The Cesium Cloud That Crossed Europe Before Anyone Named Its Source

On June 9, 1998, the Swiss government announced that its national monitoring network had picked up airborne cesium-137 at levels up to 1,000 times background. It could not say where the cesium came from. France and Italy, it added, were seeing the same thing.

The answer was a steel mill on Spain’s southern coast. In the early hours of May 30, a furnace at the Acerinox plant near Algeciras had melted a load of scrap metal with a cesium-137 source hidden inside it, and part of the cesium went up the stack. By the time Switzerland raised the alarm, the plume had spent ten days drifting across the western Mediterranean and into central Europe. That sequence comes from a 1999 account by the Atmospheric Release Advisory Capability (ARAC) team at Lawrence Livermore National Laboratory, which later reconstructed the release.

The Algeciras melt matters because of who caught it. The mill’s own radiation monitors, installed to stop exactly this kind of source from entering the furnace, did not detect it, according to 2007 Senate testimony by Joel O. Lubenau, a radiation safety specialist. Spain’s own environmental network recorded no rise. The alarm came from air samplers in other countries, run for their own reasons, and it set off a search across a continent for a source no one had reported.

A cloud with no owner

The levels were small in absolute terms. In France and Italy, airborne cesium-137 reached up to 2 mBq per cubic meter, about 1,000 times the natural background for that nuclide, according to a session report from the International Radiation Protection Association’s 2000 congress. That was enough to trigger an alert through ECURIE, the European Union’s system for urgent radiological information exchange.

It was also enough to worry people about the wrong thing. The discovery initially raised concern that an unreported nuclear power plant accident had occurred somewhere, Lubenau said in his testimony.

Within a day of the Swiss announcement, Acerinox told Spain’s nuclear regulator, the Consejo de Seguridad Nuclear (CSN), that it had found radiation in one of its furnace filtration systems. A written question to the European Commission from Undine-Uta Bloch von Blottnitz, a member of the European Parliament, dates that notification to June 9 and says contamination had been found in the plant’s smoke filters on June 2. On June 12, the Spanish government said that a medical radiotherapy source of cesium-137 had “apparently” been melted in the mill’s scrap furnace. The International Atomic Energy Agency’s Emergency Response Centre issued a bulletin the same day.

What cesium-137 emits

Cesium-137 decays by emitting beta particles, and it does so with a half-life of about 30 years, according to the U.S. Environmental Protection Agency’s cesium-137 fact sheet. Its decay product, barium-137m, survives only about 2.6 minutes before releasing a gamma ray and settling into stable barium.

That short-lived daughter is what makes the nuclide easy to count. The gamma from barium-137m is the signal laboratories measure when they put an air filter on a gamma spectrometer. For people downwind, the exposure route that mattered was internal: cesium-137 enters the body when it is inhaled or ingested, the EPA says, and then spreads through soft tissue.

Running the plume backward

ARAC learned of the accident from U.S. news reports on June 12. Neither the exact time of the release nor its size was known. So the Livermore team worked backward from roughly 100 air measurements, adjusting the size and timing of a modeled release until its predicted concentrations matched what stations had recorded. The measurements came from Spain, France, Switzerland, Italy, Sweden, Russia and the European Union, according to a companion Livermore paper.

The model put the cloud first over the Mediterranean. For seven days its axis stayed over open water, passing the Balearic Islands, before it turned north into central Europe, where the Alps split it in two. That path fits the other half of the puzzle: the CSN’s own monitoring network on land had not recorded any rise.

By June 26, ARAC had settled on a release of 50 curies (about 1.85 TBq) over 30 minutes, starting at 1:30 a.m. UTC on May 30. On June 30, the CSN supplied the plant’s location and stack details and said the release had probably happened between 1 a.m. and 3 a.m. UTC that day. The Livermore team wrote that its estimate “compared very well” with the Spanish government’s range of 8 to 80 curies (about 0.3 to 3 TBq).

Not every country downwind of the Strait of Gibraltar saw the plume. An Algerian survey published in Mediterranean Marine Science sampled air, soil, sediment, vegetation and seawater in Algiers and western Algeria from mid-June to July 1998 and found no evidence of cesium-137 contamination from the incident.

The cost of a source no one saw

The health consequences “fortunately remained trivial,” according to the IRPA session report. The financial ones did not. Clean-up, the factory shutdown and waste storage cost more than $25 million, the same report said.

The European Parliament question also said the furnace had kept running for several days after the melt, producing contaminated steel, ash and other residues, and raised the possibility that ash had been dumped at a local site.

Spain’s response was aimed at the front gate rather than the stack. The accident led to the “Spanish Protocol,” a voluntary agreement between industry and government for radiological monitoring of scrap metal, according to the IAEA.

The cesium that escaped Algeciras is still decaying. At a 30-year half-life, a little more than half of what went up the stack in May 1998 remains. The release itself was noticed not by the mill’s detectors, which were built to catch the source, but by air filters in Switzerland, France and Italy that had been put there for something else.