Experimental Breeder Reactor-II (EBR-II) was arguably the most successful of the U. S. fast reactors. See figure 7.2. It was a 62.5 MWt, 20 megawatt electric (MWe), sodium-cooled, "pool-type" reactor, i. e. the heat exchangers for transferring heat to a secondary loop of liquid sodium were submerged in the reactor vessel. It was designed by ANL and constructed, beginning in June 1958, at the National Reactor Testing Station (today the Materials and Fuels Complex in the Idaho National Laboratory). Criticality at low power without sodium coolant was achieved on September 30, 1961; criticality with sodium coolant on November 11, 1963; and design power on September 25, 1969.

EBR-II demonstrated the feasibility of a sodium-cooled fast breeder reactor operating as a power plant. It operated initially with metallic HEU fuel. A hallmark feature was that it had an adjoining Fuel Cycle Facility (FCF) (now called the Fuel Conditioning Facility) that permitted continuous reprocessing and recycling of fuel to keep the working inventory down.25 EBR-II spent fuel was processed and fresh fuel fabricated at the FCF from 1964 to 1969.26 In 1967, the EBR-II was reoriented from a demonstration plant to an irradiation facility.

After cancellation of the Clinch River Breeder Reactor (CRBR) in 1983, the EBR-II reactor and the FCF became the research and demonstration facilities for the Integral Fast Reactor (IFR) concept promoted by ANL. The IFR program was terminated and EBR-II began shutdown operations in September 1994, after 30 years of operation.

The EBR-II shutdown activities included defueling and draining the primary and secondary sodium loops. The FCF has been converted to a Fuel Conditioning Facility whose mission is to electrochemically treat spent EBR-II fuel to create radioactive waste forms that are acceptable for disposal in a national geological repository. The fuel is not considered suitable for direct disposal in a geological repository because it contains sodium to provide a good thermal link between the fuel pellets and the fuel cladding. Sodium would react with any water that penetrated the cladding to generate hydrogen. The laboratory has signed an agreement with the state of Idaho that the fuel conditioning work will be completed by 2035.

The Monju accident triggered a significant shift in Japan’s fast breeder reactor program. After the accident, the JAEC established an ad-hoc "Roundtable Committee on FBR" to develop new policies. Prof. J. Nishizawa of Tohoku University, who was not a fast breeder reactor expert, chaired the committee. The Committee also included experts from outside the nuclear community, including Mr. Yukio Okamoto (ex. Ministry of Foreign Affairs), Prof. Sawako Takeuchi, an economist, and Prof. Hitoshi Yoshioka of Kyushu University, a nuclear critic. Although the Committee confirmed the continuation of fast breeder reactor development, it recommended a more realistic and flexible approach, declaring that the fast breeder reactor should be considered as a promising option (rather than the ultimate goal) and suggested "periodic review of R&D programs from the standpoint of technological and economic feasibility."5 It also endorsed a more diversified R&D program to explore technical alternatives to existing fast breeder reactor technologies.

Following this report, the JAEC’s Long Term Plan, published in 2000, established a goal "to maintain the technological option of the fast breeder reactor and its associated fuel cycle…in order to prepare for future energy problems," and recommended programs to explore "various alternatives to currently developed sodium-type fast breeder reactor and PUREX (wet) reprocessing technology."6

The inner and outer breeder sections of the DFR were originally loaded in 1958 with natural uranium elements clad in stainless steel. Early in 1965 it was found that a few of the lightly irradiated elements in the outer breeder were difficult to remove, although the inner breeder elements were in good condition. A comprehensive survey of the outer breeder was carried out in September 1965, and a number of elements were found to be distorted or swollen. Investigation showed that this had been caused by higher than normal uranium temperatures due to abnormal coolant flow conditions in some regions of the breeder. This will not occur in future fast reactors since coolant flow conditions will be different, and the breeder fuel itself will be ceramic and therefore not subject to the temperature limitations of natural uranium. It was decided to remove 500 breeder elements, and to carry out the work. Special cutting tools and removal equipment had to be manufactured.

The work was completed by the end of December and the reactor went critical again on 23 January after loading new experiments.

