Abstract
Bad ideas seldom die: they simply go into hibernation, ready to burst forth when conditions ripen.
A decade ago, the transmutation of high-level nuclear waste was widely seen as a dead end. It was too complex, too congenial to proliferation, and too expensive. But today, transmutation research is again fashionable. At international energy conferences it is lauded as “bringing to the table new concepts that could be relevant for next-generation power producing systems” and as being “rather seductive to all of us” because it will require “new reprocessing techniques, new fuel developments, additional nuclear data, new reactors and irradiation facilities.” In 1999, Europe's Nuclear Energy Agency said transmutation research in Japan would help bring “the nuclear option into the twenty-first century in a healthy state.” And in January, Pete Domenici, New Mexico's Republican senator, secured $34 million to test the technology at Los Alamos National Laboratory.
What's going on? How did moribund, poorly coordinated research programs on transmutation spring back to life? In a word: politics.
Proponents claim that transmutation–once developed and refined–will help solve the nuclear waste problem. Transmutation has even been described by Los Alamos as the solution to the proliferation problem. Neither claim is likely to prove out.
Rather, a close look suggests that transmutation research in the field has been driven by political forces intent on propping up the nuclear power enterprise.
A great deal of work and money–perhaps hundreds of billions of dollars–would be needed to fully develop and build the types of reactors needed for transmutation. These “fast neutron” machines would include subcritical reactors driven by neutron-producing proton accelerators, as well as fast-neutron “pluto-nium burner” reactors–varieties of breeder reactors first conceived during the Manhattan Project.
Transmutation was intensively investigated in the 1980s and looked at again in the mid-1990s, and found wanting. It was an inefficient way to address problems in nuclear waste management; it was too expensive; and it presented serious proliferation concerns.
Although a 1996 study by the National Research Council of the National Academy of Sciences supported research on “selected topics” related to transmutation, it concluded that “None of the S&T [separation and transmutation] system concepts reviewed eliminates the need for a geologic repository.”
Regarding proliferation, the study said: “Widespread implementation of S&T systems could raise concerns of international proliferation risks … and require the United States to change its policy against reprocessing.” Indeed, the study concluded that “Proliferation risks would generally be greater with widespread implementation of S&T systems in the many nations using nuclear power.”
Finally, the study's support of limited research was in the context of further developing nuclear power, which it described as “possible applications of S&T as part of more efficient use of fissionable resources,” and for laboratory experiments relating to potential reduction in repository space requirements for nuclear waste.
Nearly a half century ago, nuclear power was viewed as the solution to many of humankind's ills, a source of power so plentiful that it would be “too cheap to meter.” But many billions of dollars later, the global nuclear industry has been greatly chastened.
Nuclear power has been frightfully expensive to develop and implement. It is so unforgiving of failure that catastrophic accidents on the order of Chernobyl are always possible. The nuclear industry has diffused nuclear technology globally, which has proliferation implications. (The Comprehensive Test Ban Treaty explicitly recognizes the proliferation problem by requiring all 44 countries with nuclear reactors of any kind to ratify the treaty before it can come into force.)
Given the difficulties in finding a
And finally, no one seems to know what to do with the highly radioactive waste produced by nuclear power plants, much of which–when measured by a human time frame–threatens to hang around more or less forever.
Over the past two decades in particular, the nuclear industry in the developed world has stumbled into a bottomless bog in one country after another. In the United States, nuclear power failed on Main Street in the aftermath of Three Mile Island, and then it took a dive on Wall Street because of staggeringly high capital costs.
Meanwhile, Japan and France, which had pinned their hopes on breeder reactors and a “plutonium economy,” have instead faced mounting technical problems and soaring costs. Breeders could make somewhat more fuel than they consume, but the reality is that the cost of ordinary uranium fuel used in “once-through” light-water reactors has been so low that breeders simply cannot compete. Despite more than $20 billion (in 1999 dollars) spent on building breeders worldwide over five decades, and billions more trying to operate them, the technology is plagued by technical problems.
