The probable culprit behind a mysterious cloud of radioactive particles detected floating above much of Europe in 2017 appears to have been identified.

The radiation spike – in the form of an extremely high airborne concentration of the radioactive isotope ruthenium–106 – was detected by scientists in October 2017, but the source of the dramatic radiation surge (almost 1,000 times normal levels) was never definitively confirmed.

At the time, many speculated that nuclear facilities in Russia were responsible for what was perceived as an accidental ruthenium–106 release – despite denials at the time by Russian authorities.

Now new research looks to back up the Russian origin hypothesis, according to an international team of almost 70 scientists led by radionuclides researcher Olivier Masson from the Institut de radioprotection et de sûreté nucléaire (IRSN) in France.

"Based on airborne concentration spreading and chemical considerations, it is possible to assume that the release occurred in the Southern Urals region (Russian Federation)," the researchers explain in their new paper.

In what they claim is the most comprehensive assessment of the incident to date, Masson and his team analysed over 1,300 readings taken of the radioactive cloud, recorded by 176 measuring stations in almost 30 countries.

While the airborne radioactive matter released was not harmful to human health, it nonetheless constituted the most serious release of radioactive material since the Fukushima accident in 2011 – with maximum values of 176 millibecquerels of the isotope per cubic metre of air.

Shortly after the release, Russian officials suggested the radiation surge might have been due to a crashing satellite, with the isotope being released from the battery of a spacecraft re-entering Earth's atmosphere.

Not so, according to the new analysis.

"The measurements indicate the largest singular release of radioactivity from a civilian reprocessing plant," says one of the researchers, radioecologist Georg Steinhauser from the University of Hanover.

Specifically, the new evidence – based on modelling of air mass movements around the time of the accident – indicates Russia's Mayak nuclear complex in the southern Urals "should be considered as a likely candidate for the release", the researchers conclude.

That lines up exactly with early suspicions dating back to November 2017, although the Russian state nuclear corporation Rosatom has since insisted that normal readings in the soil around the facility show the Mayak plant can't have been responsible, as if the ruthenium–106 release had originated there, the concentration of the isotope would have been thousands of times higher.

The new findings cast more doubt on the veracity of those claims, however, with the team speculating the accident might have occurred at Mayak while scientists were trying to produce the isotope cerium–144, for subsequent use in neutrino experiments at the Gran Sasso National Laboratory in Italy, as has previously been speculated.

"We were able to show that the accident occurred in the reprocessing of spent fuel elements, at a very advanced stage, shortly before the end of the process chain," says Steinhauser.

"Even though there is currently no official statement, we have a very good idea of what might have happened."

If the researchers' modelling is correct, the accident occurred sometime in late September 2017, on either the 25th or 26th of the month – almost exactly 60 years to the day after one of the worst nuclear accidents in history at the same site: the Kyshtym disaster, ranked as the third most serious nuclear accident ever on the International Nuclear Event Scale.

While nothing in the new research can definitively prove an accident at the Mayak site was the origin of the radioactive plume, other evidence points the same way.

The neutrino research to be conducted in Italy – called SOX (Short distance Oscillations with boreXino) – was cancelled only a number of months after the radioactive cloud was released in late 2017, after the required cerium–144 necessary for the experiments failed to be obtained.

Communications in relation to the project cancellation seen by Science News stated: "we learned that during the purification of the material unexpected problems occurred, yielding to a loss of activity as well as an increase of impurity levels".

For neutrino physicist Jonathan Link of Virginia Tech, who was part of the SOX research team awaiting the isotope, news reports of the radioactive cloud floating over Europe carried a deeper significance that at the time was understood by very few.

As Link told Science News, "I already suspected … that I knew what happened".

The findings are reported in PNAS.

Update 31 July 2019:

A representative from Rosatom told ScienceAlert the agency disputed the new findings and denied the alleged accidental leak at the Mayak facility had taken place, saying the accidental release scenario was "inconsistent with the established facts".

"We maintain that there have been no reportable events at any Rosatom-operated plants or facilities," Rosatom said in a statement.

"Both the national regulator and experts from an independent international inquiry inspected the Mayak facility back in 2017 and found nothing to suggest that the ruthenium-106 isotope originated from this site, nor found any traces of an alleged accident, nor found any evidence of local staff exposure to elevated levels of radioactivity."