After the Meltdown,
Lessons From a Cleanup

Wreckage at Three Mile Island yields 150 tons of debris.


New York Times
April 24, 1990

AFTER the most arduous cleanup effort in nuclear history, workers at the Three Mile Island nuclear plant last week finished shipping 150 tons of radioactive wreckage from the damaged reactor vessel. They are now preparing, 11 years after the accident, to begin monitoring the plant over the long term.

Their painstaking chipping, sifting and packing of the rubble from the nation's worst nuclear accident has shed new light on the course of events in the nuclear core, which were largely hidden at the time because of misleading instruments and ill-trained operators. The cleanup has yielded new insights into the strengths and weaknesses of nuclear plants, and the unforeseen difficulties of cleaning up after destroying the reactor core.

when the last shipment of debris from Three Mile Island arrived in Idaho last Thursday for long-term storage at the Department of Energy's Idaho National Engineering Laboratory, the first major phase of the cleanup was over. Engineers at the headquarters of the Nuclear Regulatory Commission in suburban Washington say that getting to the bottom of the mess has demonstrated one cardinal point: it is far easier to wreck a reactor core than any of them had believed, but apparently harder to push an accident to the next step, rupturing the reactor vessel, the great steel pot that holds the core and forms a major line of defense against radiation release.

The cleanup process itself also reemphasized the principle that the nuclear field is especially vulnerable Murphy's law, which states that anything that can go wrong will go wrong. The cleanup was unexpectedly plagued with algae that flourished in water exposed to radiation fields that would kill a human in minutes, making it impossible to see objects hidden deep beneath the water. Even television cameras lowered adjacent to the work could not see what was going on, and radiation fogged the camera lenses.

Even so, the work showed that a cleanup was possible. In contrast after the Chernobyl accident, the Soviet authorities decided simply to entomb the reactor in concrete, much of it hastily dropped by helicopter. The Chernobyl accident was much worse than the one at Three Mile Island, which began on March 28, 1979. At Chernobyl an explosion destroyed reactor building, while the heavy metal structures at Three Mile Island remained intact, and the containment dome was not breached.

But the extent of destruction in the core at Three Mile Island was a surprise, gradually revealed since 1985, when the vessel head was removed. Then workers discovered that where once there had been a precise arrangement of 12-foot-long fuel rodsó made up of ceramic pellets of uranium painstakingly sealed into long metal tubes and gathered in precisely configured bundles -- now there was a void.

Lower down was a solidified mass of what had once been molten fuel, and a jumble of metal parts. At the very bottom was more re-solidified fuel, and metal pipes that once guided instrument lines that had been melted like candles left on a radiator.

Condition of Reactor Vessel

The reactor vessel itself, on which metallurgical analysis began only a few weeks ago, shows little sign of damage, according to Government experts. This is of particular importance because a reactor has many backup systems, but is not designed to cope with the failure of the vessel.

The "worst case" scenario for a reactor accident is that once the fuel melts and operators lose the ability to control the geometry of the core, a nuclear reaction re-establishes itself and the core burns through the vessel and then the reactor basement, and from there into the earth beneath, and thence, in the engineers' just, all the way to' China. In practice, if the core reached the ground water, result would be a steam explosion that would spew vast quantities of radiation, experts say.

Although the finding at Three Mile Island suggest that the chance of a catastrophic accident is smaller that critics had predicted, many people in the nuclear industry say they believe that another accident on the same scale in this country would be the last because public pressure would then force the closing of all remaining plants. The cleanup demonstrated other reasons than public health to avoid a nuclear accident, among them, the technical difficulty and cost of the cleanup job. Removing the fuel required the labor of 400 workers over four and a half years, and consumed most of $970 million spent on the partial cleanup. Decontaminating surfaces around the plant was another expense. Another cost was the loss of the generating station, which was built for $700 million and would cost far more to replace.

It was an exceptionally cumbersome task that resembled taking apart a shipwreck in a bottle, and its completion leaves several questions open, including how eventually to dispose of the debris sent to Idaho and the reactor, still contaminated with radioactive residue.

Disposal of Radioactive Water

Also ahead is the evaporation of 2.3 million gallons of slightly radioactive water left over from the accident and the cleanup, a project that has been plagued by legal challenges and mechanical breakdowns.

"In a way, this marks the beginning," said one critic, Eric J. Epstein a spokesman for Three Mile Island Alert, a Harrisburg citizens' group. Noting that the plan is to leave the damaged reactor sealed until the adjacent Unit I reactor is ready for decommissioning, he said, "that plant

Engineers now estimate that molten fuel reached 5,000 degrees.

will be a de facto low-level and high-level repository in the middle of the Susquehanna River for an indefinite period of time."

