Appendix 7 LMFBR development in decline

The cancellation of the CRBRP in 1983 was the watershed event that essentially halted all development of the LMFBR in the U.S. and subsequently in every European country except Russia. Prior to that event, development activity was relatively high and technological progress was being made on a variety of fronts. There were five more or less independent forces at work that led to the CRBRP cancellation.

First, a widespread malaise had set in to the LWR industry. Throughout the latter half of the 70’s aggressive regulation by the NRC led to significant increases in plant cost beyond anything that had been anticipated by the utility industry users of the technology. The early nuclear plants, although somewhat more expensive than equivalently sized coal plants, more than made up for their higher capital cost with significantly lower fuel cycle costs. As the plants became significantly more expensive, utility companies began experiencing resistance from their state public utility commissions when it came time to put the plants into the rate base. As plant construction schedules became impacted by mandated retrofits and unanticipated changes many utility companies decided they were in over their heads and cancelled plans for new nuclear capacity. At the same time, the same companies began to experience a decline in load growth. While widespread installation of central air conditioning along with other home appliances had led to load growth in the 4-6% range for the preceding 20 years, the 70s saw load growth dropping to the 1-2% range, removing the need for many of the plants utility companies had in early stages of construction.

Second, the 1970s witnessed the unbridled growth of the environmental movement in the U.S. The word “ecology did not come into widespread use before the late 1960s. By the early 70s, the movement was at a full gallop with the creation of the EPA and a host of environmentally motivated private entities such as the National Resources Defense Council whose focus was activism. Pre-existing societies such as the Sierra Club rebranded themselves from a focus on conservation to environmental activism. By the mid 70s, there were a large number of these entities, and every one without exception became an anti-nuclear crusader. The central themes of these crusaders were (and by and large, still are) that the plants were not adequately safe and there was no way to dispose of nuclear waste. Their activism clearly had an effect on the NRC, which was under just as much attack as were the utilities building the plants. The NRC responded in the fashion to be expected from any government bureaucracy by turning the screws tighter on the utility companies.

Third was the decision in 1974 to split the Atomic Energy Commission (AEC) into separate promotional and regulatory bodies. Thus was born the Energy Research and Development Administration (ERDA), the predecessor of the Department of Energy (DOE) and the Nuclear Regulatory Commission (NRC). Prior to that split, the AEC carried out the spirit and the law of the atoms for peace initiatives launched by the Truman and Eisenhower administrations. It both promoted and regulated nuclear power and attempted to maintain a balance between the two, being careful not to over-regulate for fear of frustrating nuclear power development. Once the NRC was turned loose, there was little in the way of a moderating influence over its regulatory activism. Coincident with the split of the AEC, the Joint Committee on Atomic Energy was disbanded. The Joint Committee on Atomic Energy had previously been a powerful body in congress composed of senior senators and members of the House that promoted nuclear power development in the U.S. Its disappearance left a huge vacuum.

Fourth was the accident at Three Mile Island in 1979. All the environmental activists saw their opportunity and immediately went on the attack. Every utility company in the country that had committed to building or was operating nuclear plants became the object of a media blitz and was forced to answer why they were continuing with their construction programs or operating facilities that were “demonstrably unsafe”. When the molten fuel zirconium cladding at Three Mile Island reacted with the water coolant to produce hydrogen which collected at the top of the reactor vessel, the media had a field day. Everyone knew about hydrogen bombs in those days without having a ghost of an idea how they worked, but the media made the connection anyway setting off a panic. The whole episode scared the pants off a lot of utility company executives and their board members. Following the Three Mile Island accident, there wasn’t a new order for a nuclear plant for 30 years and all those that had been ordered after 1973 were cancelled.

Fifth was the non-proliferation bugaboo. In 1977, then President Carter ordered a halt to CRBRP licensing on the basis that it could contribute to nuclear proliferation.1 The technical merit for this position was apparently based on the observation that the plutonium that was bred in the outer reached of the axial blanket was almost pure Pu239, making it ideal for use in a weapon. There was no effort on the part of the administration to find a technical solution for this issue. The axial blanket thickness could have been reduced, the blanket material could have been changed to Thorium, or the blanket material could have been poisoned with small quantities of Pu241 if this alleged concern had been legitimate. In reality, the issue was really a stalking horse to placate Carter’s environmental activist buddies and prevent progress on the plant.

