Thomas B. Cochran, Harold A. Feiveson, and Frank von Hippel

Immediately after the bombing of Pearl Harbor on December 2, 1942, research on plutonium production for atomic weapons was consolidated at the University of Chicago under Nobel Laureate Arthur H. Compton. The "Metallurgical Laboratory" (later to become Argonne National Laboratory) was the code name given to Compton’s facility. It was here that a small group of scientists, led by Enrico Fermi, built the world’s first reactor, Chicago Pile-1 (CP-1), which achieved initial criticality on 2 December 1942. During the next two years, work on the development of plutonium production reactors shifted to Oak Ridge and then Hanford. By early 1944, Compton and the Chicago scientists began thinking about the role of the Metallurgical Laboratory after the war.1

On the morning of April 26, 1944, Enrico Fermi, Leo Szilard, Eugene Wigner, Alvin Weinberg and others gathered to discuss the possibilities for using nuclear fission to heat and light cities.2 The scarcity of fissile material was on everyone’s mind. It was unclear at that time whether there was sufficient uranium even for producing highly enriched uranium and plutonium for a significant number of nuclear weapons. Fermi and his colleagues at the Metallurgical Laboratory therefore cast around for ways to produce maximum power — or plutonium for weapons — with minimal resources.3 They recognized that some reactor configurations might permit the conversion of uranium-238 to fissile (chain-reacting) plutonium at a rate faster than the fissile uranium-235 was consumed, hence the term "breeder reactor."

Walter Zinn, one of the nation’s few reactor experts and a close colleague of Fermi, was soon recruited to the cause.4 By summer of 1944 he had begun a more detailed investigation of breeder reactor designs. By the end of 1945, he had abandoned the idea of breeding uranium-233 in thorium and confirmed the original plan of breeding plutonium-239 from uranium-238 using fast fission neutrons.5 In 1945 Enrico Fermi said, "The country which first develops a breeder reactor will have a great competitive advantage in atomic energy."6

The world’s first fast-neutron reactor was Clementine, a 25 kilowatt thermal (KWt), mercury-cooled experimental reactor built at Omega Site (TA-2) at Los Alamos.7 It was proposed and approved in 1945. High intensities of fission-spectrum neutrons were needed by the bomb designers. Also, operation of the reactor would supply information about fast reactors that would be relevant to their possible use for production of power and fissile materials.8

Construction began in August 1946, criticality was achieved in late-1946, and full power in March 1949. The fuel was plutonium metal with natural uranium slugs at each end of the steel-clad rods. The rods were installed in a steel cage through which the liquid-mercury coolant flowed, driven by an electromagnetic pump. The core was surrounded concentrically with a 15 cm thick natural uranium reflector, a 15 cm thick steel reflector and a 10 cm thick lead shield.9

Clementine was shut down in March 1950 due to a control rod malfunction. Operations resumed in September 1952. It operated only until 24 December 1952, however, when a fuel rod ruptured. The uranium slugs swelled, burst the cladding and released plutonium into the mercury coolant.10 The reactor was subsequently dismantled.

After Clementine, Los Alamos developed and briefly operated one additional fast reactor, LAMPRE-I. This sodium-cooled reactor was fueled with molten plutonium. It achieved initial criticality in early-1961 and operated successfully for several thousand hours until mid-1963. Designed to explore issues associated with using plutonium fuel in fast breeder reactors, it was originally intended to operate at 20 megawatt thermal (MWt). It became apparent, however, that knowledge was inadequate about the behavior of some of the core materials in a high-temperature, high-radiation environment.11 The design power therefore was reduced to 1 MWt, with the plan to follow LAMPRE-I by a 20 MWt LAMPRE-II. By mid-1963, LAMPRE-I had served its intended purpose and was shut down. Funding for the construction of LAMPRE-II never materialized.12

Admiral Hyman G. Rickover briefly experimented with fast-neutron reactors for naval submarine propulsion. This effort began with General Electric’s development and operation for the Navy of the land-based S1G prototype at the Knolls Atomic Power Laboratory in West Milton, New York. The S1G, which was HEU-fueled, operated from the spring of 1955 until it was shut down in 1957 after Admiral Rickover abandoned fast reactors for naval propulsion. During its brief operating history, the sodium-cooled S1G experienced trouble with leaks in its

і 3

steam generators.13

The S1G prototype was followed by the deployment of the S2G fast reactor in the nuclear submarine, USS Seawolf (SSN 575). According to Atomic Energy Commission (AEC) historians, Hewlett and Duncan, in their history of the U. S. nuclear navy from 1946 to 1962:

