NUCLEAR ACCIDENT IN JAPAN: THE SCIENCE; A Flash, and an Uncontrolled Chain Reaction

By JOHN NOBLE WILFORD and MATTHEW L. WALD © New York Times, 1 October 1999.

Workers at a Japanese nuclear fuel factory were mixing a liquid batch of uranium yesterday when suddenly there was a flash of blue light, the ominous flash of radiation at the moment when the uranium begins a chain reaction or, in the language of the age, goes critical.

''Inadvertent criticality'' is the term that nuclear engineers use for this accidental chain reaction.

Since the dawn of the atomic age, scientists and engineers have worried about this type of accident in the handling of nuclear materials like uranium and plutonium.

Accidents at power plants are the most familiar, but accidents like yesterday's, which occurred at a processing plant, are not unknown.

As the atoms are split, their fragments, called fission products, are quickly spread and are far more radioactive than the uranium was in the first place.

But nuclear regulatory officials and scientists said that though serious, the accident, at the Tokaimura plant northwest of Tokyo, was not a catastrophe on the scale of a large reactor accident like the partial nuclear meltdown at Three Mile Island, near Harrisburg, Pa., in 1979, or the explosion at the Chernobyl nuclear reactor in Ukraine in 1986. The Ukraine accident killed 31 people and contaminated vast areas of the former Soviet Union.

''I don't think that it will be a long-term problem,'' David Lochbaum, nuclear safety engineer with the Union of Concerned Scientists, told Reuters. ''This appears to be a local disaster and not a global one.''

For one thing, scientists said, the Japanese accident involves only a few pounds of uranium, which should not be enough fuel to sustain a prolonged chain reaction or cause widespread destruction. But the criticality does give off a shower of radiation.

In a reactor, the reaction is sustained because the nuclear material is held in place, at the proper density and shape for continuing the fission process. Outside a reactor, though, the reaction should die out quickly because the energy released by the fission literally blows the mass apart in a small radioactive burst.

It was still too early to determine exactly what happened at the Japanese plant.

Hans Meyer, a spokesman for the International Atomic Energy Agency in Vienna, said that the uranium involved was from ore and was being cleaned in a chemical process.

Some American experts said that if, as reported, the tank was filled with uranium in the form of uranyl nitrate, this implied that the uranium had come not from ore but from uranium being reprocessed after a prior use.

According to several reports, the uranium had been enriched to 19 percent, meaning that 19 percent of the atoms were uranium 235, the form that splits readily.

A breeder reactor, a type intended to produce more fuel than it consumes, needs the richer form of uranium. But in most commercial reactors, uranium enrichment levels are 3 percent to 5 percent. The higher the enrichment, the less uranium is needed to form a critical mass, the minimum amount needed for a chain reaction.

Dr. Peter Parker, a professor at the Nuclear Structure Laboratory at Yale University, said the workers had apparently been mixing too much uranium in a small vessel. ''That made the density of the uranium so high that the neutrons didn't get absorbed by the other material around,'' he said, and this led to an out-of-control reaction.

Several criticality accidents have occurred at fuel processing plants in the United States. One worker received a lethal dose of radiation in a similar accident at a plant in Charlestown, R.I., in 1964. Similar accidents at Los Alamos, N.M., killed two bomb researchers during the Manhattan Project, the effort in World War II to develop the first atomic bombs.

In 1991, at a Wilmington, N.C., plant owned by General Electric, workers mistakenly transferred liquid wastes containing uranium into a tank that, because of its size and shape, could have posed a criticality hazard. The plant declared an emergency, but the crisis passed without an out-of-control reaction.

The Department of Energy has repeatedly fined contractors for lax controls over uranium and plutonium handled at its weapons plants. The Nuclear Regulatory Commission, which regulates civilian atomic projects, licenses seven processing plants and inspects them twice a year for criticality safety. There are two installations in Lynchburg, Va., and one each in Hematite, Mo.; Columbia, S.C.; Erwin, Tenn.; Richland, Wash., and Wilmington, N.C.
Victor Dricks, a spokesman for the Nuclear Regulatory Commission, said that to obtain a license, plants must ''have at least two independent barriers to inadvertent criticality.''

They could simply assure that the vessel is too small to contain a critical mass, or that it has a safe shape -- a column, for example, rather than a round bowl. In a sphere, the neutron given off by a uranium atom is far more likely to strike another uranium atom, sustaining a chain reaction, than if the material is in a long, thin pipe.

''You control the geometry,'' said Felix Killar, director of materials licensing at the Nuclear Energy Institute, a trade group in Washington. ''You put it in containers that no matter what happens, the material can't go critical, because it's in a safety geometry.''

Some plants weigh the contents of a tank and shut off the incoming flow before it is deep enough to assume a round shape, industry experts said. They also keep radioactive liquids in a dilute form.

Dr. Arjun Makhijani, president of the Institute for Energy and Environmental Research in Washington, said that the Tokaimura plant had had other accidents in recent years and should be shut down for a thorough, independent investigation.

Responding to reports that the reaction there could be continuing hours after the accident, he said: ''This is a most, most unusual accident. This is going to fascinate the people with the sharp pencils for a long time.''