Such incidents in no way weakened AEA confidence in the concept of the fast breeder. On the contrary: while they pressed on with detailed designs for the PFR they had already satisfied themselves that the prospects were excellent:

The design study of a 2×1000 MWe fast reactor power station in general endorsed the conceptual design of the prototype fast reactor as representing the most likely features of the first commercial fast reactors. A capital cost estimate for this study indicates a cost similar to that of the best thermal reactor available at the same time, with potential for further reductions. (AEA Annual Report 1965-66, paragraph 157)

The Enrico Fermi Atomic Power Plant (Fermi 1) was the brainchild of Walker L. Cisler, who in 1951 became president and general manager, and later CEO and chair of the board of Detroit Edison. Nuclear energy had caught Cisler’s attention in 1947 when he joined an AEC advisory committee on how to make connections with private industry. In December 1951, Cisler presented to the AEC a Dow-

Detroit Edison study, one of four industry studies that found that "atomic energy had an important potential for power production even if reactors were not yet economical for that purpose alone."27

In 1952, Cisler assumed the leadership responsibilities for organizing electric utilities to develop the Enrico Fermi Breeder Reactor Project. The project was formally organized in 1955 as the Power Reactor Development Company (PRDC) with 34 companies participating. In January 1956, PRDC applied to the AEC for a construction permit to build the reactor on the shore of Lake Erie at Lagoona Beach (near Newport, 30 miles from Detroit), Michigan. The construction permit was granted on August 4, 1956, groundbreaking took place four days later, and the pouring of concrete began in December 1956.28

In terms of core size and power, the Fermi 1 reactor was the largest fast-neutron reactor built up to the time. Criticality was achieved on August 23, 1963. The 200 MWt (66 MWe) sodium-cooled HEU-fueled power reactor differed from EBR — II in that it was based on a loop design in which the liquid sodium primary coolant transfers its heat to secondary sodium in an external intermediate heat exchanger.29

In October 1966, a blockage of the flow of sodium through part of the core caused a partial core meltdown. The accident was attributed to a zirconium plate that had become unfastened and obstructed the sodium flow into a fuel assembly. Two of the 105 fuel assemblies melted during the incident, but no contamination was recorded outside the containment vessel. This accident inspired the book, We Almost Lost Detroit.30

Damage to the reactor and fuel assemblies took approximately four years to repair. In May 1970, the reactor was ready to resume operation, but a sodium explosion delayed startup until July. In October, the reactor finally reached a power level of 200 MWt. During 1971, it only generated 19.4 gigawatt-hours (GWh) of electricity, however, corresponding to an average capacity factor of 3.4 percent. The PRDC therefore declined to purchase additional uranium fuel to continue plant operation. In August of 1972, upon denial of the extension of its operating license, shutdown of the plant was initiated. Operation ended on September 22, 1972. The decision to decommission the plant was made November 27, 1972. It was officially decommissioned on December 31, 1975.

The 2005 Long Term Plan was renamed the Framework for Nuclear Energy Policy and established a new fast breeder reactor commercialization target of 2050.7 In 2006, the Sub-committee on Nuclear Energy Policy of the Government’s Advisory Council on Energy published Japan’s Nuclear Power National Plan, which laid out detailed policy measures based on the JAEC’s framework.8 The Nuclear Power National Plan reiterates the 2050 commercialization target for the fast breeder reactor and announced a goal of developing a post-Monju demonstration fast breeder reactor by 2025. The associated Phase II "Feasibility Study on Commercialization of Fast Reactor Cycle Systems" compared various types of fast reactor designs and associated fuel cycle technologies, and tentatively identified a sodium-cooled fast reactor with advanced wet reprocessing technology as the preferred option.9

The study compared four fast-neutron reactor designs: sodium-cooled (1.5 gigawatt electric (GWe)) with metallic fuel, helium-cooled (1.5 GWe) with nitride fuel, lead-bismuth-cooled (0.75 GWe) with nitride fuel, and water-cooled (1.356 GWe) with mixed oxide fuel (MOX). Unit construction cost estimates for a sodium- cooled fast breeder reactor would be the lowest ¥180,000/kilowatt compared with approximately ¥200,000/kilowatt electric for the other designs. Four basic options for advanced reprocessing and fuel technologies were evaluated:

1. Advanced wet reprocessing plus simplified pelletized MOX fuel;

2. Metal electro-refining reprocessing plus injection cast metallic fuel;

3. Advanced wet reprocessing plus vibration packing (Sphere-pack) MOX fuel; and,

4. Oxide electro-refining reprocessing plus vibration packing (Vipac) MOX fuel.

The most economical option would be the advanced wet reprocessing plus simplified pelletized MOX fuel in a large (200 ton/year) plant (~¥0.5-0.66/kWh) with the alternatives costing up to ¥1.6/kWh. None of these cost estimates are engineering estimates. All represent development targets required for fast breeder reactors to be competitive with light-water reactors.