By 1981, Carter was out of office and President Reagan ordered resumption of CRBRP licensing. At some point during the previous administration, some politico coined the term “technological turkey” to describe the CRBRP. What was meant by this was never made particularly clear, but the term stuck and somehow implied that the technology was no longer relevant or the plant design was behind the times or the plant cost was greater than could be justified. An unholy alliance was formed between anti-nuclear democrats and budget conscious republicans and the plant was cancelled in 1983.

Following the CRBRP cancellation, the European programs in Germany, the U.K. and France were all terminated one by one, leaving only a modest continuing effort carried on by the Russians. Most notable among these European countries is France, the host country for Superphénix. Superphénix was plagued with numerous operational problems. In addition, the responsible French design entity, notably Novatome, was not able to satisfy itself that the LMFBR concept as developed could compete with LWRs.

The most logical plant design for the follow-on to CRBRP would have been in the 1000-1500 MWe size range, oxide fueled, 1000-1050°F outlet temperature, probably of the loop design, probably with helical coil once through steam generators and possibly EM primary system pumps. The core design would probably use a somewhat larger fuel pin diameter, possibly adopting the CRBRP heterogeneous core design to take advantage of its flux flattening, better breeding, and better utilization of blanket assemblies. The primary heat transport system probably would have abandoned the elevated loop concept although the designs in existence in 1984 retained this concept. The fuel assemblies would probably have been wire wrapped hexagonal cross section with more pins per assembly. In the main however, the plant would be a scale-up from CRBRP. In fact, such a design was underway under a joint DOE/EPRI project known as the Large Scale Prototype Breeder, LSPB.

The LSPB Project began in 1982 and was managed out of an office formed by EPRI in Naperville, Illinois and staffed primarily by personnel loaned to EPRI by NSSS vendor companies and architect engineering companies associated with CRBRP. When CRBRP was cancelled late in 1983, DOE personnel decided they could no longer justify supporting a project that had been advertised as a CRBRP follow-on. Nonetheless, work continued on the project through 1986 with a focus on pool-type designs until the DOE arrived at a strategy for sustaining some activity on the LMFBR concept that would be politically acceptable. That strategy began to emerge in 1986.

The basic idea of the new strategy was to focus efforts on “small innovative LMRs”. The “FB” was dropped from the LMFBR acronym because was seen as being particularly annoying to the political opponents, who were mostly arrayed around the environmental movement and non-proliferation concerns. “Innovative” appears to have been intended to convey the notion that somehow the CRBRP was an old fashioned scale-up of the FFTF and the DOE intended to break ranks with the industrial conglomerate supporting the earlier effort and plow new ground. The origins of this counterintuitive (and counter common sense) idea of “small” reactors are not known – one could speculate but to no advantage – but the DOE picked up on this theme with their revised LMFBR program and solicited proposals from the three reactor vendors. At the same time, they withdrew support for the LSPB causing EPRI to similarly withdraw support and close the office that had been set up to pursue the effort.

The DOE invited proposals from the three LMR reactor vendors that had been engaged in the design of CRBRP, Westinghouse, General Electric, and Atomics International. The DOE decided to go forward with the proposals submitted by General Electric and Atomics International. In 1988 the DOE down selected to the General Electric design for further development, which was a small modular pool design. About three years later, DOE funding dried up for the General Electric design, but the company continued to support a token activity surrounded around this concept. Although there have been some initiatives and activity at the national labs, neither the DOE nor the electric utilities have done anything of substance to advance the cause of the breeder reactor in over 30 years.

Five prevailing conditions from the 1970s affecting LMFBR design

There were five conditions that were characteristic of in the 1970s that had a negative effect on breeder reactor design and development. None of the five were lasting nor, in retrospect, should any of them have had the effect they did. The importance of these prevailing conditions and their impact on decision making was not widely recognized at the time, but as always seems to be the case, hindsight is 20-20.

