The Energy Spectator

Three Mile Island — Thirty Years After

If we haven't recovered by now, we never will. A special report.

By 3.31.09

Send to Kindle

On March 28, 1979, a minor valve failed in the cooling loop of a power plant on a sand bar in the middle of the Susquehanna River and the world became familiar with the name "Three Mile Island."

The country's first major nuclear accident was perhaps the greatest psychodrama of the era. For almost a week the world stood still and focused on Harrisburg, Pennsylvania. Reporters clamored into the area by the busload, one swearing he saw radiation dripping down the side of the reactor building like water. TV cameras remained riveted on the menacing cooling towers hovering over the rural landscape. The world held its breath while puzzled technicians wrestled with the mysterious technology, trying to figure out whether an elusive hydrogen bubble could expand into a nuclear holocaust.

Only a year before Canadian scientist Herbert Inhaber had issued a report claiming, through elaborate fault-tree analysis, that the chances of being killed in a nuclear accident were "less than the chances of being struck by a meteorite." (One of his fiercest critics was a young Berkeley physics professor named John Holdren, who has just been appointed Science Advisor to the President by Barack Obama.) Now it seemed we had already overstepped the boundaries of Dr. Inhaber's predictions. In a crowning irony, only three weeks before The China Syndrome, a Jane Fonda movie depicting a nuclear accident, had opened in theaters around the country. Harper's droll cartoonist David Suter would shortly mock Inhaber's style in the following feature:

We think the chance of THAT [i.e., a complete meltdown] is nearly infinitesimal, about one in 10[to the 4th] reactor-years of operation, or roughly the same as, say, the likelihood that a feature-length movie describing a reactor accident should open in New York just as a similar accident actually happens in a neighboring state.

So thirty years later, at a moment when the industry is all abuzz with talk of a "nuclear renaissance," what can we say about Three Mile Island, the lessons learned, and the concerns that remain? The following points, at least, have become clear:

1) Nuclear accidents can happen, but the consequences are much less than generally imagined.

2) The Three Mile Island accident was very much a part of the regulatory regime in which nuclear power had been developed. It had less to do with the technology.

3) The regimen under which nuclear plants operate today has improved immeasurably. The safety communications among reactor owners, for one, is orders of magnitude better than it was in the 1970s. This is both the result of industry efforts and the rise of merchant energy companies that now own and operate the majority of reactors.

4) In terms of overall environmental impact, nuclear power is undoubtedly the cleanest and safest form of electrical generation we have today. However, the failure to build new nuclear plants and the continuing operation of a few aging reactors creates the possibility that another Three Mile Island-type accident could occur.

Let's look at these one at a time:

Nuclear accidents can happen. A nuclear reactor cannot explodelike an atomic bomb. This has always been known by experts but little understood by the public. (During Three Mile Island, Herblock of the Washington Post drew a cartoon of a mushroom cloud rising out of the cooling tower.) The reason an explosion cannot occur is that the concentration of fissionable material in a reactor is nowhere near the necessary critical mass.

There are two isotopes of uranium, U-235 and U-238, only one of which (U-235) can split it two, giving an energy release. U-235 makes up only 0.7 percent of the natural ore. In order to get to "reactor grade," the isotopes must be separated -- a very difficult process -- and the ore "enriched" up to 3-4 percent U-235. (Iran has been trying to do this for year.)

In order to make a bomb, the uranium must be enriched up to -- can you guess? -- 90 percentU-235. At that point, two portions of a critical mass must be fired together at the speed of a cannon, merging in less than 0.00001 seconds so that the initial release of energy does not blow the mass apart. A nuclear reactor is not, nor never has been, nor never will be, a bomb.

If a reactor cannot explode, however, it can overheat. This is called a "meltdown." There is an important safety factor, however. There cannot be a runawayreactor because of an important fail-safe mechanism in all commercial plants. The water that bathes a reactor core serves two purposes -- a coolant and a "moderator," slowing down the neutrons so they can split other atoms. Without the moderator, the reaction cannot continue. If a reactor suffers a "loss of coolant," however, it also loses the moderator,which means the reaction stops. All that is left is the "decay heat," which will raise temperatures to melt the fuel rods but not penetrate the reactor vessel of the containment structure.

