A note on German uranium stores

Timothy Koeth and Miriam Hiebert’s fascinating detective work on the German uranium cubes (Physics Today, May 2019, page 36 ) sheds new light on Germany’s often overlooked wartime nuclear program.

The article prompted me to look at an aspect of the German work I did not consider while revising my History and Science of the Manhattan Project, Koeth and Hiebert’s reference 1: How did the amount of uranium the Germans had available compare with that used in Enrico Fermi’s Chicago Pile-1 (CP-1)? It turns out that the Germans had much less: Their 1064 natural uranium cubes would have had a total mass of about 2.5 tons; CP-1 incorporated about 5.6 tons of pure uranium metal and 37 tons of uranium oxide.

Much of Germany’s uranium was in the form of plates, most of which probably ended up in the US at Oak Ridge or Hanford. In his study of the German program, Mark Walker comments that Werner Heisenberg’s “large-scale” plate experiment, the B-VII pile, was planned to contain 3 tons of uranium metal, much less than a single full fuel load of the Oak Ridge X-10 reactor at about 106 tons or of one of the Hanford reactors at 255 tons. ¹ The cylinder for the German cube-based pile would have held about 1.7 tons of heavy-water moderator (less, if the volume of the cubes is accounted for); CP-1 boasted nearly 350 tons of graphite moderator. Those numbers drive home the immense difference in scale between the German and Allied programs.

A striking aspect of all the German pile experiments was their lack of any control mechanisms. We—and they—can be grateful that they did not succeed. I hope that Koeth and Hiebert’s article will lead to the discovery of more uranium cubes.