Retooling a Leviathan
The challenges of keeping an aging nuclear infrastructure alive
Sometime in the late fall or early spring of 2009, a team of engineering experts from the Bechtel Corporation will cut a hole 28 feet square in the four-foot-thick concrete dome that contains the San Onofre Nuclear Generating Station Unit II. After freeing the unit’s steam generator from its moorings, they will lift its 700 tons of metal through the top of the dome with a specially designed crane, and replace it with a new generator hauled from Japan. Then the hole will be closed, its steel reinforcements welded shut and its protective concrete replaced. Ninety to 120 days after the process began, at a cost of over $700 million, steam will once again race through the generator’s super-alloy tubing to the spinning turbines that power a million Southern California homes.
Two years later, Bechtel will have to go through the whole process again: San Onofre’s Unit III generator is in only slightly better shape than its sibling.
For the first time since 1979, when the near-meltdown at Three Mile Island dashed what was left of the nuclear industry’s fading ambitions, nuclear power in the United States is experiencing a mini-renaissance. But in order just to keep the current nuclear capacity going beyond 2020, dozens of aging plants like San Onofre must be massively overhauled. And that’s nothing compared to the enormous expense of building new power plants, a necessary step if the industry is to grow, or even survive, beyond the next 50 years.
The bathtub curve
Engineers often employ a curve shaped like a bathtub to describe the reliability of machinery over its lifespan. If you plot the incidence of mishaps over time, the curve begins at a high point, during the breaking-in phase, and then drops sharply as the machine — whether coffeemaker or airplane — enters its relatively stable middle years. Then, as the machinery begins to age, the line climbs at the same angle that it once dropped. The 103 nuclear reactors across the country — which supply one-fifth of the nation’s power — have been headed up the back end of the bathtub curve for years.
Last spring, San Onofre’s second generator shut down for 108 days when routine maintenance revealed that a corroded gasket had uncoiled into an injection tank. More recently, San Onofre Unit I, which has been shut down since 1992, was found to be leaking small doses of radioactive tritium into the water it discharges into the Pacific Ocean, and possibly into the groundwater that its landlord, Camp Pendleton Marine Base, uses for drinking.
An even bigger problem with San Onofre’s reactors, however — and with 72 other similar reactors around the country — is that Inconel-600, the nickel-based superalloy metal used to fabricate their steam-generator tubes, has proved more vulnerable to cracks than anyone anticipated. Ray Golden, Southern California Edison’s public information officer for San Onofre, has a way of impressing even a hardened skeptic with the wonders of radiation. It’s only scary when you don’t understand how it works, he says. But even Golden can’t talk about cracks in steam-generator tubes without sounding a little worried: Tubes of the same type burst two years ago at a plant in Mihuma, Japan, killing four workers instantly with 392-degree Fahrenheit steam.
Like the Mihuma reactor, San Onofre is a pressurized light water reactor, or PWR, most of which were built by Westinghouse in the late 1970s. PWRs use the heat created by the controlled splitting of uranium atoms to produce steam to spin turbines. To regulate the temperature of the reactor core, pressurized water circulates through the coolant loop, made of tubes three-quarters of an inch in diameter and about as thick as a dime. The Nuclear Regulatory Commission allows plants to operate with microscopic fissures in up to 21 percent of those tubes; San Onofre II has cracks in 13 to 14 percent.
"We’ve taken conservative measures at every refueling outage to inspect these tubes and take them out of service before problems arise," says Golden, a linebacker-big, red-haired man who has spent 23 of his 45 years explaining nuclear power to the media and public. "But if we don’t replace the entire generator, we’ll be looking at a shutdown in 2010. Unit III would last a few more years, but neither generator would make it to the end of its licensing period in 2022." Golden expects the alloy in the new generators, the tougher Inconel-690, to last well into a second 20-year license extension — not that SCE has asked for one. "We haven’t made that decision yet," he says.
Subsidies provide life support
The San Onofre plant produces 2,200 megawatts. When SCE compared the cost of replacing the steam generators to building the two new natural gas power plants required to put out that much power, the $700 million replacement operation looked cheap: Constructing and fueling two natural gas plants over the next dozen years would cost $1.6 billion.
But it’s one thing to rehabilitate an aging plant to keep it running, quite another to construct and fire up a whole new plant. Although the operating costs of nuclear generation compare favorably to other power sources (1.5 cents to about 3 cents for natural gas), its startup costs do not: A 2003 Massachusetts Institute of Technology study put the price of a nuclear-generated kilowatt hour, amortized over a 25-year-period, at 3 cents higher than one produced by coal or natural gas.
But that 3 cents disappears, some analysts say, when you factor in the toll fossil-fuel pollution exacts upon the atmosphere. "The Need for Nuclear Power," an influential article by Richard Rhodes and Denis Beller that appeared in Foreign Affairs magazine in 2001, points out that nuclear plants cost a lot to build only because their radioactivity has to be contained in a leak-proof structure. "If fossil-fuel plants were similarly required to sequester the pollutants they generate," write Rhodes and Beller, "they would cost significantly more than nuclear power plants do."
It’s a compelling argument, one that may prove more true as the United States moves to regulate carbon dioxide emissions. At present, however, nuclear generation still looks like a shaky deal. Never in its 50-year history has it survived without generous government subsidies such as the Price-Anderson Act, a federal insurance policy for nuclear reactors that limits a utility’s liability to $300 million. The Energy Policy Act of 2005 gives new nuclear plants a 1.8 cent-per-kilowatt-hour production tax credit during their first six years of operation. Should regulatory snafus delay their opening, the first six new plants built will be eligible for $2 billion in compensation.
Those subsidies have given nuclear power, which a few years ago seemed on its way to the grave, a bit of hope: Twenty-seven companies have told the Nuclear Regulatory Commission that they want to build plants. Regardless, many utility officers remain unconvinced, and plan on investing in cleaning up their coal plants rather than going nuclear.
Amory Lovins, the director of the Rocky Mountain Institute, would rather see the subsidies go to renewable energy, believing that even government life support can’t make up for all of nuclear’s economic shortfalls. "You can make a corpse jump by defibrillating it," Lovins says. "But you can’t bring it back to life."
Judith Lewis is a senior editor at the L.A. Weekly, where she covers public health and the environment.
This story is a sidebar to the feature:
With global warming an increasing threat, some are urging a return to nuclear energy, but the industry’s own checkered past reminds us that a nuclear renaissance will be neither easy nor cheap