231 December 3, 1982 THE CLINCH RIVER FOLL Y INTRODUCTION When a measure to terminate funding for the Clinch River Breeder Reactor was defeated by just o ne vote last September it became. clear that support for the program had eroded far more than most people realized funding was further underscored by the fact that three Senators known to be unsympathetic to the project were out of town at the time of the vote The tenuous support for Clinch River's It now appears that there will be another attempt to terminate funding for the Clinch River Breeder during the lame duck session, either as an amendment to an appropriations bill, or as a part of the continuing resolution, should one. become necessary. If this move proves successful, it will end a controversy that has lasted nearly a decade--a.controversy which. is central to the entire future of nuclear energy.

It will also end a project that has become a multi-billion dollar folly.

For many conservatives, Clinch River presents a dilemma.

They are, on the one hand, strongly supportive of nuclear energy, but they are also concerned about the burgeoning federal deficit.

Their opposition to the Clinch River Breeder, therefore, is born more out of a concern to limit federal spending than opposition to nuclear power. The stakes are high. If, as spokesmen for the nuclear industry contend, the death of Clinch River will lead i nevitably to the death of nuclear power in the United States, conservatives would undoubtedly continue to support the project. But the validity of this argument has become steadily more doubt ful as new uranium discoveries, a general slowdown in the const r uc tion of nuclear power plants, and increasingly large cost overruns bring the wisdom of supporting Clinch River into question. 2 In order to make a rational decision on continued funding for the Clinch River Breeder, Congress needs answers to a number o f questions. Among the most important of these are 1) What are the prospects for the availability of conventional uranium supplies in the medium term 2) How does the Clinch River Breeder's technology compare with other options which are competing with it f o r scarce research dollars 3) When will a breeder reactor really be needed 4) How does the development of a breeder reactor fit in with the overall development of nuclear power in the U.S THE GENESIS OF THE BREEDER When scientists first considered nuclear p ower for generating electricity, what attracted them most was a reactor's low fuel costs. Initially, the prospective fuel cycle advantages of nuclear powered electricity appeared to be outweighed by the lack of fissile uranium to fuel reactors. Indeed, in 1944, scientists on the Manhattan Project estimated the entire world's supply of fissile uranium could generate the electricity needs of the U.S for only one and a half years This suggested breeding. Only seven-tenths of one percent of the uranium found i n nature is fissionable, that is, usable for nuclear power generation. This small percentage is U 235 which gets its name from having 235 neutrons in its nucleus).

Nearly all the rest of the uranium in nature consists of the nonfissile form of the element U 2

38. But when this form of uranium absorbs neutrons in a reactor, it undergoes a transmuta tion into a material not present in nature--plutonium--which is fissionable.

This plutonium transmutation makes breeding possible. If rods of U 238 in a reactor are arranged so just the right amount of neutrons hit them, it it possible to produce more fissile plutonium from the targeted U 238 than is consumed in the reaction.

With breeding, it would be possible to extend one and a half year's worth of fuel (circ a 1945) into enough to meet the electri city needs of America for 75 to 150 years It was because the supply of uranium was thought to be very limited that a breeder reactor seemed to make sense. But Admiral Hyman Rickover's success with light water reacto r s and the massive new finds of uranium in the Rocky Mountain and Southwest regions of the U.S. during the early 1950s made the breeder far less attractive as a commercial reactor. Moreover, technical problems cast considerable doubt on its commercial viab i lity. Breeders use volatile sodium as a coolant, and so the welds and fittings in these reactors must be of much higher quality than those in light water reactors. In addition, the problem of sodium volatility 3 means. that the fueling of remote control. T hese and more attention to coolant systems in general breeders must be done entirely by other technical difficulties require back-up pumping systems and safety The Atomic Energy Commission (AEC) had been successful in launching the light water reactor as a commercial proposition.