To proponents of nuclear power, transmutation offers a glimmer of hope. Although the problems with transmutation are well known within the nuclear establishment, interest in the idea grew in the 1990s in France, the United States, and Japan. Collectively, these countries have about half the world's nuclear power reactors.
Proponents of transmutation claim that research and development could convert a theory into a practical and effective tool for long-term waste management. While they no longer suggest that transmutation would eliminate the need for a geologic repository, they claim that it could solve a great portion of the long-term waste management problem by converting long-lived radionuclides into radionuclides with relatively short half-lives.
In turn, that would bathe nuclear reactors in a more environmentally friendly light–and help finance the creation of new fast-neutron reactor technologies as a bonus.
Spent power-reactor fuel contains a number of radionuclides that have long half-lives–thousands to millions of years. Scientific analyses evaluating the possible consequences of spent fuel disposal over such long periods are fraught with uncertainties. Nevertheless, most experts agree that far into the future some leakage of long-lived radionuclides is likely.
It is also impossible to guarantee that geologic waste repositories–such as the one proposed at Yucca Mountain, adjacent to the Nevada Test Site–will be safe for tens of thousands of years from human intrusion, intentional or inadvertent. This is especially true, as with Yucca Mountain, where water is so limited that future generations may unknowingly drill too close to the repository in search of aquifers.
In the United States, which has a target date for opening its repository that could be as early as 2010–the earliest of any country–there are still no final environmental standards for the protection of future generations. Even the process for setting standards is mired in political and technical conflicts. It is noteworthy that the nuclear industry is resisting the inclusion of safe drinking water provisions in the proposed environmental regulations.
Given the difficulties and questions associated with repository siting–notably the extremely long periods of isolation required–it is no wonder that the transmutation of long-lived radionuclides into shortlived ones has attracted attention. If it worked, it could solve the waste problem by cutting down the isolation time in a repository from tens or hundreds of thousands of years to a few hundred.
Transmutation would not eliminate the need for long-term storage. Here, a worker looks out of the giant boring machine that has dug a five-mile tunnel through Yucca Mountain, in Nevada, a likely location for a nuclear waste repository.
Two kinds of transmutation reactions are important for waste management: neutron capture and fission. Either way, the goal is to transform long-lived radionuclides into short-lived ones. For example, the absorption of a neutron by the nucleus of radioactive iodine 129 (with a half-life of 16 million years) triggers a three-phase process by which it ultimately becomes stable, non-radioactive xenon gas.
To produce such a reaction, the iodine 129 must be placed for a sustained period in an environment where there is an abundance of neutrons. A nuclear reactor–in effect, a neutron-producing machine–can provide that environment. But neutrons can also be absorbed by shortlived or stable radionuclides, transforming them into long-lived ones–an outcome that needs to be avoided.
In contrast, the transmutation of heavy radionuclides such as plutoni-um and neptunium can be accomplished by fission, which converts them into shorter-lived radionuclides. Fission also yields an abundance of neutrons for more fissioning and for transmutation of long-lived fission products such as iodine 129 by neutron absorption.
Most fission products are shortlived, but the fission process itself generates some long-lived radionuclides, including iodine 129, technetium 99, and cesium 135. That leads to significant complications–the chief one being that spent fuel must undergo repeated reprocessing and separation to prevent shortlived radionuclides from being converted to long-lived radionuclides, thus defeating the purpose of the process.
Cheek-by-jowl with the reprocessing plant would be a fuel- and “target”-fabrication facility. Reprocessed and separated fuel and target elements would be loaded into a reactor designed for transmutation. Existing light-water reactors could play only a very limited role in transmutation, because their designs are not fully compatible with the necessary physics.
The preferred route would employ a fast-neutron reactor (such as a breeder reactor) or an accelerator-driven subcriti-cal reactor. These are the machines of choice because they produce highly energetic neutrons, which–in principle–could efficiently transmute a wide variety of radionuclides.