The Nuclear Power National Plan also set out important principles for the future development of fast breeder reactor and fuel cycle systems. First, it established a cost-sharing principle to distribute demonstration fast breeder reactor project costs between the utility companies and the Government. It specified that the private sector would invest an amount equivalent to the cost of a commercial light-water reactor, significantly reducing the financial risk for utilities.10

Another important principle of the Nuclear Power National Plan was that the second commercial reprocessing plant after the Rokkasho plant should be timed to match the pace of fast breeder reactor development and deployment. It suggested that planning for the second reprocessing plant start around 2010.

In 2007, the Government increased the fast breeder reactor R&D budget for the first time since the late 1990s to ¥44 billion in response to these new programs and principles. It is now approximately 10 percent of the total nuclear budget. This budget increase was prompted partially by international developments, notably the announcement of GNEP, which had an initial emphasis on using fast-neutron reactors to fission plutonium and other transuranic elements in light-water reactor spent fuel.

The socio-political factors behind Japan’s entrenched commitment to fast breeder reactor technology

Despite the marked slippage of fast breeder reactor commercialization targets, why have Japanese commitments to the fast breeder reactor remained, at least publicly, unchanged? There are three possible explanations.

The fast ‘breeder’ reactor is the system on which the long term prospects of nuclear energy generation are based … Work on this system has been increasing steadily for some ten years and the greatest effort of the

AEA’s research and development programme is now devoted to this type of reactor. Expenditure in 1966-67 was approximately £12 million and there will be increasing capital expenditure over the next few years as the construction of the DFR prototype proceeds. This system is regarded as likely to provide a very cheap source of electricity. Building costs (at 1967 prices) of fast reactor stations are expected to be as low as £50 per kilowatt installed and generating costs to be reduced ultimately to 0.3d (old pence) per kilowatt hour. The prototype, a large station producing 250 MWe, is expected to be on power in 1971.

By 1968, the AEA was looking to have at least 15 gigawatt electric (GWe) of fast breeders in operation by 1986. On the basis of "another bold decision" by government, exploitation of the fast breeder would be "the major event of the rest of the century". By 1969 the AEA was asserting that "the UK has the firm intention of introducing fast reactors as rapidly as possible after the operation of our 250 MW prototype."

Meanwhile, in May 1967, the primary cooling circuit of the DFR sprang a leak of molten-sodium-potassium. The reactor was shut down in July 1967 for nearly a year. The DFR was also manifesting other engineering problems. No reactor hitherto in operation had subjected its structural materials to intense high energy neutron radiation for lengthy periods. To find materials able to withstand the demanding environment in the core of a fast reactor was a daunting challenge.

While these practical problems occupied the attention of the staff at Dounreay, the AEA was linking up with the Central Electricity Generation Board (CEGB) and the two reactor building consortia for further design studies on commercial fast breeder power stations. On 14 October 1970, introducing the AEA annual report, chairman Sir John Hill characterized the outcome thus:

…we have had a most useful study of the fast reactor by a group consisting of engineers of the CEGB, the industrial design and construction firms and the Authority… we have now an agreed programme… which could lead to the CEGB being able to start the construction of the first civil fast reactor, possibly of 1300 MW, by early 1974 … seeing how the prototype fast reactor performs in 1972 and 1973.

As it turned out, the PFR did not perform at all in 1972 or 1973. Nevertheless, with the PFR falling steadily farther behind schedule, the 1971 AEA annual report was still confident. The cooperative study had resulted in:

the formulation of a strategic plan for the introduction of fast reactors to the CEGB network; this assumes that construction of a first commercial station will start in 1974 as a ‘lead’ station, following operation of the PFR. This would be followed by other stations after an interval of perhaps two years. This plan assumes that the technical and economic results from the development programme confirm present expectations; it will be reviewed each year in the light of progress achieved.

Despite the commercial failure of Fermi 1, the U. S. Liquid Metal Fast Breeder Reactor (LMFBR) development effort picked up momentum in the 1960s, aiming for commercialization of the breeder before the end of the century.31 In its 1962 Report to the President on Civilian Nuclear Power, the AEC specifically recommended that future government programs include vigorous development and timely introduction of the breeder reactors, which the Commission believed essential to long-term use of nuclear energy on a large scale.32 By 1967, the LMFBR was the AEC’s largest civilian power development program.33 The Commission’s program began to embrace efforts to build an industrial base and obtain acceptance of the LMFBR by utilities, primarily through planned government-subsidized construction of commercial-scale LMFBR power plants.34 The Commission came to see its program "as the key to effecting the transition of the fast breeder program from the technology development stage to the point of large-scale commercial utilization."35

