A Primer on Nuclear Safety: 1.2.2 Heat, Water and Zirconium

1.2.2 Hear, Water and Zirconium

The United Sttes Navy had been buildig sips with boilers when Hyman Rickover arrived in Oak Ridge. They United States Nay is a very proud orgabization, and not the leastcause of its pride was its boiler technology. Hyman Rickover and his team were imerced in the Navy's boiler technology tradition.

Their challenge was to find a way to to transform Weinberg's tiny water moderated e Materials Testing Reactor mock-up into a  reactor powerful enough to power a submarine.

Before I discuss this challenge I need to explain a little more about Wienberg's toy reactor, the Low Intensity Test Reactor, which was what the original mock-up was called when it was converted into a real reactor.  Although the 1.7 billion year old Oklo mine reactors had not been discovered yet, Weinberg and company had designed a reactor that mimicked the major features of the Oklo reactors. Like the OKLO reactors, the Low Intensity Test Reactor contained uranium with a high enough U-235 ratio, that it would go critical in the presence of water. Thus far from being unnatural, Weinberg's little water cooled reactor unwittingly emulated nature. In both the Oklo reactor and Weinberg's toy reactor, water slows stray neutrons down enough to promote a chain reaction with fuel that had only modestly U-235 to U-238 ratios. This slowing down of neutrons is called moderation. Because the toy reactor generated so little heat, it was fueled by aluminum clad uranium metal plates.

So the Rickover's first step was to use a water moderator. The second step was to produce enough power to boil water. That meant that Rickover's reactors had to be more powerful in order to produce much more heat. But here Rickover encountered a potential road block. The uranium used in a reactor had to be kept separate from the water for a number of reasons. One was that the uranium would corrode. Also radioactive fission products could escape from the uranium into the the cooling-moderating water. This was true whether the uranium was in metal form, or if it had been oxidized and baked into a ceramic, and had undesirable consequences. Rickover's ship board reactor clearly required that the Uranium fuel have a cladding and safety was no small consideration.

Aluminum had some good features, and had been the preferred cladding material, but it had a corrosion problem at higher temperatures, and it was was too weak a metal to be be rupture proof in a high performance reactor. Aluninum only maintain strength up to 300°C, and melt at 650°C. This proved safe in Weinbergs toy reactor, but the short comings of aluminum as a fuel cladding were dramatically demonstrated over and over, beginning in 1952 during an accident in the Canadian water cooled NRX reactor ruptured. Due to an operator error, control rods were accidentally lifted. Further errors occurred when operators attempted to shut down the reactor. Shut-off rods failed to fully descend into the reactor core. Reactivity and heat shot up. Aluminum clad fuel elements ruptured. Hydrogen gas was generated, and exploded. The lid of an inner reactor containment dome was blown off, and radioactive materials leaked into the environment. Jimmy Carter, then one of Rickover's boys, participated in the clean up.

Failure of alunium fuel cladding in the Windscale reactor played a major role in the 1957 fire that released a great deal of radioactive material into the environment. An accident with the Canadian NRU reactor in 1958 further illustrated the dangers of aluminum cladding. Several aluminum clad fuel elements inside the reactor over heated and burst, One caught on fire. While being extracted from the reactor it was torn in two. The fragments fell into a pit, and continued to burn. The resulting release of radioactive materials created quite a mess.

Clearly then aluminum cladding would not be up to snuff as fuel cladding for Rickover's reactors.

Alvin Weinberg suggested to Rickover that Zirconium had potential as a reactor fuel cladding. Zirconium is stronger than aluminum, had good heat transfer properties, can tolerate 1,800° C heat and does not corrode in the presence of water and heat. There was one rub. Radiation testing of zirconium samples show them to be a neutron poison. Using zarconium as a reactor fuel element cladding would kill a chain reaction, or so it was thought. But Oak Ridge scientist Herbert Pomerance, Herb to those who cultivated his friendship, traced the problem to hafnium, a chemically similar element that was present as a natural contaminant in zirconium.

Thus if zirconium could be separated from the hafnium, it would give Rickover his safe reactor cladding. The task of industrial sale separation of zirconium from hafnium was given to a group of three Y-12 chemists under Warren Grimes. Grimes asked Lyle Overhoiser to work on the problem, but when Lyle needed some assistance from an analytic chemist, my father, C.J. Barton, Sr., was brought in to help. My father and Lyle had been Lab mates as chemistry graduate students at the University of Virginia, and they worked well together. At that point my father stopped being an analytic chemist, and became an industrial chemist. With the assistance of John W. Ramsey, George Parker, Cyrus Feldman, and many others, they were able to identify a method that could be scaled up to zirconium separation Industrial levels of production. While Rickover was said to be the father of the atomic submarine, Oak Ridge chemists, including my father, broke the prophylactic: for him.