First, there was a cold war going on and both the U.S. and the Soviet Union were furiously expanding their weapons inventories creating a strong demand for plutonium. There was almost no commercial reprocessing, so the only significant source of plutonium was from weapons stocks, which were jealously husbanded. In fact, at the time, CRBRP couldn’t even use weapons plutonium because of a statutory requirement to separate the weapons program from civilian nuclear power. The only stocks of plutonium available for CRBRP were from the DOE civilian program and they were very limited. Highly enriched uranium could have been used as the fissile material, but it is very expensive, and has a poor neutron reproduction rate resulting in a lower breeding ratio. The use of enriched uranium would also have invalidated much of what was intended to be demonstrated. This led to the need to minimize fissile inventory in the core.

A requirement for low fissile inventory inevitably leads to small diameter fuel pins. In fact, both the FFTF and the CRBRP had 0.23-in. diameter pins, which are small in comparison to other LMFBRs that were operating or under construction at the time. Cores fueled with small diameter pins will have a low internal breeding ratio with much of the breeding taking place in the blankets and they will experience fairly rapid reactivity loss with operation minimizing the burnup that is achievable and requiring frequent refueling. The initial approach on CRBRP was to refuel one third of the fuel assemblies each year and a smaller fraction of the radial blanket assemblies. In the case of the initial CRBRP homogeneous core design, one year of operation resulted in about 3½ atom percent peak fuel burnup. Peak burnup of the fuel at discharge was about 11 atom percent. With heterogeneous core designs, plutonium builds up quickly in the internal blankets causing them to generate considerable power at the end of an annual cycle, potentially making it necessary to shuffle them to peripheral regions where the flux is lower during refueling outages. There is more detail on homogeneous and heterogeneous core designs in Appendix 2C, particularly including the change from the homogeneous to the heterogeneous core design on CRBRP.

A second and related condition from the 1970s was the emergence of the uranium cartel. In 1972, representatives of the five major uranium producing nations (at that time Canada, Australia, the former Soviet Union, Niger and Namibia) formed a cartel that drove yellowcake (U3O8) prices sharply higher. There was an expectation that there was no other way to defeat the cartel except with a nuclear concept that was virtually independent of uranium supply – viz. the breeder reactor. This, combined with the fact that 15-20 years of operation of a LWR is necessary to produce enough plutonium for the initial fuel loading of a single equivalently sized LMFBR meant that if there were going to be a surge in orders for LMFBRs, plutonium supplies would be pinched. A further strain on plutonium resources is the fissionable material tied up in the fuel cycle – the time required to allow for spent fuel decay heat to reduce sufficiently for shipment to the reprocessing facility, and the time tied up with shipping, reprocessing, and new fuel fabrication. The expectation was that it would be at least three years between the time that LMFBR spent fuel is removed from the reactor until it is returned again in the form of new fuel.

To fit into the scheme of low fissile inventory with good doubling time, it was necessary to have a refueling system that was fast and could complete an annual refueling cycle in about two weeks. Thus, through the head refueling became the standard for LMFBRs involving a complexity of rotating heads, in-vessel transfer positions between the core barrel and the reactor vessel wall, in-vessel transfer machines (IVTM), ex-vessel transfer machines (EVTM) and an ex-vessel sodium filled spent fuel storage tank (EVST). This subject has been extensively treated in section 6.

The fundamental point to be made from the above is that fissile inventory needs to be low on the list of design considerations and future designs should be focused on operating at high burnup with long periods of time between refueling. Importantly, the plant design needs to be simpler and more economic so that it has a decent chance of being competitive with LWRs. At some point it may become necessary to refocus on fissile inventory in LMFBRs, but that point is a long way off and only after several LMFBRs have been brought on line and begin to strain the fissile supply. For now, the world is awash with plutonium drawn from weapons stockpiles.

