Impossible. The base of the Unit 1 chimney had already been poured, and 70 laboriously laid feet of vertical brickwork was already in place. If filtering were necessary, then it should have been designed in to begin with, in the gallery between the core and the chimney base. There was no way to tear the thing down and start over, and, at this late stage, the only place to put filters would be atop the chimneys. Cockcroft was adamant, and he was in command. The filter assemblies, aptly named “Cockcroft’s Follies,” were built as specified, using 200 tons of structural steel, concrete, and bricks hauled up 400 feet to the tops of the air stacks.
Several filter packs were tried. In early 1953, a temporary lash-up was settled on. The filters would be fiberglass tissue, sprayed with mineral oil to trap dust made radioactive by going through the reactor core and any fuel that might disintegrate and escape. There was much grumbling, not only among the men who had to retrofit this ghastly mess, but from Penney’s bomb makers. To make plutonium at the necessary rate, a ton of cooling air had to go up each of the two stacks every second. That meant that the hot air had to be moving at 2,000 feet per minute as it exited at the top, but at that point it would hit the drag-inducing filter pack like a brick wall. Delicate fiberglass tissue would quickly come apart, and the oil would be blown off the fibers. The air speed would have to be reduced, and this would mean less plutonium per month. The Canadians would have to pick up the slack using their heavy-water reactor at Chalk River.[122]
With further improvements and changes in the filter packs, including substituting silicon oil for the mineral oil, Cockcroft’s follies were preventing the occasional radioactive dust particle from ejecting into the atmosphere. An exception was the volatile fission product iodine-131. As a gas, it was largely unhindered by the filters. There was nothing to worry about. When one was striving to build an arsenal of weapons, one of which can obliterate a large city, it seemed almost inappropriate to fret over probabilistic health effects on individuals. At this point, production speed outweighed all but the most basic safety concerns.
On they labored. Engineering problems were knocked down one at a time. Of particular concern were the fuel canisters. The metallic uranium, which would catch fire spontaneously if exposed to air, had to be cast into small cylinders a few inches long and sealed up in aluminum cans with heliarc-welded seams. The hot uranium would react chemically with the aluminum, so the inside of each can was coated with insulating graphite. The uranium slug would expand as it heated and burst the can, so it had to be made to fit loosely. Each can was filled with pressurized helium to ensure heat conduction to the aluminum, and the assembly was covered with cooling fins. In early summer 1950, Cockcroft threw another serious monkey wrench into the business. New calculations seemed to indicate that the critical mass of Pile. No. 1 would have to be increased by as much as 250 percent, or it would not work.
Reeling briefly with this news, the design team proceeded to inventory every material in the reactor that would parasitically absorb neutrons, affecting the core reactivity. There was one item that could be reduced enough to compensate for the lack of critical mass: the fins on the fuel cartridges. In three weeks the team managed to trim one sixteenth of an inch off each of the one million cooling fins, thus solving the problem.
In October 1950, craftsmanship triumphed over knowledge, and the Windscale Pile No. 1 achieved self-sustaining fission. By January 1951, used fuel cartridges had been chemically processed, and Tom Tuohy, the Works General Manager, held the first lump of British plutonium in his hands. In June 1951, Pile No. 2 became operational, and plutonium production was proceeding at full capacity. The people of the United Kingdom under hard postwar circumstances had put their backs into it and prevailed.
There were problems. On May 7, 1952, there was an unexplained temperature rise in Pile No. 2, only in the upper portion of the core. The operators were able to bring the temperature down with the fans, but there was not a clue as to why this had happened. In September the same thing happened to Pile No. 1, but this time there was smoke coming out of the stack. Was the graphite burning? They had been warned by the Americans: Whatever you do, do not let the graphite catch fire. Once it gets going, water will not extinguish the fire. It will only make it burn hotter, and the graphite sucks the oxygen out of the water and leaves you with explosive hydrogen. Analysis showed that the smoke was caused by lubricating oil leaking from a fan bearing, but the high temperature remained inexplicable.
In May 1952, Pile No. 2 was taken down for maintenance. The workers found that 2,140 cartridges had migrated out of the fuel channels. Some were hanging precariously out the back of the core, and some had flown over the pool of water below and into the wall behind it. The air flowing from front to back had literally blown them out of the reactor. Each fuel cartridge had a little graphite boat under it, so that the cooling fins would not make it stick in the channel, but this would have to be changed. A worse finding was that the Burst Cartridge Detection Gear (BCDG) had punched a hole in the end of a cartridge, putting fission products in the air stream.
There were eight BCDGs, called “Christmas trees,” in each reactor. A Christmas tree was a rectangular matrix of 32 vacuum nozzles, arranged to fit over any 4-by-8 array of exit holes in the back of the reactor core. It could be positioned remotely from the control room to sniff the holes, looking for escaping radioactivity that would indicate a broken fuel can, but the operation was completely blind. From the control room or even from the face of the pile where fuel was pushed in, you could not see the BCDGs in action. During normal operation the eight of them would crawl over the back end of the pile, like very large insects, looking for trouble. It was a good idea to have equipment that could tell exactly which hole was leaking, but it could be too strident and cause what it was trying to detect, punching the vacuum nozzle through the aluminum. Some fuel cartridges were even jammed into the frames of the Christmas trees, never making it to the recovery pool below for plutonium extraction.
The American delegation made another informative visit, this time bringing the big guns, Dr. Edward Teller, the brilliant Hungarian theoretician and Mother of the H-bomb, and Dr. Gioacchino Failla from the Columbia University Radiological Research Laboratory. They could explain the mysterious temperature transients in Piles No. 1 and 2.[123] Technically, it was illegal to give such information to a foreign power, but the Brits had to be informed before their entire atomic bomb program went up in smoke.
There is another Wigner effect, called Wigner Energy. The same phenomenon that would bend the crystalline structure of graphite would store a great deal of potential energy in the material. If you allowed this energy to accumulate over a long period of time, it could reach a break-down point, at which it would all release at once, igniting the graphite with extremely high temperature and sending your plutonium production machinery up the chimney. The way to keep this from happening is to stage a periodic annealing, in which the core is heated to an abnormally high temperature, about 250 °C, using nuclear fission with reduced cooling. At this point a “Wigner release” should occur, with the graphite providing its own heating without the need for fission. Shut down the reactor, and adjust the fans to slowly bring down the temperature.
122
The Canadians did indeed supply plutonium for the time-critical atomic bomb project, but unlike the Windscale reactors, the NRX reactor at Chalk River was not designed specifically to make Pu-239. The Windscale pile operations would purposefully run the fuel through very quickly, whereas NRX would use the fuel to useful depletion. As a result, the fissile material for the bomb contained a lot of Pu-240, which fissions spontaneously and threatens to melt the bomb core before it can explode. The Brits solved this problem by using a “levitated pit,” in which the bomb core was suspended in the middle of a hollow sphere of high explosive, removing it from close contact with any neutron-reflecting material. It seemed a brilliant innovation, but the Americans had been using this design feature since 1948 in the MK-4 A-bomb, tested in Operation Sandstone.
123
Inside nuclear engineering, Dr. Teller is considered to be the mother of the H-bomb and not the father because he “took it to full term.”