Many of these were on order; more were likely a problem. Domestic uranium reserves have proved to be thousands of times greater than estimated by scientists in 19

48. However the growing orders for light water reactors expected d uring the 1970s threatened a near-term shortage of cheap uranium supplies Yet this presented To meet this perceived crisis, the AEC decided to move on from mere breeder research and development to a breeder demonstra tion It was at this point that the Cli n ch River reactor design was promoted. Seen as a 375 megawatt reactor, it was to be the first step in the AEC's demonstration. Once completed, a final scale up--the Large Demonstration Plant--was to follow and be ready in time for private industry to start building its own breeders by the middle or late 1980s THE ECONOMICS OF BREEDERS 4 Although uranium's price rise in the middle 1970s seemed to fortify the case for building Clinch River, this turned out to be a false sign. There were several reasons for th i s. One was that although utilities assumed the demand for electricity would continue to grow in the 1970s at the same rate as the 1960s higher electricity prices in the mid 1970s caused the rate of growth to drop sharply. This caused many of the utilities to cancel their orders for reactors In the late 1960s and early 1970s, moreover, the nuclear industry assumed supplies of cheap uranium were limited to known reserves. But, when uranium prices began to increase, there was more incentive to prospect and ma ssive, inexpensive super grade ore finds were made in both Canada and Australia. In the U.S proven and probable reserves doubled.

This uranium surplus, moreover, is projected to increase.

Free world uranium inventories stored out of the ground are already nearly six times greater than annual uranium consumption and are expected to keep growing ted to be nearly eight times that year's uranium demand. As for U.S. uranium inventories, the Department of Energy (DOE) recently projected that it would take at le a st 15 years before they are drawn down to normal market levels By 1991 inventories are projec These developments have had a profound effect on estimates of the future size of the nuclear power industry. Official projections of nuclear capacity on line by the year 2000 dropped from a high 1500 gigawatts to the current Commerce Department low 4 of 105 gigawatts-nearly a fifteen-fold decrease. Indeed, uranium ore is now so abundant U.S. mines are being closed for the lack of any market.

None of this has been good news for commercial breeder reactors. Even after a decade of demonstration efforts both here and abroad, fast breeder reactors remain much more expensive to build than light water reactors. The French, who are now most advanced in the demonstration o f breeder reactors, freely admit that their reactor costs at least 2.28 times that of a comparable conventional light water reactor. While some improvement is expected, they expect to bring the comparative capital cost of a I'maturell breeder down to 1.68 t imes that of a light water reactor of the breeder only if there were to be significant increases in the price of fresh uranium fuel. Because fuel cycle costs account for just a small percentage of the costs of producing nuclear electricity from a light wa t er reactor--between 10 and 20 percent uranium prices would have to rise to unprecedented levels before the breeder is likely to become the more attractive option. Even the most favorable analysis suggests that,uranium prices would have to increase more th a n ten-fold, to 188 per pound in 1982 dollars, before a commercial breeder could compete with existing light water reactors DOE'S own study of the issue, published in 1980, projected uranium breakeven prices for liquid metal fast breeders of 180 to $300 pe r pound (1982 dollars Finally, in a separate contract study done for the Arms Control and Disarmament Agency in 1981, three uranium breakeven prices were determined 325 (comparing existing light water reactors with breeders 403 (comparing improved light wa t er reactors using fuels likely to come on line by 1990 with breeders) and $626 (comparing a light water reactor whose fuel is reprocessed with breeders The average of these estimates is nearly 18 times uranium's current price The General Accounting Office , the Congressional Research Service, and even DOE agree that if a full-sized breeder were now ready to start operation it would be a poor investment at any time in the next 40 years This capital cost differential would change to the advantage THE DEPARTME N T OF ENERGY AND THE CLINCH RIVER PROJECT Given the high probability that any current or planned liquid metal breeder designs will be uncompetitive, there is serious reason to question whether a near-term breeder demonstra tor in the United States makes an y sense. This is particularly the case with designs using liquid sodium. The French are so far ahead of the U.S. in their demonstration of the technology that we could learn far more, at much less cost, simply by purchasing French patents and keeping a sma ll team of observers at the French site 5 The irony is that the Department of Energy is now diverting funds from its breeder fuels research and development program just to keep Clinch River rolling.