A third condition prevailing in the 1970s was the dependence of the U.S. on foreign oil simultaneous with limited and depleted quantities of domestic natural gas – a period long before hydraulic fracturing came into the picture. It was generally believed that nuclear power was needed soon to supply the nation’s energy requirements. This situation elevated the importance of design conservatism to raise confidence in technological success. This was particularly evident with the FFTF reactor. Chronologically, the FFTF followed the Enrico Fermi reactor that was built and operated by a host of utility companies led by Detroit Edison. The Fermi-1 plant experienced two serious mishaps – unreliable steam generators and a partial meltdown caused by a loose part in the reactor. The rest of the plant operated without serious incident. In particular, the design of the Fermi plant was on a track that could have led to a reasonably competitive version at commercial sizes. Nonetheless, the Fermi plant was perceived to be a failure and rather than correcting its few woes, the designers of FFTF seem to have started with a clean slate and took a different path, particularly with the heat transport system.2 One of the key factors involved here was the desire to eliminate the double walled primary system piping system of Fermi. This move was probably motivated by the perception that the double walled system was not inspectable and never would be. The result was an increase in the cost of the plant beyond that which would have shown a promise for economic competitiveness with LWRs. Other factors weighing heavily on the FFTF design were the provisions of closed loops which turned out never to have been used and the absence of a steam generating system. The early FFTF decision makers apparently did not wish to deal with the problem of developing a satisfactory steam generating system while they had so many other irons in the fire (such as the closed loops). This turned out to have been a big mistake. The problem of resolving the steam generator design was left for the CRBRP project to solve, which contributed to excessive CRBRP project costs in funding three separate steam generator designs and a conservative design approach. Moreover, when the FFTF finally made it into operation, it had no revenues from power production to offset operating costs. The result was the operating costs became oppressive and the plant was shut down prematurely.

A fourth condition characterized by the 1970s was the development of the LMFBR more or less in parallel with LWRs. There was certainly a tendency during CRBRP development to compare its features, capabilities, and limitations with those of the LWR. Annual refueling is one example. Another example may be the presence of an operating floor within containment. For the case of the LWR, containment can be entered minutes following shutdown. Because of Na24, it is not possible to quickly enter the spaces carrying primary sodium in a LMFBR. Early liquid metal plant designers all the way back to SRE compensated for this perceived shortcoming in the concept’s operability by providing an operating floor inside containment, the region below housing the inaccessible primary system while the region above being accessible immediately following shutdown. The control rod drive mechanisms, the primary pump motors, and the refueling equipment are located in this space above the operating floor. As a result, the operating floor plus the reactor vessel head are heavily shielded, which adds to the cost and complexity of the plant. If the requirement for refueling in two weeks following shutdown were to be removed, there would be no logic supporting the operating floor concept and it could be eliminated, along with the reactor head shielding. Maintainability issues, although important, also need to be considered (and moderated) from this point of view, since because of the sheer nature of the beast, there will be some things that are feasible on LWRs which cannot be accomplished on LMFBRs and vice versa. The containments on EBR-2, SEFOR, Fermi-1, FFTF, and CRBRP are a carryover from LWRs. There was no reason for doing this other than the precedent set by LWRs.

The fifth, and probably the most important condition that prevailed in the 1970s was the widespread notion that the LMFBR was the follow-on design concept after the LWR had matured, rather than a competitor of the LWR.  That notion tied the fate of the LMFBR to the LWR and served as an apology for why the LMFBR was more expensive than the LWR – it was a concept to be realized after the LWR became obsolete as a result of uranium resource limitations.  When one considers that sodium cooled thermal reactors were once considered as potential competitors with LWRs, this follow-on notion makes no sense at all.  There is not that much difference between sodium thermal reactors and LMFBRs that would appreciably affect the capital cost of one over the other.  The entire design approach needs to be looked at from a different angle.  The basic principles of sodium coolant and breeding need to be better capitalized upon.  In both design and licensing, the LMFBR must be dealt with as a very different option from LWRs.  The LMFBR is an alternative to the LWR and should compete with it along with other sources of electric power generation.  The features which make the LMFBR unique need to be capitalized upon in a way that makes it considerably more economic and attractive to utility company users as a near term and attractive alternative to LWRs. 


1 Carter simultaneously ordered a halt to the licensing of the Barnwell reprocessing plant on the same basis, which was the only commercial reprocessing plant under construction at a time when no others were operating.  The owners of Barnwell subsequently cancelled the project.  There has been no reprocessing plant construction of any kind, commercial or government, in the U.S. since.

2 It is of some interest to note that the heat transport system of the German SNR-300 is remarkably similar to both FFTF and CRBRP.  It is not clear which one influenced the other.