Alvin Weinberg: New Life for Nuclear Power

Introductory note: This short 2003 essay captures much of Alvin Weinberg's vision. - Charles Barton

From Issues in Science and Technology
Summer 2003

New Life for Nuclear Power
by ALVIN M. WEINBERG

Most of what I wrote in "Engineering in an Age of Anxiety" and "Energy Policy in an Age of Uncertainty" I still believe: Inherently safe nuclear energy technologies will continue to evolve; total U.S. energy output will rise more slowly than it has hitherto; and incrementalism will, at least in the short run, dominate our energy supply. However, my perspective has changed in some ways as the result of an emerging development in electricity generation: the remarkable extension of the lifetimes of many generating facilities, particularly nuclear reactors. If this trend continues, it could significantly alter the long-term prospect for nuclear energy.

This trend toward nuclear reactor "immortality" has become apparent in the past 20 years, and it has become clear that the projected lifetime of a reactor is far longer than we had estimated when we licensed these reactors for 30 to 40 years. Some 14 U.S. reactors have been relicensed, 16 others have applied for relicensing, and 18 more applications are expected by 2004. According to former Nuclear Regulatory Commission Chairman Richard Meserve, essentially all 103 U.S. power reactors will be relicensed for at least another 20 years.

Making a significant contribution to CO2 control would require a roughly 10-fold increase in the world's nuclear capacity. If nuclear reactors receive normal maintenance, they will "never" wear out, and this will profoundly affect the economic performance of the reactors. Time annihilates capital costs. The economic Achilles' heel of nuclear energy has been its high capital cost. In this respect, nuclear energy resembles renewable energy sources such as wind turbines, hydroelectric facilities, and photovoltaic cells, which have high capital costs but low operating expenses. If a reactor lasts beyond its amortization time, the burden of debt falls drastically. Indeed, according to one estimate, fully amortized nuclear reactors with total electricity production costs (operation and maintenance, fuel, and capital costs) below 2 cents per kilowatt hour are possible.

Electricity that inexpensive would make it economically feasible to power operations such as seawater desalinization, fulfilling a dream that was common in the early days of nuclear power. President Eisenhower proposed building nuclear-powered industrial complexes in the West Bank as a solution to the Middle East's water problem, and Sen. Howard Baker promulgated a "sense of the U.S. Senate" resolution authorizing a study of such complexes as part of a settlement of the Israel-Palestinian conflict.

If power reactors are virtually immortal, we have in principle achieved nuclear electricity "too cheap to meter." But there is a major catch. The very inexpensive electricity does not kick in until the reactor is fully amortized, which means that the generation that pays for the reactor is giving a gift of cheap electricity to the next generation. Because such altruism is not likely to drive investment, the task becomes to develop accounting or funding methods that will make it possible to build the generation capacity that will eventually be a virtually permanent part of society's infrastructure.

If the only benefit of these reactors is to produce less expensive electricity and the market is the only force driving investment, then we will not see a massive investment in nuclear power. But if immortal reactors by their very nature serve purposes that fall outside of the market economy, their original capital cost can be handled in the way that society pays for infrastructure.

Such a purpose has emerged in recent years: the need to limit CO2 emissions to protect against climate change. To a remarkable degree, the incentive to go nuclear has shifted from meeting future energy demand to controlling CO2. At an extremely low price, electricity uses could expand to include activities such as electrolysis to produce hydrogen. If the purpose of building reactors is CO2 control rather than producing electricity, then the issue of going nuclear is no longer a matter of simple economics. Just as the Tennessee Valley Authority's (TVA's) system of dams is justified by the public good of flood control, the system of reactors would be justified by the public good of CO2 control. And just as TVA is underwritten by the government, the future expansion of nuclear energy could, at the very least, be financed by federally guaranteed loans. Larry Foulke, president of the American Nuclear Society, has proposed the creation of an Energy Independence Security Agency, which would underwrite the construction of nuclear reactors whose primary purpose is to control CO2.

Making a significant contribution to CO2 control would require a roughly 10-fold increase in the world's nuclear capacity. Providing fissile material to fuel these thousands of reactors for an indefinite period would require the use of breeder reactors, a technology that is already available; or the extraction of uranium from seawater, a technology yet to be developed.

Is the vision of a worldwide system of as many as 4,000 reactors to be taken seriously? In 1944, Enrico Fermi himself warned that the future of nuclear energy depended on the public's acceptance of an energy source encumbered by radioactivity and closely linked to the production of nuclear weapons. Aware of these concerns, the early advocates of nuclear power formulated the Acheson-Lilienthal plan, which called for rigorous control of all nuclear activities by the International Atomic Energy Agency (IAEA). But is this enough to make the public willing to accept 4,000 large reactors? Princeton University's Harold Feiveson has already said that he would rather forego nuclear energy than accept the risk of nuclear weapons proliferation in a 4,000-reactor world.

I cannot concede that our ingenuity is unequal to living in a 4,000-reactor world. With thoughtful planning, we could manage the risks. I imagine having about 500 nuclear parks, each of which would have up to 10 reactors plus reprocessing facilities. The parks would be regulated and guarded by a much-strengthened IAEA.

What about the possibility of another Chernobyl? Certainly today's reactors are safer than yesterday's, but the possibility of an accident is real. Last year, alarming corrosion was found at Ohio's Davis Besse plant, apparently the result of a breakdown in the management and operating practices at the plant. Chernobyl and Davis Besse illustrate the point of Fermi's warning: Although nuclear energy has been a successful technology that now provides 20 percent of U.S. electricity, it is a demanding technology.

In addition to the risk of accidents, we face a growing possibility that nuclear material could fall into the hands of rogue states or terrorist groups and be used to create nuclear weapons. I disagree with Feiveson's conclusion that this risk is too great to bear. I believe that we can provide adequate security for 500 nuclear parks.

Is all this the fantasy of an aging nuclear pioneer? Possibly so. In any case, I won't be around to see how the 21st century deals with CO2 and nuclear energy. Nevertheless, this much seems clear: If we are to establish a proliferation-proof fleet of 500 nuclear parks, we will have to expand on the Acheson-Lillienthal plan in ways that will--as George Schultz observed in 1989--require all nations to relinquish some national sovereignty.