I apologize for the recent hiatus in my postings here. I've been in a transition in my day job and it has definitely put a crimp in my blogging... Hopefully I'll be a little more consistent in the coming months...
My recent postings have focused on the Five Imperatives of Nuclear Power. Now it's time for the Fifth Imperative: Assure the deployment of nuclear power does not result in the proliferation of nuclear weapons.
This topic has been a subject of heated debate since the inception of commercial nuclear power in the 1950's and early 1960's. President Eisenhower faced this challenge during his "Atoms For Peace" campaign.
It is widely understood there are three paths to attaining nuclear weapons. The proliferating entity can either attain the uranium enrichment capability needed to produce weapons-grade uranium; attain the nuclear fuel, reactor, and nuclear fuel reprocessing capability needed to produce weapons-grade plutonium; or simply steal the required uranium or plutonium. History suggests both production paths have been successfully deployed by proliferating entities. As far as we know, no one has stolen sufficient material to become a member of the nuclear weapons club.
I will have additional postings on this topic in the future. The subject I want to address here is the often-used term "proliferation risk". I don't like the term because I feel it is technically imprecise at best, and terribly mis-leading at worst.
The term "risk" has a precise engineering definition. It is the sum, over all event paths and outcomes, of the product of the probability of a contributing event, and the consequence of that event.
Risk = Summation over all events of (Event Probability X Event Consequence).
We talk about the risk of smoking as "expected cancer deaths" or "expected cancers". If someone wishes to discuss the "risk of an automobile accident", one must first define the term "automobile accident", all of the accidents of interest, all of the events that lead to these accidents, and the probabilities of each of these events.
The first problem with apply this risk terminology to nuclear proliferation is that no on has ever been able to precisely define "proliferation" beyond simply "obtaining a nuclear weapon". In order to usefully apply the risk equation, one must be able to deconvolve "proliferation" into its constitute chain of events. Is enriching a gram of uranium "proliferation"? Is losing a gram of uranium or plutonium "proliferation"? Loosing 1 kg ? Think about all of the events that must occur for uranium ore in the ground to become uranium in a weapon. Get the point?
The second stumbling block in applying the risk equation to nuclear proliferation is the "probability" of "proliferation" is almost completely dominated by human will and intent. Indeed, there are two components of this probability: the probability someone will attempt to divert a technology, and the probability they will succeed in doing so. In practice, advocates of the "proliferation risk" vocabulary typically assume the magnitude of the first "probability" is unity (1.0), and then move on to the rest of the story – often without any useful definition of the chain of events under consideration.
It is true some technologies are easier to divert for clandestine purposes that others. And it is possible to build technical barriers and "self-reporting" technologies into nuclear fuels, reactors, and reprocessing facilities. But given an infinitely evil, infinitely rich, infinitely intelligent, and infinitely wealthy adversary, it's probably impossible to design a system that is "proliferation-proof". (And of course, the other challenge is that many of the technologies that enhance the proliferation resistance of nuclear technologies also make them more expensive for legitimate energy production purposes.)
So, enough musings for this posting. The topic of nuclear power and proliferation is very complicated, and men and women of good will can disagree strongly on many aspects of the debate.