For several years the breeder reactor has been the centerpiece of the Administration's long-run energy strategy. It does indeed seem to promise surcease from scarcity: unlike the generation of electricity from fossil fuels, which some fear will become increasingly scarce in real terms, use of the breeder holds out the promise of creating a plenitude of another nonrenewable resource—uranium—by making a small amount of it go a very long way.
All nuclear fuel cycles, conventional and breeder alike, exploit the energy release in the fissioning of unstable, or "fissile," atomic nuclei. But the neutrons released in the fission process, to the extent that they impinge upon and are captured by "fertile" nuclei, transform those fertile nuclei into additional fissile nuclei, such as fissionable plutonium. If a given nuclear fuel cycle produces more than one new fissile nucleus for each fissioned nucleus, it is a "breeder" cycle; if less than one, it is a "conventional" cycle. All of the light-water reactors presently operating or planned in the United States are of the latter variety. They must be continually fed fresh amounts of uranium fuel, "enriched" in its content of the fissile isotope, which is only a small part of all naturally occurring uranium. Thus they present the ultimate specter of uranium exhaustion, in a period possibly limited to decades. Hence the appeal of the breeder technology, which can use unenriched uranium and thus assure that uranium reserves will last for centuries.
This mix of fear and promise has propelled the breeder reactor R&D program into a dominant position in the overall energy R&D budget: roughly one-half billion dollars went to it in fiscal year 1974, out of a total energy R&D budget of slightly less than one billion. But, when the history of policy responses to the energy crisis of the 1970s is eventually written, 1974 may possibly also be seen as the year when the assumptions underlying the breeder program were decisively challenged. From a traditional profit-and-loss accounting standpoint, the breeder is being viewed by an increasing number of informed critics as a bad investment. On environmental and safety grounds many see it as a potential disaster.
It now appears quite uncertain when, if ever, the breeder will actually be able to compete economically with the conventional light-water reactor. That the breeder technology carries higher capital costs than the light-water reactor has never been disputed. Its proponents have argued, however, that, as uranium becomes scarcer and hence more costly, the higher fuel efficiencies of the breeder will make it the more economical electricity generator. This argument has been attacked from two flanks. First, somewhat improved, although still uncertain, estimates of uranium reserves cast doubt on the likelihood of any large climb in uranium prices for a long time to come. Second, the best present projections of the capital-cost differential between light-water and breeder reactors are so unfavorable to the breeder that even pessimistic uranium price assumptions do not make the breeder the cheaper technology.
Capital-cost estimates for the Clinch River (Tenn.) demonstration liquid metal fast breeder reactor plant, for example, have climbed to well over $3,000 per kilowatt, about five times the current price of light-water reactor capacity, and it is becoming increasingly doubtful that the groundbreaking for the plant will go forward early this year as scheduled. Breeder advocates insist that the inferior design of the plant is the culprit, and not the breeder technology. But the near tripling of capital cost estimates in the past two years has led others to question whether a major commitment to breeder technology at this time may not be a serious mistake. (Light-water capital costs—and, for that matter, the capital costs of fossil-fueled electric generating capacity—have also risen precipitously during this period, but at a lower rate.) The Atomic Energy Commission's cost—benefit analysis of the breeder reactor, the latest version of which appeared in draft form in the spring of 1974, has done little to still the critics; instead, the competence and validity of that analysis have been widely challenged.
This cost—benefit analysis appeared as part of the commission's environmental impact statement on the breeder reactor, and the portions of that statement dealing with environmental factors have done even less to resolve controversies over the potential environmental hazards of a "plutonium economy." A commitment to the breeder will carry with it the problem of radioactive wastes of high toxicity for essentially the indefinite future: the half-life of plutonium (the time span over which the plutonium nuclei in a given batch decay) is 24,000 years. Compared with the light-water reactor, the stock of long-lived wastes generated by the fast breeder has been variously projected as between three and ten times as large, and even the light-water waste output has not been universally accepted as a defensible risk, given the toxicity and weapons potential of plutonium. Some argue that the definitive solution to the waste disposal problem is at hand: burial in deep salt formations, which are believed to be stable over geologic time spans. But the risks of being wrong are awesome: plutonium is among the most toxic of known toxins, and the dispersal of tons of plutonium into the environment would have catastrophic consequences. In addition, safety evaluations of the breeder reactor are in their infancy. Unlike the light-water reactor, which is fueled by a low-enriched uranium that cannot sustain a reaction in which a sudden energy release destroys the containment, there is no evident rationale for ruling out an accident of this kind in a breeder reactor. Unless the reactor were sited underground, toxic radioisotopes, in the event of such an accident, would be widely dispersed.
A possible "sleeper" in the picture is the rather different kind of breeder reactor being developed at the Shippingport naval reactor facility under the direction of Admiral Hyman G. Rickover, the successful promoter of the nuclear-powered submarine. While disclosing little about his costs and the specifics of his methods, Rickover appears confident of developing a successful breeder reactor before the 1990 date in vogue among supporters of the plutonium breeder. The Rickover reactor breeds fissile uranium-233 from thorium, and then "burns" that fissile uranium. Apparently, however, fuel changes will be required much more frequently than in the plutonium breeder. For this and other reasons, the nuclear industry, while taking Rickover at his word that he will have a functioning fuel cycle within a few years, is extremely skeptical that it will be commercially competitive.
All of these uncertainties surrounding the breeder are reflected in the projected budget allocations for fiscal years 1975 and 1976. The dollar allocation to the breeder program now bulks relatively smaller in the overall energy R&D budget. Coal, "new sources" (solar, geothermal, and so on), and energy conservation are the net gainers, in that order. But neither has the breeder program withered away. Of the $10 billion in energy R&D expenditures projected for fiscal years 1975-77, roughly $4.1 billion is for nuclear energy, and much of that is for the breeder. These figures will change, but probably neither sharply enough nor fast enough to satisfy the critics of the breeder program.
A version of this article appeared in print in the January 1975 issue of Resources magazine.
