The room was so crowded with continuous lines of glove boxes, the only way to get from one side of the room to the other was to go under the boxes. For this purpose, each line had a sloping valley dug out under it in the middle, where a person could stoop over slightly and get under the boxes without having to crawl.[163] The air in each assembly room was kept at below atmospheric pressure using blowers and racks of expensive HEPA filters. Air could therefore not leak out of the building and carry any plutonium oxide dust with it. Air could only leak into the building from outside, through imperfections in the airtight structure or when a door was opened. The world outside the building and certainly beyond the fence line was thus protected from plutonium contamination. Of even greater importance, no enemy could tell what was going on inside the buildings or tell how much was going on by reading the radiation signature at a distance. There was no radiation signature.
Plutonium was shipped to Rocky Flats from Hanford as liquid plutonium nitrate in small stainless steel flasks. Each flask, which held seven ounces of plutonium, was purposely isolated from all other flasks by putting each in the center of a cylinder the size of a truck tire. In Building 771, the flasks were emptied using a tube connected to a vacuum pump, and the liquid was transferred to a tall glass cylinder in a special glove box that looked very similar to the one that blew up at Hanford. Hydrogen peroxide was added to the plutonium nitrate, and a solid, plutonium peroxide or “green cake,” precipitated to the bottom. Moving through the glove-box line, the solid material was washed with alcohol, desiccated using a hair dryer, and pressed into 1.1-pound biscuits.
Continuing down the line, the biscuits were sent down the “chem line” to the G furnace, were they were baked into plutonium dioxide and mixed with hydrogen fluoride to make “pink cake,” or plutonium tetrafluoride. On they went on the conveyor to another furnace, where the pink biscuits were reduced to 10.5-ounce buttons of pure plutonium metal. In a day, Building 771 could produce 26.4 pounds of plutonium.[164] This raw material was then transferred to the fabrication line, where it was cast and machined into bomb parts, touched only with rubber gloves.
In 1955, a radical design change for atomic bomb cores came up from Los Alamos. Instead of a solid sphere of plutonium with a small cutout at the center for a modulated neutron source, the new cores would be thin, hollow spheres of alternating uranium-235 and plutonium-239. This design change would result in a lighter bomb that could be lobbed by a compact missile aboard a submarine, a ground-launched anti-aircraft missile, or an air-to-air missile on a fighter plane. It could also be boosted by injecting into the empty space in the core a mixture of deuterium and tritium gases, giving the atomic bomb a kick from hydrogen fusion. The management at Rocky Flats was delighted by the news and said that it should not take more than two years of construction and $21 million to make the necessary improvements to the physical plant.
The improvement schedule was fine, but in the meantime Rocky Flats would have to accommodate the new core design with whatever they could put together quickly. The timing was critical, because the Soviets seemed to be improving their arsenal at an alarming rate, and it was critically important to stay ahead of whatever they were doing. The new hollow core would require a lot more complicated machining and casting. There would be many times more machine cuttings and shavings to be recycled back, and larger lathes would be necessary. The fabrication room in Building 771 became extremely crowded as new box-lines were installed. It was hard to walk in the modified building with all the new glove boxes in the way, and the new lathe boxes were made of Plexiglas on all sides, instead of being a metal box with a Plexiglas window. The machinists had to be able to see the work from any angle, and to make a box out of leaded glass was too expensive. The Plexiglas could catch fire easily, and it was against AEC policy to use it in a plutonium line. The policy was waived under these emergency conditions, as it had been when the windows were installed in the original glove boxes.
With the time-critical, crowded work now at the factory, accidents were inevitable. It was June 1957, and the facility was still dealing with several times more plutonium than it was designed for.
The first problem occurred in the glass column used to mix plutonium nitrate with hydrogen peroxide. When you do too much of it too fast, oxygen builds up and pressurizes the column. One day without warning it exploded, blowing the side off the tall, metal glove box and wetting down two workers with plutonium nitrate. It was similar to the Atomic Man incident that happened two decades later at Hanford, but it was minor compared to the next accident. On September 11 the fabrication space, Room 180, in Building 771 caught fire. It was the result of production stress, exactly a month before the plutonium-conversion reactor at Windscale, England, caught fire for basically the same reason. Resolution of the Rocky Flats fire would be eerily similar to that of the Windscale disaster.
A box in the middle of the room was filled with leftover pieces of plutonium and machine shavings, contained in six steel cans and amounting to 22 pounds of fissile, flammable material. In one can was the result of a casting operation. The plutonium metal had been melted as usual in a hemispherical cast-iron crucible, shaped like a punch ladle. To make a plutonium casting, you melt it over a flame in the ladle, holding it by the long handle with the glove. When the metal goes liquid, you carefully pour it into the mold and let it cool and become solid. There is a thin film of plutonium left in the ladle. You peel it off and put it in a can for recycling. It is very thin and perfect for starting a fire. It is called a “skull.”
Just on its own, the skull caught fire at about 10:00 P.M. Plutonium burns with a brilliant white heat. Soon the all-Plexiglas glove box was on fire, and Room 180 was filled with smoke. Two security guards discovered it, seeing flames out of the glove boxes reaching for the ceiling. One went for a CO2 fire extinguisher, and the other called the Rocky Flats fire department.
By 10:15 P.M., the firemen under Verle “Lefty” Eminger were unreeling hose and going in, but Bob Vandegriff, the production supervisor, and Bruce Owen, the night radiation monitor, advised that no water be used. There were 137.5 pounds of plutonium in the room. Water covering it would amount to a neutron moderator, and the fissile plutonium could go critical and become a problem even larger than a raging conflagration. The two men quickly dressed down in Chemox breathing apparatus and went in with CO2 extinguishers, which they quickly emptied into the mass of flames. Nothing happened. They retreated, ran down the hall, and found the large fire extinguisher cart, dragged it into the room, which was now fully engaged, and opened the release valve. The firemen watched as the massive blast of carbon dioxide filled the room and engulfed the fire. It had absolutely no effect on the flames. Where was all the fresh air coming from to feed the blaze?
The 771 supervisor, Bud Venable, had been called at home at 10:23 P.M. and told that his building was on fire. Venable worried that the men going into Room 180 would suffocate or pass out in the heat. He ordered that the fans be turned up to the highest speed.
163
The workers named the underpass ravine “sheep dip.” There was no drain at the lowest point in the dip, and anything dropped on the floor would eventually end up there. Visiting physicists renamed this feature the “Lamb dip,” having to do with the spectral hole in the HeNe laser cavity at 1.1 microns. The humor went over most heads.
164
This was about enough to make two bombs. The old Mark I “Fat Man” used 13.6 pounds of plutonium in the core, but the newer designs used less plutonium, and some had U-235 components as well, reducing the amount of needed Pu-239, increasing the bomb yield, and reducing the weight.