Aside from the usual industrial accidents and hazards of using dangerous chemicals, working at a nuclear facility under war footing was remarkably safe. There were no radiation injuries.[38] However, there was a reactor explosion that destroyed a building. It was not in the Western Hemisphere, and, as would prove the case in many future nuclear accidents, the reactor was nowhere near running on fission. The problem involved water.
Werner Heisenberg, a respected German theoretical physicist, had made a name for himself well before the war started. He was famous for having expanded quantum mechanics with his uncertainty principle and his matrix spin operator, and he won the Nobel Prize in physics in 1932 for “the creation of quantum mechanics,” which was a bit overstated. With the German universities cleansed of nuclear talent by Nazi anti-Jew policies in the 1930s, Heisenberg, a Lutheran, was almost all that was left for mounting a nuclear weapon project. The necessary tasks were parsed and funded by the Reich Research Council. Nicholaus Kopermann was in charge of uranium production. Paul Harteck got heavy water production. Walther Bothe drew nuclear constants measurement, and Georg Stetter was given transuranic elements. Heisenberg was assigned the core problem, to prove the validity of the chain-reaction concept and then use the resulting nuclear reactor as a neutron source for further experimentation and data collection. Oddly, a separate uranium-enrichment task was spun off for Manfred von Ardenne, a German television pioneer, funded by the German Post Office.
Truth be known, Heisenberg was a brilliant theorist but not so good as an experimentalist, and his task involved building a nuclear reactor, which was heavy on the experimental side. He was grateful to be assigned Robert Döpel, professor of radiation physics at the University of Leipzig, to assist. It was Döpel and his wife, Klara, who decided that deuterium, the hydrogen isotope in “heavy water,” would be the ideal neutron moderator in a reactor using natural uranium. The construction of the first reactor, the L–I uranmaschine, was completed in August 1940. It was far subcritical, but it did accomplish neutron multiplication, producing more neutrons than were being injected from an external source, and it indicated that they were moving in the right direction. It would have to be rebuilt, larger, using heavy water, which was a precious material available sealed in 20-milliliter vials.
The Manhattan Project was doing basically the same thing in 1941, with a slightly different approach. Deuterium was indeed a fine neutron-moderating material, but, unlike the Third Reich, the United States had not captured a heavy-water-production plant in Norway. Instead, chemically pure synthetic graphite was used, delivered by the ton from Union Carbide. Enrico Fermi, a refugee nuclear scientist from Italy, headed the project, starting with a small pile of uranium and graphite in the corner of a lab at Columbia University. It was subcritical, but it multiplied the neutrons from a source. They were also going in the right direction, but they were one year behind the Germans.
By June 23, 1942, Heisenberg and Döpel had constructed L–IV, a bigger, more sophisticated version of their reactor in a dedicated laboratory building at the University of Leipzig. A large, circular pool of water was sunk into the middle of the floor in the lab. At the bottom was a frame, made of steel girders bolted together. Held off the bottom by the frame was a hollow sphere, one meter in diameter, of cast aluminum, three quarters submerged in the water. A flange around the circumference of the sphere was holding the upper and lower hemispheres together using 22 bolts. Four lifting lugs were cast into the flange with steel cables attached to a hoist above, and a long chimney emerged from the top, bolted to a flange on the upper hemisphere. On the inside surface of the sphere was a layer of uranium metal held in place by another, smaller flanged aluminum sphere. The inner space was filled with heavy water surrounding a still smaller sphere having another layer of uranium inside. A last aluminum sphere at the very center was filled with heavy water, and the chimney extended down through the center of it. Four neutron counters were arrayed on the top hemisphere.[39]
The plan was to lower a fixed neutron source consisting of a mixture of radium and beryllium powders down the chimney to the center of the reactor. The neutrons would be slowed by the heavy water and hit the first hollow sphere of uranium from all interior directions. High-speed neutrons from the fission reactions in the uranium would fly into the second layer of heavy water, slow down, and impinge on the outer layer of uranium, causing a chain reaction and sending a portion of the resulting neutron burst back through the heavy water and into the inner uranium shell. The water immersion in the pool was supposed to keep the assembly from melting when criticality was achieved. They could not have been overly optimistic, as they had no particular plan for what to do if the thing sprang to life as a supercritical reactor, with the heat exponentially increasing. The sphere was sealed up tightly, with a gasket separating the two halves, because the metallic uranium would react chemically with any water leaking in, jerking the oxygen right out of the H2O.