Further, with some fast-reactor designs, it might be possible to attach a reprocessing plant to the reactor. This would greatly reduce the transportation and storage requirements associated with the multiple reprocessing passes the fuel would have to make for a large fraction of long-lived radionuclides to be transmuted.
Finally, such reactors could accept relatively impure nuclear fuels, which in theory reduces proliferation concerns. In contrast to the situation with “weapon-grade” plutonium 239, it is more difficult to make bombs if the plutonium is contaminated with significant amounts of uranium and americium 241. Bombs made from contaminated plutonium would be unpredictable in yield.
The fabrication process would also be more dangerous because contaminated plutonium is more intensely radioactive than weapon-grade plutonium, thus making proliferation less likely in theory. But neither the unpredictable yields nor the greater health risks facing workers would be likely to deter those intent on acquiring a nuclear weapons capability.
France, the country on which most nuclear optimists pin their hopes for rekindling an interest in nuclear power, officially has high hopes for transmutation. After protests stopped its first attempt at selecting a repository site, France took the first and possibly most important step toward transmutation with the passage of a 1991 nuclear waste management law.
In fact, transmutation is the first of three methods of nuclear waste management being investigated under the 1991 law. The other two are storage and disposal. The French envisage using all three methods.
By the late 1980s and early 1990s, France's breeder reactor program was in deep trouble. A host of operating and technical difficulties had plagued its showcase project–the Superphénix, by far the world's largest breeder reactor.
Failing to get the reactor to run for extended periods at anywhere near its rated capacity, the French nuclear establishment faced a crisis. The Superphénix program had been the principal long-term justification for spent-fuel reprocessing.
Unlike conventional light-water reactors that use a once-through fuel cycle, breeders are designed to operate with plutonium fuel. This requires that plutonium be separated from spent uranium fuel. The plutonium is then fabricated into fresh fuel and fed back into the reactor.
Meanwhile, France's fallback position–separating plutonium from spent fuel so that it could be used in mixed oxide (mox) fuel in conventional light-water reactors–turned out to be far more expensive than anticipated. Contrary to expectations, ordinary low-enriched uranium fuel has remained cheap, making traditional once-through fuel use relatively attractive.
By the late 1980s, Electricité de France, France's nationalized electric utility, was reluctant to continue its money-losing reprocessing contracts. It might have bowed out, except that it did not want a major public debate about the economics of nuclear power.
A similar situation emerged in the United States with cancellation of the Integral Fast Reactor (ifr), the last hope of the U.S. nuclear establishment for keeping breeder reactor and reprocessing technology alive.
(The ifr design combined power production and reprocessing in a single facility, hence the name. The specific electrolytic reprocessing technology used for this reactor is variously called pyroprocessing, electro-refining, or electrometallur-gical processing.)
The ifr project was killed in 1994 by the Clinton administration as an unneeded energy technology research project. But pyroprocessing survived the cut as a “waste management” technology and became part of a research program based at Ar-gonne National Laboratory-West in Idaho. The technology was also used in Los Alamos National Laboratory's proposal for using accelerator-driven subcritical reactors for transmutation.
Japan's situation is not unlike that of France, except that it has far less to show for comparable expenditures on its plutonium program. Its demonstration breeder reactor, Monju, which cost more than $5 billion, suffered a sodium leak and fire in 1995, less than two years after the reactor went critical. It has been shutdown since then.
Japan's reprocessing program has been plagued with technical problems and high costs. There was a fire at a waste facility at its demonstration reprocessing plant in 1997. Meanwhile, the full-scale Rokkasho reprocessing plant, still under construction with a start-up date of 2005, is now estimated to cost about $20 billion. The cost is so large that its builders cannot hope to ever operate it on a commercially competitive basis.