Plant engineers say that taking apart the plant would be difficult with an operating reactor nearby, and that over time the level of radiation in the plant will decline naturally, so there is no reason to rush.

The owner of the plant, General Public Utilities, says that removing 99 percent of the core and partly decontaminating the containment building leaves the plant in a condition in which it "will not be a hazard to the health or safety of the public he workers or the environment."

Now, in what amounts to forensic engineering, mettallurgists are on to a new chapter of analysis of what actually happened in the accident. "This was a poorly instrumented experiment," said G. A. Kuehn, the site operations director.

With microscopic analysis of metal samples, researchers can now determine what the peak temperature was, and hope to learn more about how 20 tons of molten uranium and other core parts managed to flow into the bottom of the vessel in about a - minute, and how the five-inch-thick carbon steel vessel survived with relatively little damage.

Confusion at Early Stage

The early analysis was been confusing, when the first sample of the vessel bottom was pulled to the surface, General Public Utilities said it looked as if the carbon steel had cracked. But Government scientists now say that only a thin stainless steel liner was damaged, with a tear that they attribute to repeated heating and cooling of the reactor. They say the tear occurred because carbon steel and stainless steel expand at different rates when heated.

Fourteen other pieces of steel will be analyzed over the next year and a half. Researchers are interested in the variations of temperature.

"You can say what the average is, but if you have a spot that is extremely hot, you may have a localized spot where it can fail," he said. "It's the weakest link In the chain. We want to know that in the worst spot, we had adequate margin."

The consensus of experts is that the heavy damage to the core is a surprising contrast to the relatively light amount of damage to the vessel. "It was a hell of a lot more serious than anybody thought in the first days following the accident," said Troy Wade, who was the manager of the Idaho National Engineering Laboratory when the core samples began to arrive there. But because the vessel and the containment functioned well, he said, "nobody died, and nothing melted through to China."

Engineers say that the melting temperature of the vessel bottom 2,760 degrees Fahrenheit, at the welds where instrumentation lines enter the vessel, it is 2,540 degrees. The normal operating temperature is 580 degrees.

Some of the molten mass of fuel, forming a new substance that engineers here call corium, reached 5,000 degrees, far above those melting points, say the engineers, but that was in the region at the center of the vessel, not at the walls or the bottom. "As the corium cascades, it gives up heat, and it has no new heat source," said Mr. Kuehn "And it would take a hell of a sustained temperature to get through the vessel."

Melting of Core

Inside the vessel, workers found that 52 percent of the core had melted, leaving tens of tons of loose debris and a 15,000-pound boulder that had melted and re-solidified, with steel and zirconium parts now functioning as reinforcing bars.

The cleanup task required ingenuity to perform jobs never tried before. Three machine shops ran almost continuously to make tools for new tasks. To break up larger pieces, workers used a tool described as a "nuclear grade Roto Rooter." To break up the boulder they drilled repeated holes, in a process they called "swiss cheesing."

At another stage, engineers wanted to remove bolts that were intended never to be taken apart, and had to develop tools to break steel bars that had been welded in place over the lugs. And steel parts never intended to be removed had to be taken apart with hundreds of cuts by a plasma torch burning at 5,000 degrees.

To remove the debris, workers used clamshell-type tool.; of the type that hang on the end of steam shovels. But metal bars that were formerly part of the fuel or the supporting structures kept holding the jaws open and letting everything fall through. And everything had to go into containers 14 inches across. The containers were kept small to limit the possibility of assembling a critical mass of nuclear material that would sustain a chain reaction.

And every time a tool was lifted out of the water, workers had to vacate the platform, to avoid radiation exposure. But because they succeeded in decontaminating much of the area where they had to work, total radiation exposure will be held to less than half the minimum predicted by the Nuclear Regulatory Commission before work started.

Contamination of Worker

Precautions against overexposure worked well until January, when a worker in a cleanup area picked up what he thought was a bolt and turned out to be a piece of fuel. Engineers are not certain about the extent of the dose to his hand and arm, but believe that it was substantially higher than the quarterly limit set by the commission.

Water was kept above the debris, as a shield to hold down radiation. But that meant that every pound of debris had to be painstakingly hauled up out of the water, at depths of up to 40 feet using long-handled tools. The workers stood on a temporary bridge above the vessel.