This fail-safe system did notexist as Chernobyl, which used graphite (carbon) as the moderator. When the Chernobyl core lost its water coolant, the reactor kept reacting. Moreover, carbon burns.The result was a runaway reactor that fired the carbon moderator and spewed radioactive debris all over the world. (As one last touch, the Soviets had failed to build a containment structure around the reactor.) Such an accident can never happen in an American reactor.

On the other hand, the catastrophic situation imagined in The China Syndrome never happened. Supposedly once the fuel assembly had melted, it would sink to the bottom of the steel reactor vessel and then keep going.It would melt through concrete containment and on down into the earth. As the actor-scientist explained to Jane Fonda after she witnesses a near-meltdown in The China Syndrome:

If the core is exposed for whatever reason, the fuel heats beyond core heat tolerance, in a matter of minutes, nothing can stop it and it melts right down through the bottom of the plant, theoretically to China. But of course as soon as it hits groundwater, it blasts into the atmosphere and sends out clouds of radioactivity. The number of people killed would depend on which way the wind is blowing, rendering an area the size of Pennsylvania uninhabitable, not to mention the cancers that would show up later. I may be wrong but I would say you're lucky to be alive. For that matter, I think we could say the same for the rest of Southern California.

(The irony of the Pennsylvania reference was not lost on anyone at the time.)

But of course it didn't happen -- nor did it ever make much sense in the first place. At Three Mile Island 60 percent of the core melted and dropped to the bottom of the reactor vessel, yet its heat wasn't even enough to penetrate the chromium lining of the reactor vessel, which has a lower melting point than steel.

But suppose by some wild and wacky happenstance the core did manage to melt through the steel and the concrete containment vessel and begin its journey to the center of the earth. Would it cause the steam explosion that would wipe out Pennsylvania?

A steam explosion occurs when you drop a superheated object into a vat of water. The water evaporates so quickly it acts as an explosion. As Bernard Cohen had long pointed out, however, a molten core making its way to China would not hit groundwater with anything near the same impact. At best it would sink a few yards per day. That would mean it might boil some groundwater, but this water would follow the core's tunnel right back into the containment structure. The result would be a geothermal site, where the earth's radioactive heat meets groundwater and sends steam shooting into the air. The China Syndrome was never anything but a cinematic fantasy.

Three Mile Island was an industrial accident. It ruined the reactor, caused a billion dollars worth of damage and nearly bankrupted the utility. What made it unusual for an industrial accident is that no one was hurt. The radioactive release -- caused when the seals on a steam overflow tank failed -- was minuscule. Exhaustive studies of the area have never found any health effects on the surrounding population. What Three Mile Island proved is that the worst-case scenario for a nuclear accident was far less than anyone realize.

The culture, rather than the technology, caused the accident. Within a year of Hiroshima, the federal government claimed a monopoly on all nuclear technology. The Atomic Energy Commission took control of research and reactor construction, while the utilities were not allowed to participate. After President Eisenhower's "Atoms for Peace" program, however, the AEC began civilian development. General, Westinghouse and Babcock and Wilcox started building reactors, although the AEC kept a tight rein on the technology.

When combined with the whims of state utility commissions, the results were a collection of one-of-a-kind reactors built around the country with very little in common. Each nuclear plant was an island unto itself. As Simpsons creator Matt Groening, who grew up with the ill-fated Trojan Reactor near Portland, once put it, "The manufacturers used to leave a note on the utility's doorstep saying, 'Congratulations, you are the owner of a new nuclear plant.'" The utilities communicated very little and shred no information. At Three Mile Island, the operating crew spent four hours per yearreviewing what was happening at other reactors.

A year after Three Mile Island, Reason published a brilliant article by Adam Reed explaining "Who Caused Three Mile Island." Reed noted that the AEC's culture of secrecy had kept nuclear technology completely isolated from a whole generation of engineering psychology that had evolved since 1945. As Reed wrote:

In the early 1950s, research established that about 80 percent of all industrial accidents were due, not to defects or malfunctions in industrial equipment, but to error and confusion on the part of human beings operating it.… Before the end of the '50s…insurance company safety consultants were bringing the new discipline of engineering psychology to bear on the design of industrial plant equipment.… By 1965, hardly any new equipment could be put into operation unless it conformed to the standards for safe human factors designed established by Underwriters Laboratory.… By 1970 no new design for a toaster or blender at General Electric could get off the drawing board without being examined by an expert in human factors. Yet the same company was designing, manufacturing, and delivering nuclear reactors that had never been seen, much less examined by an engineering psychologist.… It was only after the loss of the Three Mile Island plant in 1979 that engineering psychologists asked what the hell was going on in nuclear power plant control rooms. What they saw made them shiver.      

