotopes into the air,
causing fatal radiation sickness for those exposed to high
doses and raising lifetime cancer rates for miles around.

This sort of attack is not as easy, by a long shot, as
hitting the World Trade Center. The reactor is a small,
low-lying target, often nestled near the conspicuous
cooling towers, which could be destroyed without great
harm. The reactor is encased in reinforced concrete several
feet thick, probably enough, the industry contends, to
withstand a crash. The pilot would have to be quite a
marksman, and somewhat lucky. A high wind would disperse
the fumes before they did great damage.

Invading a plant to produce a meltdown, even given the
record of those mock attacks, would be more complicated,
because law enforcement from many miles around would be on
the place quickly, and because breaching the containment
vessel is harder from within. Either invaders or a kamikaze
attacker could instead target the more poorly protected
cooling ponds, where used plutonium sits, encased in great
rods of zirconium alloy. This kind of sabotage would take
longer to generate radiation and would be far less lethal.

Discussion of this kind of potential radiological
terrorism is colored by passionate disagreements over
nuclear power itself. Thus the nuclear industry and its
rather tame regulators sometimes sound dismissive about the
vulnerability of the plants (although less so since
Sept.11), while those who regard nuclear power as
inherently evil tend to overstate the risks. It is hard to
sort fact from fear-mongering.

Nuclear regulators and the industry grumpily concede that
Sept. 11 requires a new estimate of their defenses, and
under prodding from Congress they are redrafting the
so-called design basis threat, the one plants are required
to defend against. A few members of Congress have proposed
installing ground-to-air missiles at nuclear plants, which
most experts think is a recipe for a disastrous mishap.

''Probably the only way to protect against someone flying
an aircraft into a nuclear power plant,'' said Steve Fetter
of the University of Maryland, ''is to keep hijackers out
of cockpits.''

Being Afraid For those who were absorbed by the subject of
nuclear terror before it became fashionable, the months
since the terror attacks have been, paradoxically, a time
of vindication. President Bush, whose first budget cut $100
million from the programs to protect Russian weapons and
material (never a popular program among conservative
Republicans), has become a convert. The administration has
made nuclear terror a priority, and it is getting plenty of
goading to keep it one. You can argue with their priorities
and their budgets, but it's hard to accuse anyone of
indifference. And resistance -- from scientists who don't
want security measures to impede their access to nuclear
research materials, from generals and counterintelligence
officials uneasy about having their bunkers inspected, from
nuclear regulators who worry about the cost of nuclear
power, from conservatives who don't want to subsidize the
Russians to do much of anything -- has become harder to
sustain. Intelligence gathering on nuclear material has
been abysmal, but it is now being upgraded; it is a hot
topic at meetings between American and foreign intelligence
services, and we can expect more numerous and more
sophisticated sting operations aimed at disrupting the
black market for nuclear materials. Putin, too, has taken
notice. Just before leaving to meet Bush in Crawford, Tex.,
in November, he summoned the head of the atomic energy
ministry to the Kremlin on a Saturday to discuss nuclear
security. The subject is now on the regular agenda when
Bush and Putin talk.

These efforts can reduce the danger but they cannot
neutralize the fear, particularly after we have been so
vividly reminded of the hostility some of the world feels
for us, and of our vulnerability.

Fear is personal. My own -- in part, because it's the one I
grew up with, the one that made me shiver through the Cuban
missile crisis and ''On the Beach'' -- is the horrible
magic of nuclear fission. A dirty bomb or an assault on a
nuclear power station, ghastly as that would be, feels to
me within the range of what we have survived. As the White
House official I spoke with said, it's basically Oklahoma
City plus the Hart Office Building. A nuclear explosion is
in a different realm of fears and would test the country in
ways we can scarcely imagine.

As I neared the end of this assignment, I asked Matthew
McKinzie, a staff scientist at the Natural Resources
Defense Council, to run a computer model of a one-kiloton
nuclear explosion in Times Square, half a block from my
office, on a nice spring workday. By the standards of
serious nuclear weaponry, one kiloton is a junk bomb,
hardly worthy of respect, a fifteenth the power of the bomb
over Hiroshima.

A couple of days later he e-mailed me the results, which I
combined with estimates of office workers and tourist
traffic in the area. The blast and searing heat would gut
buildings for a block in every direction, incinerating
pedestrians and crushing people at their desks. Let's say
20,000 dead in a matter of seconds. Beyond this, to a
distance of more than a quarter mile, anyone directly
exposed to the fireball would die a gruesome death from
radiation sickness within a day -- anyone, that is, who
survived the third-degree burns. This larger circle would
be populated by about a quarter million people on a
workday. Half a mile from the explosion, up at Rockefeller
Center and down at Macy's, unshielded onlookers would
expect a slower death from radiation. A mushroom cloud of
irradiated debris would blossom more than two miles into
the air, and then, 40 minutes later, highly lethal fallout
would begin drifting back to earth, showering injured
survivors and dooming rescue workers. The poison would ride
for 5 or 10 miles on the prevailing winds, deep into the
Bronx or Queens or New Jersey.

A terrorist who pulls off even such a small-bore nuclear
explosion will take us to a whole different territory of
dread from Sept. 11. It is the event that preoccupies those
who think about this for a living, a category I seem to
have joined.

''I think they're going to try,'' said the physicist David
Albright. ''I'm an optimist at heart. I think we can catch
them in time. If one goes off, I think we will survive. But
we won't be the same. It will affect us in a fundamental
way. And not for the better.''

Bill Keller is a Times columnist and a senior writer for
the magazine.