That morning, Döpel had noticed something odd about L–IV. Bubbles were coming out of the sphere and bursting on the surface of the water in the pool. They had been experimenting with it since June 3, and it had seemed complacent, even dull and unresponsive, but now it looked angry. As he stood and tried to figure out what was wrong, the bubbles stopped. Nothing to be concerned about. Döpel struck a match over the last bubble as it surfaced, and it popped with a bang. Yep. The gas leaking out was hydrogen, or it could be deuterium. Somehow, water was getting to the uranium.
After lunch, Döpel and Paschen, the lab mechanic, winched the thing out of the water and started to loosen the bolts. A gasket must have failed and it would require replacement. As soon as the seal broke, the sphere made a sudden hiss. A vacuum had developed inside, and air was rushing in. They stood frozen for a second. It was quiet, then suddenly flames started shooting out around the flange, followed by molten uranium, scattering all over the lab. Döpel doused it with water as Paschen tried to re-tighten the bolts, and the flames seemed to subside.
Heisenberg was summoned. He did not know exactly how to handle this situation. A nuclear reactor had never caught fire before. He ordered the ball to be lowered into the pool. At least that would cut off the oxygen and keep it cool. Nothing burns, he thought, under water. He left it to Döpel and went to the adjacent building to hold forth at his weekly nuclear physics seminar.
At about 6:00 Döpel barged in, saying “You must come at once!” Heisenberg spun around to upbraid him for interrupting, but he saw a look of cold terror in his face. “You’ve got to come look at the thing!”
They hastened to the lab, and Döpel pointed down into the central pool. Steam was rising from the sphere. It looked as if it were … expanding? It gave a little shudder. Both scientists spun in unison and lunged for the door. The L–IV exploded with a roar, sending flaming uranium against the 20-foot ceiling and setting the building on fire. For two days it burned, with no amount of effort able to extinguish the burning remnants of the reactor, and it finally settled down into a gurgling swamp of radioactive debris.
Although the project was supposed to be a secret, the explosion was not, and Heisenberg had to endure winks and hearty congratulations from associates on his success with his atomic bomb. By the time the story had leaked across the ocean to the United States, it had grown considerably. An entire room full of German scientists had perished in a nuclear weapon test. For Heisenberg, it was no success at all. The metallic uranium in their pathetically subcritical assembly had simply caught fire. It was a setback. By December, the Americans had caught up with the Germans and passed them with a self-sustaining chain reaction. Security was so tight, the Germans did not even know they had been beaten. At the end of the war, their only accomplishment had been the world’s first nuclear reactor accident, caused by water leaking past an inadequate gasket.
38
Well, maybe a few. On June 6, 1945, the exact critical mass of a uranium-235 bomb core was still unknown, and an experiment using 35.4 kg of 79.2 % enriched uranium metal was devised. Blocks of the uranium were built into a pseudosphere surrounded by paraffin. The assembly was placed in a large tank of water. As the tank was being filled, the thing went unexpectedly critical, and there was no way to scram it. Someone lunged for the drain valve, which was 15 feet away, and finally brought it under control. Three scientists received “significant” radiation exposure, but they seemed unscathed.
39
This incident is fairly well known, but there is some confusion about the exact configuration of the reactor. The uranium is often described as being in powdered form, which makes no sense. If it were powdered, the uranium metal would have reverted to uranium oxide before the sphere was loaded, and metallic uranium was thought to be essential. It was most likely uranium metal formed into marbles, so that they could be packed between aluminum shells. To machine hollow hemispheres of uranium would have been unnecessarily difficult. An alternate, more plausible configuration of L–IV is in a drawing captured by the Alsos Mission in 1945. It shows flat plates of alternating uranium metal and deuterated paraffin stacked inside the aluminum sphere. With ten plates of uranium (551 kilograms) and ten plates of paraffin, only the top hemisphere was filled. The plan was to load two plates at a time while monitoring the neutron multiplication, stopping when criticality was achieved.