Although makeshift repairs permitted the Seawolf to complete her initial sea trials on reduced power in February 1957, Rickover had already decided to abandon the sodium-cooled reactor. Early in November 1956, he informed the Commission that he would take steps toward replacing the reactor in the Seawolf with a water-cooled plant similar to that in the Nautilus. The leaks in the Seawolf steam plant were an important factor in the decision but even more persuasive were the inherent limitations in sodium-cooled systems. In Rickover’s words they were ‘expensive to build, complex to operate, susceptible to prolong shutdown as a result of even minor malfunctions, and difficult and time-consuming to repair.’14

Consolidation of breeder reactor research at Argonne National Laboratory

In 1946, the newly formed AEC took control of the nation’s nuclear research facilities and tapped Zinn to head the Chicago laboratory, which by then had been reorganized and renamed Argonne National Laboratory (ANL). The next year, the AEC Commissioners decided to consolidate the entire AEC reactor program at ANL.15 The Commission needed reactors not only to produce plutonium for weapons but also for the production of radioisotopes and for general research. There was also widespread public interest in using reactors to generate electric power.16

In drafting his section of the General Advisory Committee report, Zinn stressed power reactors. Here (as had been the case since 1944) a fact of supreme importance was the shortage of fissionable material. Existing stocks of uranium ore seemed scarcely large enough to sustain production of a modest number of weapons, to say nothing of providing fuel for power plants. Zinn believed the only hope for power reactors lay in those which would breed more fissile material than they consumed.17

Zinn convinced the AEC to give the breeder project a high priority and insisted on directing the effort himself. Fermi promoted it by giving lectures extolling the goal of extracting almost 100 percent of the fission energy from natural uranium.18

Experimental Breeder Reactor-I

On November 19, 1947, the AEC authorized ANL to design and build a liquid- metal-cooled, fast-neutron reactor, the second fast reactor in the United States, Experimental Breeder Reactor-I (EBR-I), alternately known as "Chicago Pile 4" and "Zinn’s Infernal Pile."

The EBR-I team decided to cool the reactor core with a sodium-potassium (NaK) alloy. Since they knew little about the effect of this liquid-metal coolant on materials and worried that the control rods might stick or corrode, they decided to cool them with air, which introduced the complexity of designing two completely separate cooling systems. This was especially hard because the sodium-potassium metal would burn in both water and air. Therefore, there could be no fluid leakage.19

From the beginning of the Manhattan Project, questions had been raised about the public safety concerns associated with building reactors in the Chicago area. By summer 1948, Zinn was convinced the project needed to be built at a remote site and asked the AEC to find one.20 The Commissioners chose a site near Arco, Idaho, that had been a proving ground for navy ordnance. It came to be known as the National Reactor Testing Station, now part of the Idaho National Laboratory (INL) and soon housed other ANL reactor projects as well as other government reactors.21

EBR-I was the first fast-neutron reactor designed to both breed plutonium and to produce electric power. The 1.2 MWt (0.2 megawatt electric)22 sodium-cooled reactor went critical on December 20, 1951, and lit four 200-watt light bulbs, thereby becoming the world’s first electricity-generating nuclear power plant. See figure 7.1. EBR-I was fueled with weapon-grade (94 percent-enriched) uranium. On June 4, 1953, the AEC announced that EBR-I had become the world’s first reactor to demonstrate the breeding of plutonium from uranium.

Unfortunately the reactor was designed with a prompt positive power coefficient of reactivity (increases in power had a positive feedback). On November 29, 1955, during an experiment to obtain information about this instability, the reactor had a partial (40-50 percent) core meltdown. The damaged core was removed and the reactor was repaired and operated until it shut down on December 30, 1963.

The accident at EBR-I focused attention on safety issues associated with liquid — sodium fast-neutron reactors and especially the possibility of an explosive criticality due to the partial melting and collapse of the core. This possibility was first studied by Bethe and Tait.23 By 1983, the effective end to the U. S. fast reactor commercialization program, U. S. analysts had concluded that the Bethe — Tait analysis was overly conservative regarding the magnitude of the potential energy release in a fast-reactor accident, but that there were no "universally accepted estimates of upper limits on consequences of hypothetical fast-reactor accidents."24

The one kilowatt (KW) ANL Fast Source Reactor was also built at the National Reactor Testing Station to produce neutrons for the fast reactor development program. Reactor startup occurred on October 29, 1959 and the reactor was operational until sometime in the late-1970s, when it was moved to a new location on the Idaho site.