Charles Julian Barton, Sr. at Y-12

I talked with my father, Charles Barton, Sr., yesterday about his ORNL career. He is not a good communicator. His speech as always been halting, and he does not organize his memories into well formed stories. I can see that Interviewing him will be a process, and that information will come out in little snippets. My father's view of the the importance of the information might not be the same as mine, or of histories. What I have learned so far:

My father views his work on the separation of Zirconium and Hafnium. Zirconium and Hafnium are "rare earths." They are chemically similar, and thus not easy to separate. About 1% to 3% of refined Zirconium is Hafnium. Inside a reactor Zirconium and Hafnium behave very differently. Zirconium has a low neutron cross section. That means it is unlikely to capture neutrons inside a reactor. Capturing neutrons slows down or even stops chain reactions. Zirconium also resists corrosion. This makes it an ideal metal to use inside a reactors, especially as a cladding for fuel elements in light water reactors. Hafnium is has a high neutron cross section. It is 600 times more likely than Zirconium to capture neutrons inside a reactor, and unless separated from Zirconium, will poison chain reactions. Hafnium is also used inside reactors as control rods.

The chemical, and metallurgical properties of Zirconium made it an ideal material for light water reactors. During the 1940's the Navy saw that reactors could revolutionize the propulsion of submarines. They looked at two designs, one using sodium as a coolant. The history of sodium cooled reactors has always been a troubled one, and the Navy did not master the technology. The second naval reactor concept, patented by Alvin Weinberg, was the light water reactor. Pure zirconium was needed in order to get good performance from the light water reactors. Thus the development of both the atomic submarine and civilian light water reactors became possible. Today 85% of the world's commercial reactors are light water reactors that use Zarconium fuel cladding.

My father's first job at Y-12 in 1948, was to work along with Lyle Overholser, and J.W. Ramsey, to develop an industrial process for separating Zirconium and Hafnium. Lyle and my father had been a PhD students together at the University of Virginia in the 1930's. J.W. Ramsey was the father of a long time friend Jim Ramsey. Previous literature reported the use of ether . But the volitility of ether made it difficult to work with. My father and Lyle Overholser tried various organic solvents with little sucess. Then one day Ramsey showed up with a jug of hexone, and suggested that they try it. The hexone workes well, and the hexone process is still used for seperation in the United States. The separation process turned out well, and the light water became the corner stone of the first nuclear age. The names of L.B. Overholser, C.J. Barton, Sr., and J.W. Ramsey are on the patent. The patent describes the separation process:

The separation of hafnium impurities from zirconium can be accomplished by means of organic solvent extraction. The hafnium-containing zirconium feed material is dissolved in an aqueous chloride solution and the resulting solution is contacted with an organic hexone phase, with at least one of the phases containing thiocyanate. The hafnium is extracted into the organic phase while zirconium remains in the aqueous phase. Further recovery of zirconium is effected by stripping the onganic phase with a hydrochloric acid solution and commingling the resulting strip solution with the aqueous feed solution. Hexone is recovered and recycled by means of scrubbing the onganic phase with a sulfuric acid solution to remove the hafnium, and thiocyanate is recovered and recycled by means of neutralizing the effluent streams to obtain ammonium thiocyanate.

The History of ORNL states:

"Herbert Pomerance later that year discovered that zirconium's capability for neutron absorption had been vastly overstated because of its contamination by the element hafnium, which had a much greater poisoning effect.

Zirconium minerals have traces of hafnium, whose chemical characteristics are nearly identical to zirconium's, making economical separation of the two difficult. With funding from Captain Rickover and the Navy, laboratory researchers across the country investigated ways to separate the two elements. In 1949, chemical technologists at the Y-12 Plant, under the direction of Warren Grimes, developed a successful separation technique and scaled it to production level under the direction of Clarence Larson, then superintendent of the Y-12 Plant.

Zirconium alloys became essential first to the Navy's reactors and later to commercial power reactors. Zirconium rods filled with uranium pellets made up the fuel cores of nearly all light-water reactors, and hafnium was used in the control rods to regulate nuclear reactions. "

After the industrial facilities for purifying Zirconium were established at Y-12, the Y-12 Chemistry group was transfered administratively to ORNL. My father was moved to X-10 to work on the aqueous homogeneous reactor. Although little known now, the Aqueous homogeneous reactor was quite successful. It might have received a great deal more attention had not ORNL been also developing an even more promising concept, the Molten Salt Reactor.

At least one ORNL technical report reflects my father's aqueous homogeneous reactor research,
PHASE STABILITY OF HOMOGENEOUS REACTOR HOT FUEL SOLUTIONS. He was the lead writer along with J.S. Gill, GM Habert, WL Marshall, and RE Moore.

The History of ORNL reports:

"In 1952 the Lab built a small (1-megawatt) ``homogeneous'' reactor, one in which a liquid uranium solution was used both as fuel and as the source of steam to spin a generator's turbine. Besides offering potentially higher generating efficiencies than solid-fuel designs, it offered and important operation advantage: Its fuel solution could be routed continuously through a processing plant for purification and replenishment so the reactor would not require shutdowns for refueling. In 1957 ORNL built a larger homogeneous reactor, one modified to irradiate thorium and ``breed'' uranium while it generated power. But by then work on a solid-fuel breeder was well under way, and the AEC soon abandoned the liquid-fuel alternative."

The next post on my father's ORNL career will deal with his role in the development of Molten Salt Reactors.