One of the major advances of industrial psychology had been making certain that control devices were clearly marked and differentiated, ideally requiring the operator to perform an act similar to what was being put into effect. A lever that moved the control rods up or down, for example, should move up or down itself in identical fashion.

What the engineers instead discovered in nuclear operating rooms was a sea of identical light and switches completely unrelated to their operating functions. Nor was there any hierarch of importance. As Samuel Walker wrote in Three Mile Island: A Nuclear Accident in Historical Perspective:

Within a few seconds after the accident began, the plant's alarm systems, including a loud horn and more than a hundred flashing lights on the control panels, announced the loss of feed-water in the secondary loop, the turbine trip, the reactor trip, and other abnormal events. But they offered little guidance about the cause of those occurrences and did not differentiate between trivial and vital problems.

In several reactors, lights and gauges indicating important information were so high up on the control panel that operators needed a ladder to see them. In one, two key switches that had to be thrown simultaneously were so far apart that it took two operators to perform the task. In another example that became infamous, the control rods were moved up or down by two identical, unmarked levers sitting side-by-side. The operators had so much trouble differentiating them that they eventually attached two different beer cans in order to remember which was which. At Three Mile Island, one crucial error occurred because a maintenance tag obscured a light warning of a serious malfunction.

The confusion in the operating room was mirrored by the lack of communication within the industry itself. The pilot relief valve, whose malfunction set off the Three Mile Island accident, had failed nine times previouslyat other reactors, including once at Toledo's problem-plagued Davis-Besse Reactor only a few months before. Yet none of the utilities communicated with each other and the manufacturer, Babcock and Wilcox, had not deemed the malfunction worthy of anybody's attention. As the Kemeny Commission, which investigated the accident, finally concluded: "[G]iven all the above deficiencies, we are convinced that an accident like Three Mile Island was eventually inevitable."

Safety and operating procedures at nuclear reactors have improved immeasurably. Prodded by both a draconian Nuclear Regulatory Commission and an outraged public, the industry began healing itself. INPO -- the Institute for Nuclear Power Operations -- a voluntary, industry-run operation in Atlanta, was set up to supervise and develop safety procedures. Although funded by the utilities, INPO remained independent and was given wide latitude in overseeing and inspecting reactor operation. The utilities were more than willing to cooperate. One more accident and they knew every reactor in the country would have to close down. As one executive put it, "We are all hostages to each other."

The philosophy of reactor operation changed completely. The original premise had been that the technology was too complicated for ordinary human beings but the engineers who designed reactors were such geniuses they could build them so no human being could possibly foul them up. Early operators were high school graduates. Indeed, the Three Mile Island accident occurred when the operators misread several signals and overrode safety measures that were preventing the meltdown.

Now things changed completely. Operator training became a five-year regimen more demanding than that of airline pilots. Five years experience in the Nuclear Navy -- where a lot of operators get their start -- barely got you in the door. Trainees now sat in classes for ten months before ever touching a valve. Then they did a two-to-five year apprenticeship in the control room before they could begin their licensing courses. They would then spend another 14 months on the job and in the simulator before taking the NRC-administered exam. Only then could they become licensed operators.

A simulator was required at every reactor. These were detail-by-detail duplications of the actual control room, serving both as classroom and laboratory for continuing research. Even after they have finished their five-year training, licensed operators are required to spend one week out of six in the simulator honing their skills and refining their knowledge of the plant. "We've made equipment changes in the real plant based on things that happened right in this simulator," says Mike Goskamp, who supervises training at Vermont Yankee. As a result of all this, reactor operations slowly improved over the 1980s and 1990s and there were no more serious incidents.

Things didn't really take off, however, until a group of new merchant energy companies entered the field in the mid-1990s. Deregulation of the electrical industry had suddenly turned most of the nation's 100-or-so reactors into "stranded assets" -- white elephants that were costing the utilities so much money it would make them incapable of competing in an unregulated market. Then Southern Utilities and Chicago's Commonwealth Edison, which had both improved their nuclear performance, decided to set up subsidiaries that would start buying distressed reactors.

"At the time, reactors were running at about 60 percent capacity, which was the standard for the industry," says Gary Taylor, CEO of Entergy. "It was a holdover from the coal days. You would shut them down every two weeks or so to give the boiler a rest. But we had some Navy guys on board who said, 'We run these reactors for five years at a time in the Navy. Why can't we do that here?'