Breeder research and development is desirable and importa nt to developing advanced fuels for both existing reactors and the next generation single, unattractive, dated breeder design (completed more than a decade ago) is likely to be very harmful to research should focus research and development efforts on a br e eder that has 1) significantly lower capital costs; 2) the ability to be initiated without the cost and delay inherent in plutonium repro cessing; and 3) that is less concerned with maximizing breeder ratios rather than with simply achieving a breeding or conversion ratio of about one This is a major mistake.

But to link these efforts to the prototype of a DOE instead In December 1981, DOE stated that $1.1 billion had been spent procuring components for Clinch River and that it would take another 2.57 bill ion to complete the project. Adjusting these figures to 1982 dollars would give a total DOE project cost estimate of approximately 3.6 billion. Even this figure, however is probably well below the project's likely final cost because the DOE either ignored or grossly underestimated five major cost factors 1) The cost of borrowinq money from the U.S. Treasury.

This is isnored in DOE'S Clinch River estimate, yet all private utilities-must include the cost of borrowing money in. their construction estimates. If Treasury Bill rates are used as a guide to future interest rates (General Accounting Office practice th e interest costs for Clinch River could amount to 4 billion 2) Plutonium fuel for Clinch River. Clinch River will require 6 metric tons of plutonium for its first five years of operation. Commercially reprocessed plutonium valued at 40 a gram would cost th e project nearly 250 million reprocessing is not available, and more refined weapons-grade plutonium fuel must be obtained from DOE'S defense programs, the fuel cost would soar to 1 billion. Yet only $10 million was allowed in DOE'S original Clinch River c o st estimate If commercial 3) Costs Due to Delay. DOE assumes Clinch River, which is a unique breeder demonstration facility, can be constructed in seven years-less than one-half the time it currently takes to build a standard light water reactor In additi on, DOE has allowed only 200 million as a contingency for cost overruns.

Assuming an interest rate of 12 percent, however, a delay of just 1 to 3 years would increase the project's cost by between $400 million and $1.3 billion 4) Reprocessing of Breeder Fu el. Clinch River's own fuel must be reprocessed. Yet, DOE'S cost estimate makes no provision for such reprocessing. The department is planning to build a $1 billion breeder fuel reprocessing plant, and assuming that 30 percent of the plant is dedicated to Clinch River's needs, this would add $300 million to the bill 5) Breeder Fuel Fabrication. DOE'S estimate includes money for fabricating breeder fuel for Clinch River. Rather than using this money to buy this service directly from current suppli ers, DOE intends to build a new national facility. (the Secure Automation Fabrication Facility). This will cost nearly 500 million, and should be added to the cost of Clinch River.

Including these costs gives a final project cost ranging between 8 to $11 billion. I t should be noted, however, that these higher estimates are themselves conservative, in that they exclude all costs associated with Clinch River's waste management requirements, the plant's eventual decontamination and decommis sioning, and any further co st escalation.

CLINCH RIVER AND THE PRIVATE SECTOR In 1971 the Atomic Energy Commission estimated that the Clinch River project would cost.$400 million. Private industry was so convinced that breeders would be economical before the 199Os, that it promised to meet over one-half the project's costs--$257 million. But when projected costs jumped to nearly 700 million the following year, the industry stuck to is original 257 million pledge. Since 1972, the cost estimates have escalated ten-to twenty-fold, but t he industry has contributed a total of only $122 million, half of which has been in the form of in-kind services. Moreover, in hearings held last year before the House Science and Technology Committee, the utilities' legal counsel argued that because of d elay in the project, private industry was no longer obligated to contribute any additional money to Clinch River.

