100,000 DEAD

So this month we must waste much of our issue on a new hoax perpetrated by two of America's cruder brainwashing institutions, Rep. E. Markey (D-Mass.) and Prof. H. Kendalls Union of (allegedly) Concerned (so-called) Scientists.

"The most detailed government study of potential consequences of a nuclear accident of atomic power plants has concluded that the worst case death toll could exceed 100,000 persons. The new estimates, which surpass the worst case estimate of 3,300 early deaths contained in the last NRC study in 1975 come from the NRC by Sandia National Laboratories," begins the report by the Washington Post (1 Nov). All of these statements are false; they set the tone for the rest of this "revelation" by the ever-suckered dummies of the newspaper for which Janet Cook won a Pulitzer Prize.

First, what is the "worst case" of an accident? A despicable abuse of language that makes as much sense in probability theory as the four corners of a circle in geometry.

With unimportant exceptions, the severity of an accident keeps increasing without limit as the probability of that severity approaches zero. What, for example, is the worst-case accident that can be caused with a hairpin? There is no "worst case." Mathematically, there is a non-zero probability that the same hairpin can cause the deaths of 100,000 people (mesmerized, waiting in line for their turn, let us say), and the reason why this is not the worst case is that as the 100,000^th victim drops dead it may, with an even smaller probability, fall onto an innocent bystander, rupture his spleen, and raise the death ton to 100,001. And as the funeral procession approaches a banana peel...

In fairness, the "worst case" misnomer is part of the US nuclear industry jargon and not, for once, Prof. Kendall's concoction; however, unless the M.I.T. tolerates mathematical illiterates on its physics faculty, he must be familiar with the elements of probability theory and well aware of the misnomer that is being abused by his "scientists." (In the German industry, the term grobter annehmbarer Unfall, maximum assumable accident, makes eminent sense, for it is defined as the most severe accident with which a nuclear facility can still cope.)

For the same reason, the Washington Post's reference to the 1975 report, allegedly giving a "worst case estimate of 3,300 early deaths" is uncanny gibberish. At that death toll, the probability of such a nuclear accident has decreased to 0.0000001/year (one in 10 million reactor years), and since such a probability lies on the borders of mathematics and philosophy, the authors of the [1975] report chose to end the figure there, for it must end somewhere if the page is to be finite. The line continues, of course, into even smaller probabilities down to next to nothing at all. The trouble here seems to be that extending a straight line beyond the point where the picture ends taxes the top writers of the Washington Post well beyond their utmost power of imagination and intellectual capacity.