On 25 March, Tokyo Electric Power Company (TEPCO), which runs the Fukushima Daiichi plant, reported that chlorine-38 had been found among several radioactive isotopes in water taken from the turbine building of reactor number 1. Chlorine-38 can form when neutrons hit atoms of naturally occurring chlorine-37 in seawater.
Some chlorine-38 is to be expected because seawater was used to cool reactor number 1, and fuel rods continue to emit neutrons even after a reactor is shut down. But according to calculations by Dalnoki-Veress, a dead reactor core would not produce enough neutrons to account for the amount of chlorine-38 measured – 1.6 million becquerels per millilitre.
Despite the presence of the control rods, he says, partial or total melting of the fuel rods could rearrange the fuel in ways that would allow chain reactions to occur. Such reactions would probably be short-lived, he says, as the heat they would generate would tend to push apart the chunks or molten globs of fuel involved, reducing their ability to react further. But continued melting and rearrangement could cause repeated bursts of chain reactions.
John Lee of the University of Michigan in Ann Arbor says that achieving criticality in a reactor – initiating a chain reaction, in other words – requires a precise arrangement of fuel.
Even if the fuel rods partially or fully melted, it is hard to see how they could achieve such an arrangement by accident. He recently ran a simulation showing that a molten sphere of uranium fuel would be further from criticality, not closer, than an intact reactor with the control rods inserted.
"I cannot say it's absolutely impossible," he says. "[But] for me, it's difficult to imagine that a reactor after shutdown somehow had achieved a critical configuration again."
That still leaves the puzzling neutron and chlorine-238 measurements, but there are other possible explanations.
Lee says that gamma rays from the radioactive material that has leaked outside the reactors could be confusing the measurements. Neutron detectors are supposed to distinguish between the two kinds of radiation, but do a poorer job of this when the flux of gamma rays is very strong, he says.
A similar discrimination problem could explain the chlorine-38 measurement, says Jim Rushton of Oak Ridge National Laboratory in Tennessee. Other radioactive isotopes give off gamma rays of about the same energy as chlorine-38 does, so they could be confusing the measurements, he says.
He agrees with Lee that spontaneous chain reactions are very unlikely. Even if they did occur briefly, they would not add much radioactivity or heat beyond what workers are already dealing with from the radioactive material that accumulated when the reactors were running at full power, he says. "You'd get a spike, but for seconds or fractions of a second," he says.
If you liked this article, please give it a quick review on ycombinator or StumbleUpon. Thanks