LIVERMORE, California — Thirty years ago, half the core of a reactor at the Three Mile Island nuclear complex melted down, but government officials and the utility running the place didn’t know that. And they wouldn’t know for six more years.
In fact, as the crisis extended from its start on March 28, 1979, the amount of information available about the nature of the accident remained slim. Key pieces of data were missing. Nobody knew exactly what was happening inside the containment vessel and, more importantly, what was coming out of it. The sensors designed to measure radioactive release were overwhelmed.
The limitations of data collection and computation made precise predictions and good decisions difficult to make. As intermittent emissions of radioactive gas tumbled into the sky, the uncertainty about how severe the accident had been rose with them.
The information vacuum didn’t just impact what was happening on the ground during that frightening week. It has rippled down through time as the debate over what the Three Mile Island accident meant to society became clouded with doubts over what actually happened out in the Susquehanna River. Both pro- and anti-nuclear groups used the accident to illustrate their points, but material facts about the reactor’s core were still coming out six years after the partial meltdown.
“The instrumentation that was put in that plant and in several other plants at the same time, was not designed to handle the scope, the scale of the release that happened at that plant. Even though it was small, it was beyond the range of the instruments that were in the containment vessel,” said Tom Sullivan, a former Livermore scientist who helped respond to the disaster. “So, what happened is that the accident moved into the realm of the unknown and as soon as that happened people started to imagine things.”
If a Three Mile Island-like accident happened today, we have a datacenter-full of new capabilities to assess any damage. The accident, in part, called into being a new force within the Department of Energy: a group of meteorologists whose sole job is to model how dangerous particles from anything from nuclear fallout to chemical explosions will cross a landscape. Headquartered at Lawrence Livermore National Laboratory, the National Atmospheric Release Advisory Center is an under-appreciated link in our chain of defenses against nuclear and chemical accidents or attacks.
We only truly built its capability in the aftermath of the worst nuclear power accident in American history. In one of the most underrated outcomes of the disaster, Three Mile Island confirmed the value of having better data, and the geeks who could understand it.
Without it, the authorities found themselves scrambling. The utility was trying to figure out what had gone wrong. The public was fearful and angry. The government was in damage-control mode. Everyone wanted to know the thing that no one could provide: Exactly how much of the tasteless, odorless, invisible radiation was released and where was it heading? The impact of the confusion would have much longer-lasting impacts than anyone could have imagined.
|
Walter Cronkite led off the CBS Evening News broadcast by delivering the bad news to the world.
“The world has never known a day quite like today,” he said from beneath his bushy eyebrows. “It faced the considerable uncertainties and dangers of the worst nuclear power plant accident of the atomic age.”
Those words reflected the terror that a fundamentally new kind of danger had entered the world, one that we manifestly were not prepared to measure. Into the data vacuum rushed fear and DOE staff.
Marv Dickerson, another former Livermore scientist, got the call in the United Red Carpet Room at O’Hare Airport.
“We cleared out an area and started work right there in the airport,” Dickerson said. Shortly thereafter, he was on a plane bound for Harrisburg, where he’d become intimately involved in the response to the accident, briefing decision makers.
Meanwhile, Sullivan headed back to Livermore to crank out calculations based on the data that was beginning to flow into the system. They pushed everyone else off the Lab’s computers and started working. The setup back then was crude.
“The big computers were at the other end of the lab, and we had no electronic connections. People would run back and forth with a big reel of tape,” Sullivan said, recalling a time when memory was stored in BPI, or bits per inch of tape. “We would run it from the big computers to the little computers we had that would then make [a map] that we would make a hard copy of, hand label, and put in a fax machine. Talk about primitive.”
Even with that ’70s setup, which is about as foreign to us now as a polyester leisure suit, they were able to produce a rudimentary model of the release in Harrisburg, but it was missing some key details.
“When we first responded to Three Mile Island, we didn’t even have topography in the model,” Sullivan said. “Basically it was a Flat Earth Society model. We were very worried about that. The first 48 hours, we probably put in 24 man hours trying to get the topography into the little grid.”
There was no accounting for turbulence, and the resolution wasn’t too high, either. But their projections were still immensely valuable in guiding the measurement taking of helicopters and ground crews.
At Livermore’s modern NARAC facility, they are several orders of magnitude more precise in their calculations now, and they’ve responded to more than 150 events. What used to take hours now takes minutes. Real-time data availability might not have headed off the divisive clash over what happened at Three Mile Island, but it would have helped protect citizens while providing a common basis for argument over the atomic energy’s risks.
As it was, the confusion made having a real conversation about the meaning of the accident at Three Mile Island’s second reactor impossible. While the accident was certainly not as bad as the darkest fears of some, it turned out to be considerably worse than industry and government officials believed (or let on) at first.
