It was 7:00 P.M. In the basement was a long, narrow room, having two 1,000-liter tanks bolted to the concrete floor. Four and a half feet above the concrete was built a raised floor with the tops of the two tanks protruding through it. The tanks were used to temporarily hold very dilute mixtures of plutonium salt and water originating upstairs. The pipes had been changed, and there was some resulting confusion as to what the tanks were connected to. Along the walls were two shelves. You entered the room by climbing seven wooden steps to the open doorway, and on the shelf to your left sat an unauthorized stainless steel bucket with two handles. It looked like a cookpot stolen from the cafeteria kitchen, probably used to make soup. It had no business being in the same room with plutonium extract. It could hold 60 liters of fluid.
They had tried two extracting solutions the previous day, and somehow the results of both experiments had wound up in tank 2, the one closer to the doorway. The shift supervisor thought it prudent to find out how much plutonium was sitting in the tank, and he instructed one of the operators to take a sample and have it analyzed. The operator lowered a small glass vial into the tank through a sensor port on top, filled it with fluid, and sent it upstairs to be evaluated.
The sample contained about 0.6 grams of plutonium per liter of fluid. The tank was nearly full, holding 800 liters of liquid. Supervisor made a quick calculation. That meant that there were 480 grams of plutonium in tank 2, which was a nearly ideal shape for criticality. The safety limit for the tank was 400 grams, and he found this puzzling. How could the tank have 480 grams of plutonium in it when 400 was the cutoff for criticality? He ordered operator to take two more samples and confirm it.
While lifting out another sample, operator noticed that the tank was not completely filled with the organic solutions, as they had feared. In fact, it was mostly filled with a weak solution of plutonium in water, with their experimental, oily, concentrated solutions having floated to the top and sitting in a thin disk-shaped layer in the tank. This explained why the tank had not gone supercritical when filled, and supervisor felt slightly better. There may have been enough plutonium in the organic layer to go critical, but only if it were shaped like the entire tank and not spread into a thin layer on top of the water. The process downstream of tank 2 was not set up to work on organic solvents, so they had better decant it off the top of the water before something happened. They found a 20-liter glass bottle, normally used to hold chemical reagents, a two-hole stopper, and two rubber hoses. Operator put the bottle on the shelf, connected the hoses to the stopper, and connected one hose to an active vacuum port on top of tank 1, which was completely empty. He connected the second hose to the stopper and lowered the open end of it into the wide access port on tank 2, being careful to let it suck fluid only from the thin layer of organic solvent on top. The oily stuff was dark brown, indicating that it was thoroughly loaded with plutonium.
Satisfied that all was well, the supervisor left his two operators with the task and went to see how the rest of the shift was performing. When the bottle looked about full, the operators pulled the stopper loose, hefted the 17 liters, and poured it into the big pot, still up on the shelf. One operator had to go to his other duties, and the other found the supervisor to ask him, what next? The supervisor told him to make sure it was all gone out of the tank. By being mindful of the depth of the end of the rubber hose, operator was able to suck out another 20 liters into the bottle, filling it to the top.
Operator pulled the stopper and started emptying the bottle into the big pot up on the shelf. The liquid, a mixture of two plutonium-extraction experiments, was thick, with globs in it. It poured slowly, but the bottle was gradually getting lighter and easier to hold. BANG! A flash of blue light filled the room, and the operator felt the sudden blast of heat hit him right in the face. He instinctively dropped the bottle. It crashed to the floor, sending shards of broken glass and dissolved plutonium scattering across the floor, out the door, dripping down the steps, and to the floor drain. The criticality alarm in the building went off immediately, along with the criticality alarm in the next building, 50 yards away. Everybody in both buildings dropped what they were doing and hastened for the escape tunnels.
The building radiation-control supervisor was the only one not leaving. He switched into emergency mode and made sure that everyone surrendered his or her personal dosimeter before getting away. He then ran into the operator, looking frightened and in shock, and he directed the injured man to decontamination and medical care. Everyone was ordered to not go back into the building until the reason for the criticality alarm could be determined and made safe.
The oily mixture in the stainless steel pot had heated instantly to the boiling point upon going critical, and the resulting thermal expansion turned off the chain reaction. The supercritical condition in these very small, unplanned reactors teeters on a knife-edge, and the slightest modification of the density, the total mass, or the shape of the reactor can shut it down as quickly as it came into being. There it sat, undisturbed, until 11:30 P.M., when, having cooled, it suddenly lapsed again into criticality and faded out when the liquid re-heated. This time, the reaction was too weak to set off the criticality alarm, and nobody was there to be harmed by the mixed radiation pulse.
The shift supervisor was feeling dread about how completely he had botched his experimental program. He had a strong urge to erase what he had done, resetting the situation to normal conditions, but to set everything right he had to get back into the room. The radiation supervisor would hear none of it, but the shift supervisor was adamant. Finally, radiation agreed to follow shift to the area and scope out the extent of the contamination. As they neared the doorway to the tank room, the rate meter on the gamma-sensitive “cutie pie” instrument slid off scale. The room was hot with fission products, and it would be crazy to go in there. Wait for the clean-up team to dress out in radiation suits and come up with a plan to disable the reactor.
Shift supervisor was too impatient for that. Somehow, he talked the radiation supervisor into leaving him standing in front of the room while he went to check something.[179] Shift supervisor, seeing radiation supervisor turn a corner, raced up the steps into the tank room. He saw the pot on the shelf and quickly scoped the problem. The pot was cooling down, and he had to do something very soon, before it had a chance to go critical again. He took the handles in both hands and lifted the pot, planning to dump its contents down the steps and into the floor drain. The plutonium mixture would be so spread out into a thin puddle, it could not possibly regain criticality. The thing was a lot heavier than it looked, and he managed a controlled fall to the floor. It hit with a wet thud, right in the middle of the puddle of plutonium solution.
This time, the supercriticality pulsed like it meant it. As the barely subcritical pot hit the floor, its tendency to fission was extended by the flat field of plutonium-239-bearing solvent now under it. Not only were neutrons reflected back into the pot, they were multiplied by causing fissions outside the critical mass, giving back as many as two neutrons for every one lost by leakage from the surface of the pot. Alarms in both buildings went off again, and supervisor was drenched in fluid as the reactor boiled explosively. Supervisor staggered down the steps and made it to decontamination. He had absorbed 2,450 rem of mixed radiation, and he was a dead man walking.[180]
179
The existing record of what happened at Mayak was given to the team at Los Alamos compiling
180
We may never know his name, but the night shift supervisor at the Mayak plutonium extraction building was awarded the not-coveted 1994 Darwin Award. The Darwin is given to that individual who has taken his or her (usually his) self out of the gene pool by doing something really stupid, therefore proving that evolution works by not allowing people who should not reproduce to do so. http://darwinawards.com/darwin/darwin1994-25.html