Keeping High-Tech Machines at Low Risk

A LATE-SUMMER sun cut through the clouds as workers at a nuclear plant in the heart of North America's breadbasket went about their tasks, maintaining and refueling the 1,170-megawatt reactor at Wolf Creek, Kan.

Suddenly a control-room alarm sounded. Through a misopened valve, extremely hot water under high pressure was surging from the reactor's cooling system into a nearly empty holding tank. In 66 seconds, the reactor lost 9,200 gallons of water - 10 percent of its coolant.

Had the valve taken 3 to 5 minutes longer to close, the accident would have disabled the reactor's emergency cooling system, according to analyses by the plant and by the US Nuclear Regulatory Commission. In another half hour, the core would have been left high, dry, and hot, leading to a core meltdown and a release of radiation - just an hour and a half's drive from the state capital, Topeka.

The accident represented an unforeseen weakness in the reactor's design that may be common to other plants in the US, according to the NRC. And as such, Wolf Creek has one important thing in common with other accidents involving complex technologies: They exposed a weakness in design that adds to the level of risk involved in the operation.

Many observers compare risk at nuclear plants to that of other complex machines, such as the Challenger space shuttle. In that case too, the accident uncovered a design flaw.

"If we focus on operator error or the response of mid-level managers, we miss the big picture. It deflects attention away from the risks" of the technology involved, says Diane Vaughan, a professor of sociology at Boston College, whose book "The Challenger Launch Decision" details the failures that led to the shuttle disaster.

"If you look at the [estimated] probability of a shuttle accident before and after the Challenger accident happened, it's striking," adds Arjun Makhijani, president of the Institute for Energy and Environmental Research, a nonprofit organization based in Tacoma Park, Md. "Before Challenger, the risk of losing a shuttle was put at 1 in a million launches; after the accident, it fell to 1 in 100."

The risks associated with serious nuclear accidents likewise shifted after the accident at the Three Mile Island nuclear plant outside of Harrisburg, Pa., in 1979. Before TMI, the likelihood of a major accident at a nuclear plant was put at 1 in 10 million. "After TMI," he adds, "the risk estimates increased by three orders of magnitude."

The difference between spaceflight and nuclear power, he explains, is that the people who climb into the shuttle's cabin do so by choice, knowing that what they are doing is risky. "Chernobyl affected millions of people who had no sense of the risk and no choices in accepting it," Mr. Makhijani says.

These issues are now looming largest in Asia, the region anticipating the largest increase in new nuclear plants. The Pacific Rim is experiencing rapid economic growth, says Wilfrid Kohl, director of Johns Hopkins University's International Energy and Environmental Program in Washington.

"Electricity has been the fastest-growing fuel sector in the industrial and developing countries" in Asia because most of their economies are shifting from heavy industries to service and high-tech companies," says Mr. Kohl. "High-tech oriented societies are willing to place their bet on nuclear" to make up for a lack of indigenous energy sources and to ensure energy security.

Yet even in Asia, the growth in nuclear energy may be less spectacular than anticipated, according to Ronald Hagen, an energy analyst at the East-West Center in Honolulu.

Over the last few years, for example, public opposition to nuclear energy in Japan has stiffened following several accidents. During the most recent mishap, the loss of 3 tons of coolant at the $5.7 billion Monju reactor, plant officials were caught trying to hide the severity of the problem from the public.

"If you look at the schedule for nuclear plants in Japan, there is a mysterious three-year gap where nothing is scheduled to begin," notes Mr. Hagen, suggesting that there may be a policy shift to build fewer plants. He also notes that companies supplying equipment to the nuclear industry are shifting workers to nonnuclear jobs.

In the meantime, the first glimmers of competition are appearing, with the emergence of independent power producers. These new entries into the field are shunning nuclear plants for those using fossil fuels. As these companies emerge and take root, they could subject nuclear utilities in Japan to the same economic pressures expected to develop in the US.

Experts inside and outside the nuclear industry in the US are worried about the potential for nuclear plant operators to cut corners on safety to keep their prices competitive - possibly introducing other risks in using nuclear technology.

"As a society, we need to remember that there will be certain costs" to using complex technologies, "even under the best of circumstances," says Dr. Vaughan.

To reduce risk, it is important to become more imaginative about the ways technologies can fail, says Charles Perrow, a Yale sociologist who has written about managing high-risk technologies. Systems are engineered and regulated with specific accidents in mind. Yet when systems fail, it often is because elements malfunction in ways no one anticipated. "We need to examine the complexity of our technologies and imagine even the crazy things that could happen," says Dr. Perrow.

As for nuclear energy, Western reactors are generally safer than those in the former Soviet bloc, adds Makhijani. "But that doesn't mean that a Chernobyl is impossible here."