Almost all of the major reactor vendors and nuclear utilities still support the project. One top industry executive privately explained The nu clear vendors have supported the project to please their utility customers. The utilities, in turn, support the project because it costs them virtually nothing and because they view it as the ultimate test of government's commitment to nuclear power. If t h e federal government is willing to fund Clinch River, the utilities figure, it will be willing to fund anything nuclear including the utilities' own errors.Il fuel, private industry is far less'interested than it was in the early 1970s. Indeed, it seemed a t that time that reprocessing spent fuel from existing light water reactors in order to extract plutonium would be profitable by the early 1980s for use in light water reactors as an alternative to fresh uranium fuel. But the costs of reprocessing have ri s en far more than the cost of produc- ing fresh uranium fuel--to such a level, in fact, that private industry has been unwilling to finish even the one commercial reprocessing plant in Barnwell, South Carolina, which is over 50 percent complete As for prov i ding the Clinch River project with plutonium 7 The principal reason for this dearth of industry interest is the lack of any civilian market for plutonium DOE'S total annual plutonium research demand is only a small fraction of the 15 tons of plutonium the Barnwell plant is designed to produce per year. The second factor that has discouraged private industry from engaging in plutonium production is the high storage cost of the fuel. Industry experts estimate that it would cost between 1 to $3 per gram per y ear to store.

Barnwell plant this translates into storage liabilities in excess of 40 billion. So, in the absence of a clear commercial market for plutonium fuel, private industry is loath to assume the risks of proceeding further Over the lifetime of the In an effort to keep Barnwell alive, despite the industry's lack of interest, the federal government is willing to guarantee loans of up to 1 billion to corporations willing to finish work on the plant. The government has also promised to buy much, if not all, the plutonium from Barnwell for 'lfuturell breeder research It appears that the government may pay the price of Barnwell's plutonium at the higher level charged for weapons-grade material in the breeder deomonstration program by offering money for wo r k on a follow-on Large Demonstration Plant LDP) breeder, with little or no effort at cost sharing. The department is requesting 15 million for work on the LDP, and has divided the money such that each nuclear architect engineer in the United States will h a ve the opportunity of about the same amount of work to avoid favoritism, this operation is being coordinated not by any nuclear firm, but by Boeing, a company that has no significant reactor experience. This 15 million runs counter to the Office of Manage m ent and Budget's order last year that all work on the LDP must be funded entirely by the private sector DOE is also trying to bolster industry's flagging interest Possibly CONCLUSION A new energy technology should not be moved to the demonstra tion phase merely because it is feasible and may supply energy.

A demonstration should be funded only when it is clear that the technology can compete economically against existing alternative energy sources. This is true for any technology, not just for energy. Rese arch is one thing, an expensive demonstration is quite another. Research and development on a supersonic transport SST) airplane continues at NASA and the major aircraft manufact urers, for instance, but no demonstration project has gone forward because t he commercial potential of the SST is far from certain.

Demonstrating a design too early carries a stiff economic penalty, as the British and French governments are now learning.

They went ahead with the Concorde SST, which required massive subsidies and is now being abandoned as a commercial venture. 8 In choosing energy technologies to demonstrate, it makes far more sense first to concentrate work on projects likely to produce near-term benefits, and only then move on to those that are less likely to do so. Improved light water reactor fuels could save billions starting in 19

90. Advanced centrifuges and laser isotope enrichment could also bring enrichment costs down sharply in the next 20 years. Research and development and demonstration money should be focused on these projects first. While breeder research and development should continue, breeder demonstrators should wait until a commercially attractive design is developed reactor of some sort may be needed in the future, just when it should be needed remains unclear. Given this fact, the Clinch River Project, which will yield at best a dated technology, can no longer be justified. Whether the project is modified to bring it more in line with the state of the art in breeder technology or scrapped entir e ly, one thing is certain-it should not go ahead as planned. Like the SST, Clinch River represents a techno logy we could develop. However, also like the SST, it is a technology which will not be able to compete in the marketplace in the near future. Congr ess should not provide further funding to the demonstration project In the final analysis, it is clear that while a breeder Henry Sokolski Consultant