Zinn’s concept was a reactor that burned the artificial nuclide Pu-239 to make power. Manufactured plutonium was almost 100 percent pure, containing no worthless isotopes, and it did not need the advantage of thermalized neutrons to fission efficiently in a fast-neutron environment. The fission process in pure Pu-239 was not marginal and barely able to make it, as was the natural uranium fission scheme. The plutonium reactor core, bleeding fast neutrons from every external surface, would be surrounded by a blanket of pure U-238. Wasted fast neutrons escaping the fission process would be captured in the uranium shield, converting in a few days into more Pu-239. Calculations showed that such a pile would produce more Pu-239 than it burned in the fission process. It seemed utopian. It was a power source that produced more fuel than it consumed. There was enough U-238 piling up from the weapons production to provide the total energy needs of civilization for thousands of years into the future, without drawing down the weapons material or the submarine fuel. This gorgeous concept would be known forevermore as the “fast breeder reactor.” There were some minor complications to be worked out.[94]
A power reactor must have some means of transferring energy from the fission process to the power-generation process. In Rickover’s submarine reactor, being assembled for testing just over the horizon from the EBR-I site, the medium of transfer was water. The water absorbed heat from the fission by direct contact with the fission neutrons, which were born going 44 million miles per hour. Zinn’s breeder had to be different. There would be no neutron moderation, no slowing them down to thermal speed. The coolant in the breeder had to be heavy, metallic, and liquefied, so that it could absorb heat from the reaction by conduction to metal surfaces in the reactor structure without slowing anything down. Neutrons hitting heavy metal nuclei would simply bounce off and retain their speed. Both transfer methods, water and metal, kept energy from accumulating in the machine and causing it to melt. The fluid medium would flow into the bottom of the reactor core, picking up heat and exiting through a pipe at the top. For the EBR-I, the logical choice of coolant was a mixture of sodium and potassium metals, pronounced “nack.”
By mixing sodium with potassium, the melting point of the metal could be brought down to about room temperature. It was a very sluggish fluid, but it would flow, particularly when heated to 900 degrees Fahrenheit with the reactor at full power. Unlike water it was, of course, completely opaque. There was no way to visually inspect the reactor internal structures with the coolant in it, the way it was possible using water moderator. The sodium tended to activate under the neutron bombardment of fission, becoming sodium-24, throwing beta and gamma radiation with a 15-hour half-life. The potassium was a lesser activation risk, but it would contribute some to the induced radioactivity. The worst characteristic of nack is that it would react vigorously with air, water vapor, or particularly liquid water, becoming sodium and potassium hydroxides. In their undiluted form, these are nasty chemicals. Step in a puddle of it, and it will first dissolve your heavy leather boot followed by the foot inside the boot. Anything aluminum touched by it, like an airplane or a Pullman railway car, turns to white powder. There were entire reactors made of aluminum, but this one would have to be stainless steel.
Zinn found that plutonium-239 was impossible to obtain for reactor experiments. By 1951 the AEC was turning out plutonium-based implosion bombs on an industrial scale, and surplus material was simply unavailable at the time. It was easier to obtain 97-percent enriched uranium-235 for the EBR-I fuel, and it was an adequate substitute for the plutonium.[95] At Argonne, the 52 kilograms of bomb-grade uranium was fabricated into 179 stainless steel-clad fuel rods with some spares, made pencil-thin for good heat conductivity to the nack. The rods were mounted vertically in a steel tank, separated from each other in the preferred “hexagonal pitch” configuration so that the coolant could flow through them from bottom to top.
Control of a fast reactor is a bit touchy, because a smaller percentage of the fission neutrons are delayed than in the usual thermal reactors. With most of the neutrons born instantly upon fission and no slowing-down time allowed, the smoothness of the controls is diminished. The fission in EBR-I was managed using boron neutron-absorption rods, moved up and down into the reactor core by electric motors. To shut it down quickly, you could bring your palm down on the red scram button and operate the motors at maximum in-speed, or you could hit the reflector release button. Under this condition, the bottom half of the breeding blanket would unlatch and fall away by gravity with a floor-shaking CLUMP. Just losing the blanket and its tendency to reflect some neutrons back into the core was enough to kill the fission process and make the reactor quickly subcritical. In the middle of the control room there was also an alternate scram control. It was a triangular metal handle, hanging by a chain. Anyone in the room could pull that handle at any time and shut down the pile instantly.
In late May 1951, the experiment was ready for a power-up, and dignified guests and visitors, having endured a dreary ride out to the site, were eagerly anticipating success. With all the controls carefully pulled out, the reactor remained serene and subcritical. Zinn remained calm and estimated that the core was so completely dead, he would need another 7.5 kilograms of fuel to make it work. With a hint of reluctance the AEC granted him some more U-235 from military stockpiles, and it took three months to get it fabricated into fuel.
On August 24, 1951, the world’s first fast breeder reactor was the first reactor at NRTS to go critical. At zero power it worked fine, but measurements indicated that the fuel assemblies would have to be modified if the thing was going to boil water and make steam, as was necessary for a power reactor. Two thirds of the fuel rods were shipped back to Argonne, where they were made shorter and wider by a hydraulic press. Arriving back in Idaho, they were arranged as a belt-line around the thinner fuel tubes.
On December 20, the reactor was run up to about 45 kilowatts. The hot metal coolant was pumped continuously through a heat exchanger with water on the other side of the heat transfer process. At 1:23 in the afternoon, an operator opened a steam valve connected to the heat exchanger, and steam ran through the power turbine. Four 200-watt light bulbs on the turbine deck began to glow white-hot. At that instant, electrical power was first generated by a nuclear reactor.[96] Zinn took a piece of chalk, wrote his name on the concrete wall of the building, and invited his entire team to do the same. Those names remain on that wall to this day.
94
Recently the phrase “fast breeder reactor” has been changed to “integral fast reactor.” I’m not sure why. A sign on Highway 20 leading into the NRTS cautioned travelers, “warning: do not disturb breeding reactors.” It was a real knee-slapper at the time.
95
The use of plutonium in a fast-neutron reactor had already been proven. In 1945, an experimental reactor was built at Los Alamos in a deep canyon, far removed from other activities in case it misbehaved. At the time, the code-name for plutonium was “49,” and the scientists involved in the project were called “49ers.” You can probably guess the name of the reactor: Clementine. It ran experimentally at a maximum power of 25 kilowatts for six years, until a fuel-cladding failure shut it down for good. The dangerous coolant was pure mercury, which would enthusiastically activate into radioactive Hg-203. The purpose of Clementine was not to advance power-reactor technology, but to test bomb materials under the neutron-soak expected in a nuclear explosion.
96
So it is written in the official history of the world, but nothing is that simple. On September 3, 1948, Logan Emlet opened the valve on a toy steam engine connected to a water pipe running through the air-cooled X-10 graphite reactor at Oak Ridge, Tennessee. The reactor, running at one megawatt, was making steam in the pipe, and the little engine started puffing. The flywheel was belt-driving a classroom demonstration generator, which was connected to a 1/3-watt flashlight bulb. It wasn’t much light, but it was definitely the first. Everything done on the X-10 was still secret, so this historical generation of electrical power by nuclear means was buried. Buried deeper was the first use of a reactor to heat a building, which occurred in England in the fall of 1948. The reactor was named “Bepo,” meant to sound like an unknown Marx brother. I’m not making this up.