In furtherance of these objectives, the Commission, in 1968, issued a 10-volume LMFBR Program Plan prepared by ANL. The dual objectives of the plan were to:

1. Achieve, through research and development, the necessary technology; and,

2. "(A)ssure maximum development and use of a competitive, self-sustaining industrial LMFBR capability."36

The aim was to develop an economically viable, commercial-scale LMFBR by the mid-1980s.37 In a 1969 cost-benefit study of the breeder program prepared by the AEC, the LMFBR commercial introduction date was assumed to be 1984.38

With growing concern about a possible energy crisis, rapid commercial implementation of LMFBR technology had become a national mission.39 It would remain AEC’s highest priority development program until 1977, when President Jimmy Carter sought to cancel the Demonstration CRBR project; and it remained a high priority program until 1983 when the CRBR project was terminated by Congress.

In the style of President Kennedy’s 1960 commitment to put an American on the moon by the end of the decade, President Nixon, in his June 4, 1971 Energy Message to Congress, announced as the highest priority item of his energy program "(a) commitment to complete the successful demonstration of the LMFBR by 1980."40 This goal was endorsed by Congress’ Joint Committee on Atomic Energy.41

In 1967, a special law established PNC with the mission to develop indigenous fast breeder reactors and their associated fuel cycle technologies. This mission endured after the Monju accident in 1995 when PNC was renamed the Japan Nuclear Fuel Cycle Development Institute (JNC). JNC subsequently merged with the Japan Atomic Energy Research Institute (JAERI), a national research institution responsible for fundamental nuclear technology (including fusion) and nuclear safety research and in 2006 it became the Japan Atomic Energy Agency (JAEA). JAEA was established with the continued mission of developing fast breeder reactor and fuel cycle technologies. With this legal commitment to fast breeder reactor programs, it may not be easy for Japan to change its nuclear research agenda.

It is estimated that, in only thirty years from now, over three quarters of all electricity in the United Kingdom will be generated from nuclear power and that more than half of this nuclear generation will stem from fast breeder reactors (to the development of which almost half the effort on the Authority’s reactor programme is currently geared).

Sir John Hill had already expounded to the fourth U. N. Conference on the Peaceful Uses of Atomic Energy in Geneva in September 1971 on the "strategic plan" endorsed by the electricity authorities, the nuclear power industry and the AEA:

By 1979-80 we should have had seven years’ operating experience with the PFR, constructional experience of perhaps three or four large commercial stations, and initial generating experience from the first of these larger units. On this basis we would expect that by approximately 1980 we would have sufficient confidence and experience to incorporate fast reactors into the United Kingdom generating system to the maximum extent consistent with the availability of fissile material and the growth of demand for new generating plant. Whether such a timetable can, in fact, be achieved will depend on technical developments over the next few years. This, however, is the plan to which we are working and so far we see no reason why it should not be achieved.

In 1975 the U. S. Government Accountability Office (GAO) estimated that the "AEC’s total LMFBR program funding through fiscal years 1948-74 was approximately $1.8 billion."42 GAO gave the LMFBR Program costs for fiscal year 1975 as $481 million,43 which, in 2006 dollars would be approximately $1.6 billion (figure 7.3). The commercialization effort featured two components, a base program R&D effort focused on two test reactors, and a demonstration plant effort, the CRBR.

Southwest Experimental Fast Oxide Reactor

All early fast breeder concepts were based on metallic fuel. In the 1960s, however, work was begun on the use of ceramic, mixed plutonium-oxide/uranium-oxide (MOX) fuel. The Southwest Experimental Fast Oxide Reactor (SEFOR) was a 20 MWt sodium-cooled MOX-fueled fast-neutron reactor designed to determine the operating characteristics of a reactor with MOX fuel, and, in particular, to examine the implications of the Doppler thermal feedback coefficient associated with the use of MOX.44 SEFOR did not produce electricity.

Located near Strickler, Arkansas, SEFOR was built and operated for the AEC by General Electric Company under the Southeast Atomic Energy Associates, a nonprofit consortium formed by 17 power companies and European nuclear agencies including the Gesellschaft fur Kernforschung of Karlsruhe, West Germany.

Experiments at SEFOR confirmed that the negative temperature coefficient of reactivity associated with the use of mixed-oxide fuels would improve the safety of fast reactors under accident conditions involving increases in the fuel temperature.

SEFOR began operating in May 1969, and was shut down three years later. The fuel and irradiated sodium coolant were removed and taken offsite later in 1972, and some dismantling performed. The reactor was acquired by the University of Arkansas in 1975 and is still owned by the university, although the university has never operated it.45