Fear the attacker but don't panic

How easy would it be for terrorists to build a nuclear weapon? The first step would be to find enough nuclear explosive material - weapons-grade plutonium or highly enriched uranium (HEU).

According to the International Atomic Energy Agency, the "significant quantity" - that is, the amount necessary to construct a nuclear weapon - would be 8kg or 25kg respectively. This isn't a huge amount. There could be as much as 5 tonnes of separated civil plutonium in stockpiles around the world, with a further 250 tonnes of military plutonium. As for HEU, it can be found in over 130 HEU-fuelled civil nuclear research and test reactors around the world, in more than 40 countries. About 128 of these reactors and facilities are claimed to hold at least 20kg of HEU, with a global stockpile of non-military HEU amounting to some 200 tonnes. Security standards - particularly at non-military sites - have been known to be lax. Some analysts claim that as little as one tenth of the IAEA's "significant quantities" would be sufficient to make a bomb.

Having acquired enough uranium or plutonium, the next challenge would be to gather all the expertise, materials and technology necessary to build a working bomb. Some argue that a chain reaction could be achieved simply by dropping one lump of HEU on to another - a primitive "gun-type" device. Others claim that a more sophisticated effort would be required, involving reasonably well-educated nuclear physicists, electronics and explosives engineers, and a budget of a few hundred thousand pounds. Few if any analysts would claim that it would be impossible.

The answer to my initial question is that we don't really know how easy it would be for terrorists to build a nuclear bomb. Neither can we be sure why they should wish to do so. Would the object be a terrorist "spectacular" exceeding the carnage of 9/11 by a factor of ten or even a hundred? Or, as some suggest, are nuclear weapons being sought as a capability in waiting, to be revealed once the caliphate is restored in order to negotiate with the rest of the world on more equal terms?

With so much uncertainty surrounding the prospect of nuclear terrorism, there is a tendency - reasonable enough in the circumstances - to be consumed by worst-case analysis. But the problem with uncontrolled worst-case analysis, as Michael Levi argues so persuasively in On Nuclear Terrorism, is that the terrorist adversary becomes ten feet tall, omnicompetent and bound to succeed, while governments in the West become obsessed with this or that defensive measure in the expectation (or hope) that it will guarantee protection. The reality, claims Levi, is that the nuclear-inclined terrorist is likely to be about five feet tall, failure-averse, resource-limited and just as likely as anybody else to cock things up or succumb to Murphy's Law.

Levi contends that a comprehensive, systems-based approach is needed, integrating a wide range of nuclear and non-nuclear measures to multiply the chances of preventing terrorist success. Levi also calls for realism and honesty - particularly with the public. The prevention of nuclear terrorism cannot be guaranteed, and so it is essential to prepare for the gruesome consequences of such an attack. Effective consequence management will require participation of an informed public, and for that reason alone, On Nuclear Terrorism should have the widest possible readership.

Paul Cornish heads the International Security Programme at Chatham House and is the author of The CBRN System: Assessing the Threat of Terrorist Use of Chemical, Biological, Radiological and Nuclear Weapons in the United Kingdom (Chatham House, 2007).