"We soon found that most shut-downs had nothing to do with the nuclear side. It was the electrical equipment that kept breaking down. A turbine would trip or a wire would short out and you'd be offline for a few days. We started paying much more attention to this stuff and concentrating on keeping the plant up and running."

Soon capacity factors -- the portion of time the reactor is up and running -- began to creep up toward 80 percent -- unprecedented for the utility industry. Safety records improved correspondingly. "The more we paid attention to little things, the better these reactors ran and the better they ran, the more we could afford to pay attention to little things," says Taylor.

Refueling operations -- which usually took three months and were regarded as a vacation by plant employees -- became carefully choreographed operations taking three weeks and planned years ahead of time. Maintenance operations were scheduled to coincide with the shutdowns. Soon Entergy and Exelon had special teams touring the country to supervise refueling operations. The companies also set up special safety teams that can be dispatched to a reactor site at a moment's notice. "These days if a reactor incident occurs anywhere in this country, the whole industry knows it within a half-hour," says one executive.

By 2000 capacity factors were climbing toward 90 percent and reactors were running for 18 months without going offline. In 2002 the whole industry went over 90 percent and has remained there since. In the decade from 1998 to 2008 only one reactor -- the infamous Davis-Besse -- has shut down for more than a year because of safety problems. In the previous ten-year period 23 reactors had yearlong shutdowns and there were 22 in the decade before that.

America's fleet of 104 nuclear reactors now runs at a level of safety and efficiency unprecedented in any industry. Reactors now run for nearly two years without interruption. The record -- 688 straight days -- is held by Unit 1, Three Mile Island, the one that didn't melt down.

New construction is needed if this record is to be maintained. The only looming threat to this stellar recover is our failure to continue with the technology and build new reactors. The American nuclear industry has atrophied to the point where barely exists. Babcock and Wilcox has decided not to design any new reactors but will only maintain the ones it has already built. Westinghouse's Advanced Pressurized Boiling Water Reactor is a strong candidate for new construction, but the company was bought by Toshiba in 2006. On the other hand, General Electric's Economic Simplified Boiling Water Reactor (ESBWR) has fared poorly in Nuclear Regulatory Commission reviews and the last American manufacturer may be about to drop out of the business. Areva, the French giant, is now the major competitor in the American market.

This decline is reflected in the engineering profession. A whole generation of nuclear engineers and nuclear scientists eventually dropped out of the field because they saw no future in the field and because they got sick of trying to explain their profession to skeptical friends and relatives. There is hardly anyone in the nuclear industry these days under age 50 and this generation will be retiring soon. If their accumulated wisdom is to be passed on, we must have a revival soon.

Operating performance has improved so much that more than half the country's reactors have successfully applied to have their licenses extended for another twenty years. A few of these renewals, however, are approaching a danger zone. The Oyster Creek Reactor in New Jersey, for instance, the oldest operating reactor in the country, is now scheduled for an NRC relicensing decision in April. Completed in 1969 under a much earlier technology, the reactor is showing its age. Ideally, it would be shut down and replaced by a new one. Under present conditions, however, this is next to impossible. Oyster Creek provides New Jersey with 60 percent of its electricity. Shutting it down would cripple the New Jersey economy and make a mockery of Governor Jon Corzine's grandiose plans to power the state with wind and solar facilities.

Yet the continued operation of Oyster Creek poses a danger to the nuclear industry. One incident and we will be right back in 1979 and another Three Mile Island. The Nuclear Renaissance will be strangled in its cradle.

Despite being the object of scorn for three decades, nuclear energy delivers enormous benefits to the American public. With only 9 percent of the nation's generating capacity, it produces 20 percent of our electricity. Natural gas, on the other hand -- the favorite of environmentalists -- makes up 39 percent of our capacity but delivers only 20 percent of our electricity because the fuel is so expensive.

The nation's disdain for the nuclear industry and lack of will in constructing new reactors has put us in a precarious position. Our entire energy future may be riding on the fate of a few 40-year-old reactors. Somebody had better pay attention to this before we have another nuclear accident and the enormous promise of nuclear energy in this country ends for good.

(Portions of this article are excerpted from William Tucker's recent book, Terrestrial Energy.)

Like this Article

Print this Article

Print Article
About the Author
William Tucker is news editor for RealClearEnergy.org.