Memo to File
February 22, 1999

Q:  Does LANL think that accidents resulting from contained vessel experiments with plutonium are so unlikely as to be effectively impossible?
A:  No -- or at least they didn't think so during the Cold War.

From a memo from Roland Jalbert to Dean Meyer, Group Leader, H-1, 1/22/70, entitled "PLUTONIUM HAZARD EVALUATION FOR GMX-11 CONFINEMENT EXPERIMENTS:"

The proposed continuation of the GMX-11 confinement experiments involving explosively driven plutonium-239 has been reevaluated in the light of recent experimental work and hazard analyses and the larger amounts of plutonium and high explosive that are anticipated compared to those that were originally planned.

A serious release of plutonium would take place in the wake of a major failure of the confinement vessel which would not be contained by the safety vessel.  Most of the plutonium, it is believed, would be deposited in chunks and smaller particles within a radius of a few hundred meters.  However, a substantial fraction of the total amount would be of respirable size (10 microns or less) or somewhat larger and be carried downwind to be deposited as fallout or to expose those in the path of the cloud...[emphasis added]

This memorandum calculates (much too conservatively, by today's analyses) that, under stable air conditions with wind from the south, each downwind person standing at Trinity Drive would receive 469 rems of committed lifetime lung dose per kilogram of Pu lost.  Under today's assumptions, about 23% of such a population would eventually die from cancer caused by this event.  Catastrophic failure, involving a loss of 3.5 kg Pu, leads to proportionally worse outcomes in the old H-1 analysis.  Fallout is also considered to be a serious risk:

...Thus, in the above worse case situation, decontamination of a part of the townsite would probably be required.  Areas of higher contamination would mostly be government land which are presently controlled or could be, if necessary, if decontamination proved ineffectual.

Sandia National Laboratories has examined the decontamination question in detail in a similar context (accidents with nuclear weapons), and has concluded that in many cases it would be cheaper to condemn and buy land wholesale rather than to attempt to clean it up.  (This voluminous report is available for examination at the Study Group.)

Source:  Roland A. Jalbert, H-1 to Dean D. Meyer, Group Leader, H-1, "Plutonium Hazard Evaluation for GMX-11 Confinement Experiments," LANL Memo, 1/22/1970.