In short, we believe it is no accident that transmutation of high-level waste has found champions in those parts of the nuclear establishment most attached to reprocessing and breeder reactors. Transmutation programs would revive hopes for reprocessing in all three countries, and they would provide hundreds of billions of dollars for subsidized reactors.
Trouble is, neutrons can be
The Energy Department believes that existing U.S. nuclear power reactors will have generated 87,000 metric tons of spent fuel by the end of their service lives. Using transmutation to reduce the amount of only a few of the 20 long-lived radionu-clides in those 87,000 metric tons would cost about $300 billion, according to the department's “Road-map for Developing Accelerator Transmutation of Waste,” a report to Congress issued in October 1999. The “Roadmap” estimates that much of the cost would eventually be recovered from the sale of electricity produced by transmutation-capable reactors, reducing the net cost over a period of several decades to about $14 billion.
(Most of the $40 billion pricetag for a geologic repository would have to be added to that. Even with transmutation, a repository would be needed.)
But the Energy Department's record for coming up with reasonably accurate cost estimates for new technologies is not good. For instance, the initial cost estimate for the laser fusion machine now being built at Lawrence Livermore National Laboratory was $2.1 billion. That was just a “come-on” price. The current General Accounting Office estimate is $3.9 billion. Moreover, the construction schedule has slipped by six years and the project is mired in technical difficulties.
The National Academy of Sciences study on transmutation criticized the kinds of optimistic estimates for reprocessing costs that have been used by the Energy Department. It estimated the excess cost of transmutation compared to direct disposal as “no less than $50 billion” and said that costs “easily could be over $100 billion.”
The global total of spent fuel is about three times the amount in the United States. This provides the basis for a very rough extrapolation. We estimate that the worldwide cost of transmutation of spent fuel from the current stock of power reactors could be roughly a trillion dollars, possibly more. Net costs, after electricity sales are taken into account, could be hundreds of billions of dollars if the reactors work as hoped–an outcome that is far from guaranteed, given the troubled history of fast reactors.
Even if tens or even hundreds of billions of dollars were spent worldwide, transmutation would not be able to convert most long-lived ra-dionuclides into shortlived ones. For example, uranium 238, which constitutes 94 percent of all spent fuel and which has a half-life of 4.5 billion years, is not a significant part of any transmutation scheme.
Theoretically, a vast number of breeder reactors could be built to convert a portion of the uranium 238 into plu-tonium 239. Then the plutonium and unused uranium could be separated and put back into the reactor. The plutonium could be used as fuel to produce more plutonium in the uranium. But such reprocessing would have to continue over dozens of cycles.
That is not going to happen. It would create the ultimate “plutonium economy,” which would last hundreds of years–the very outcome that U.S. nonproliferation policy has sought to avoid for more than two decades.
Moreover, complete recycling of uranium is not technologically feasible–its repeated reintroduction into a reactor causes it to become so contaminated with unwanted isotopes that fuel fabrication becomes costly and dangerous.
For engineering, cost, and safety reasons, it would be necessary to discard most contaminated uranium as a waste in any transmutation scheme. That means the cost of a deep geologic repository would still be borne even under a massive transmutation regime. Without transmutation, the cost of burial is roughly estimated to be $40 billion. Most of this is fixed cost, which would be incurred even with transmutation.
Transmutation also faces basic engineering problems in dealing with most of the long-lived fission products (such as cesium 135) and activation products (such as carbon 14). For instance, the engineering challenges of separating cesium 135, the long-lived isotope, from the other cesium isotopes are so great and so insurmountable that no one has even proposed research into the problem.
Cost questions aside, the implementation of any transmutation program would also have implications for nuclear proliferation, the environment, human health, and reactor safety.
•
This is true. Nonetheless, pyro-processed plutonium would still be usable in weapons–and it would be entirely satisfactory to a terrorist group that was not fussy about predicting the exact yield of a bomb or about subjecting workers to health hazards during the manufacture of weapons.