Richard Lyons, reporting on the front page of The New York Times, wrote that John G. Herbein, Metropolitan Edison vice president, said “that the mishap was not all that unusual in his eyes since similar accidents had happened ‘two or three times’ to the first reactor unit at the plant that opened in 1974. But he conceded that there was one major difference: the leak of contaminated water into the building next to the reactor this time.”
It’s clear, in retrospect, that Herbein didn’t really understand what had happened in the three-month old reactor.
“It was a bit of an engineering nightmare because they didn’t understand how the plant was functioning at that time,” Sullivan said.
At first, utility and government officials contended that only 180 or maybe 360 of the 36,000 containment rods had melted. Those numbers were up to 9,000 rods shortly after the accident, but the cooler heads prevailed a little too well in this case. For a while in the early 1980s, nuclear power industry officials maintained that no melting had occurred — and that was a major reason for claiming the coverage of the affair was overblown.
“Little, if any, fuel melting occurred, even though the reactor core was uncovered. The safety systems functioned reliably,” said D.B. Trauger, a nuclear engineer at Oak Ridge National Laboratory, at an engineering conference eight months after the accident. “Based on the conservative licensing analyses, the core was subjected to conditions that would have produced a total melt…. This accident has revealed that reactors are orders of magnitude safer than previously assumed.”
In fact, it took excavation of the containment vessel in the mid-’80s to glimpse the true extent of the damage. What workers found was shocking. The Washington Post‘s Three Mile Island timeline summarized the new severity estimates, “Core temperatures reached 5,000 degrees Fahrenheit; as much as 50 percent of the fuel melted.”
Today, the Nuclear Regulatory Commission recognizes that half the core melted.
On the other hand, fears of massive radiation releases and increased cancer, though they continue to surface, have not played out. No one died from the initial accident and the official line is that no deaths can be statistically attributed to the radiation released in the area.
In other words, both sides got pieces of the story right, but neither of them could have known exactly what was happening at the time of the event. What’s more unclear was the impact that TMI had on the nuclear industry as a whole. The public reaction to the accident was clearly real, but Three Mile Island’s importance could be overestimated.
While it’s true no new nuclear plant has been built since the accident, the industry’s problems began years earlier, and the chief reason for its declining momentum in the United States was economic.
The myth continues, though, because it serves everyone to blame the nuclear industry’s problems on the meltdown and subsequent increased anti-nuclear power protests. Environmental groups get to claim victory in slowing nuclear construction projects and the nuclear power industry gets to blame their failure to fulfill the promise of “the nuclear age” on the irrational response to a “minor” accident.
But the industry was showing major signs of weakness before the accident. Plants were turning out to be expensive and taking exceedingly long times to build. There was (and remains) no solution for the long-term storage of radioactive waste. And the public had already begun to lose trust in the government-industrial complex to regulate itself adequately.
An anti-nuke protester in Harrisburg, Pennsylvania.
On Feb. 27, 1979, two months before the accident, influential New York Times columnist, Tom Wicker, noted the trouble the industry was having, particularly with the credibility of its safety assessments.
“Now the industry is paying the price in reduced public support and confidence and a nuclear energy construction program slowed almost to a standstill,” Wicker wrote.
Regulators had touted a landmark 1975 safety study, which Marcus Rowden, a Nuclear Regulatory commissioner said showed “the risks from potential nuclear accidents would be comparable to those from meteorites.” But as doubts continued to surface about the report’s methodology and intellectual independence, the NRC was forced to back away from the report. Later, a 1982 report found that the old safety study had underestimated the risk of a nuclear accident by a factor of 20.
Perhaps, if there is a nuclear renaissance, the increase in our computational power and the commitment to transparency that the Obama government espouses will create a more rational national discourse about the atomic age. The New York Times opinion page, in the wake of Three Mile Island, actually struck exactly the right tone in its critique of the industry, “Credibility Meltdown.”
“For years, the industry and its supporters in Government and elsewhere have insisted that nuclear energy is safe. Relatively speaking, they are right. But under the pressure of emotional protests, the nuclear spokesmen have gone further. Nuclear power, they have proclaimed, is The Answer; it is not just reasonably safe but comfortingly so; a serious accident is a million-to-one shot,” they wrote.
“Then what are people to think when a Three Mile Island comes along? This event may turn out to be not so serious, but when safety is oversold, even the smallest accident delivers an exaggerated blow to industry credibility. As a result, a public that has been educated to think that there are virtually no risks involved in the use of nuclear energy will find it that much harder to take reasonable ones in its pursuit.” See Also:
Images: All images drawn from The National Archives.
WiSci 2.0: Alexis Madrigal’s Twitter, Google Reader feed, and book site for The History of Our Future; Wired Science on Facebook.
