On Nuclear Terrorism

Using Murphy's Law Against Nuclear Terrorists

By William C. Potter

A review of On Nuclear Terrorism by Michael Levi

"Michael Levi's slender volume On Nuclear Terrorism is a valuable addition to the burgeoning literature on catastrophic terrorism. Unlike many recent studies, it neither hypes the nuclear threat nor discounts it. Instead, Levi sketches the obstacles a terrorist would need to overcome to successfully implement a nuclear attack and then discusses the panoply of means available to preclude that outcome..."

Click here to return to the full article.

1 comment:

Jonathan I Katz said...

1. ``Simple hand-held detectors'' possess neither the sensitivity nor the
discrimination to be useful against masses of HEU, even minimally shielded.
The Customs agent with his ``pager'' detector can find many things, most of
them harmless, but he will not find HEU or even Pu. There are promising
developments in detector technology, mostly stalled by DNDO's refusal to
support any R&D in this direction. One can only speculate about their
reasons (my favorite is that there is no bureaucratic reward for sponsoring
the development of a system that will probably find no real threats for
decades, in contrast to one that finds trifling hazards, mostly not terrorist,
that can support press releases).

2. The analogy to drug smuggling is poor, and leads us in the wrong direction
regarding development of counter-IND systems and intelligence. There is a
large market for drugs everywhere, with many millions of retail consumers and
a robust and redundant infrastructure of distributors at several levels.
Illegal drugs are fungible (can be traded for money) almost everywhere, and
anyone who cares to can easily find out where to go, either to buy or to sell.
In contrast, neither the buyer nor the supplier of illegal fissile material
knows how to find the other, who he is, or even if he exists (a rare event
in itself). There is no market, and nearly every apparent bidder or offerer
is actually conducting either a sting or a scam. This is true of most nuclear
weapons-related skills and components, not just fissile material.

The threat is a vertically integrated terrorist organization that does every
phase of the operation itself. Only basic technical training (think of all
the Pakistani, Iraqi and Iranian engineering students in the West, or AQK at
Urenco) is likely to have been provided outside the terrorist organization,
many years or decades before the attack and before the perpetrators could be
associated with extremist ideology. Hence Levi's ``Murphy's Law'' argument is
valid; the terrorists (unlike the drug economy) have no redundancy. Their
weakest link is probably getting the fissile material or bomb at all, making
border or checkpoint security secondary (painful to say, for I have been
involved in two technically ``sweet'' projects here).

An implication of the foregoing is that our greatest vulnerability is to a
group operating in the grey area between non-state (al Qaeda, Aum Shinrikyo,
etc.) and state organizations, supplied with fissile material from a state
stockpile in a deniable mode, or by a dissident faction within a government.
Think of a future ``Hezbollah'' (with a different name, though possibly the
same people) and a nuclear-armed Iran, or AQK and a hypothetical customer.
This points us to a countermeasure: ensure that forensics will permit such
material, either detonated or seized beforehand, to be traced to their source,
because state sponsors have assets (capital cities, industry and militaries
with known locations, etc.) that make them deterrable by the threat of

A longer-term countermeasure is to limit the proliferation of related,
generally unclassified, technical training. In other words, no Iranian
or Pakistani graduate students in nuclear engineering or several other
fields, and no future AQK working at Urenco. For a fuller discussion,
see my paper ``Limiting the Nuclear Club'' in Strategic Review XXII, 74--80
(Winter 1994). This defunct journal is hard to find, so I will send a copy
to anyone who asks.

Jonathan I. Katz
Professor of Physics
Washington University
St. Louis, Mo. 63130