Moreover, although pyroprocessing is often labeled “proliferation resistant,” it is a far more compact technology than conventional chemical reprocessing, and therefore more readily hidden.
Finally, the separation of isotopes such as neptunium 237 and americi-um 241 would also increase proliferation risks because both can be used to make nuclear weapons.
•
Historically, reprocessing has led to large discharges of radioactive wastes into the environment. Several European countries are currently demanding that Britain and France stop discharging chemical reprocessing wastes into European waters. Because it seems impossible to stop the discharges, many European countries are now demanding an end to commercial reprocessing.
•
Of the 11 medium and large breed-er reactors that have been built worldwide, one did not open, one remains closed because of an accident, five are permanently shut (one because of a partial meltdown), one will soon close, and one is on standby.
Only two breeders–a relatively small reactor in Japan and the Russian BN-600, which has used primarily uranium fuel (rather than the plutonium fuel breeders were conceived for)–are expected to continue operating beyond a few years.
While accelerator-driven reactors have been described as “inherently safe,” that is misleading. For proposed accelerator-based systems, the ability to shut off the neutron source and the fact that the reactor ordinarily would be subcritical would provide a margin of safety. On the other hand, these systems would rely strongly on the ability to shut off the neutron source in an emergency. Like any other human system, that shut-off system could fail.
Also, it may be necessary to ensure that the external neutron source is not operating at full power when fresh fuel is in the reactor. If not, the reactor could go supercritical.
The Energy Department has such low confidence in the possibility that workable and cost-effective transmutation schemes could be developed that it has not explicitly requested development funds for it. However, transmutation expenditures have been hidden under other headings.
Energy Department laboratories in Los Alamos and Idaho are pursuing transmutation, partly because Congress has seen fit to add millions of dollars of unrequested money to Energy's budget for that purpose.
In fiscal 2000, Congress added $9 million in unrequested funds for transmutation research. In the current fiscal year, a new office of Advanced Accelerator Applications has been created with $34 million in funding. Most of this money is headed for Los Alamos, but $3 million is intended for the University of Nevada-Las Vegas for pyroprocessing.
Congressional delegations from New Mexico, Nevada, and Idaho have been especially active in promoting funding. In particular, Republican Sen. Pete Domenici of New Mexico, home to Los Alamos and Sandia National Laboratories, is a well-placed and vigorous advocate of transmutation, both for its supposed energy and its waste management value.
Domenici, who chairs the Senate Energy and Water Appropriations Subcommittee, said in 1999: “Solely relying on underground storage ignores the fact that this spent fuel still has most of its energy potential. This funding [for transmutation research] will enable the nation to reexamine the advisability of continuing the current path for spent fuel.
“Transmutation technologies could enable energy recovery, along with significant reduction in the toxi-city of the resulting final waste. However, while transmutation is technically feasible, much research and development will be required to determine its economic implications.”
In France, the relationship between the industry and government is even closer. Indeed, the French nuclear lobby resides within the heart of the government. The Commissariat a l'énergie atomique, France's counterpart to the old U.S. Atomic Energy Commission, is an enthusiastic proponent of transmutation.
The world's largest commercial reprocessing company, Cogéma–which would benefit most from the reprocessing aspects of transmutation–is 81 percent government owned. France's only electric utility is entirely government owned. No wonder that the annual budget for various transmutation-related items adds up even in the preliminary stage to roughly $100 million. Last year, Japan devoted $29 million to transmutation-related projects.
The age of nuclear power began with a promise that nuclear energy would be “too cheap to meter.” It was a promise that even the official and industry studies of the time showed was false.
That promise has, among other things, led to an enormous burden in long-lived nuclear waste, a problem that has no good solution. Instead of facing up to the terrible burden that the first half-century of nuclear power will impose on future generations, the votaries of nuclear power technology are trying to persuade the public to cough up billions more in the pursuit of another nuclear chimera, transmutation.