On 26 April 1986, one of the four reactors at Chernobyl power station exploded after an early morning safety test went wrong, causing arguably the most catastrophic nuclear accident in history.

Now, a new hypothesis has been proposed that challenges the accepted version of events in the initial moments of the disaster: namely, that an initial steam explosion was followed by a nuclear blast a few seconds later.

Research published in the journal Nuclear Technology suggests that the first explosion was actually a nuclear blast, turning the original theory on its head.

A series of nuclear explosions within the faulty reactor caused a jet of radioactive debris to be spewed out high into the sky, according to the new study – which involved researchers from the Swedish Defence Research Agency, the Swedish Meteorological and Hydrological Institute and Stockholm University. This was followed by a steam explosion within three seconds that ruptured the reactors, sending further debris into the atmosphere, albeit at lower altitudes.

The new version of events is based on analysis of findings made by Soviet scientists from the Khlopin Radium Institute in Leningrad – now Saint Petersburg – four days after the incident. These scientists detected traces of xenon – a chemical element – at Cherepovets, a city more than 200 miles north of Moscow, which was far from the trajectory of the majority of the Chernobyl debris. The main radiation cloud travelled northwest over Central Europe towards Scandinavia.

Analysis showed that the xenon traces were the product of recent nuclear fission, suggesting they came from a nuclear explosion. After examining weather conditions at the time, the researchers say this xenon could have come from the initial nuclear blast that shot radiation 2 miles into the atmosphere. This was then guided north east by the conditions at high altitude, as opposed to the debris that came from the subsequent steam explosion, which was pushed in a north-westerly direction by a different weather pattern.

Eyewitness accounts of an initial "blue flash" over the power station may also back up the claim that the nuclear explosion came first. Furthermore, damage to the reactor's core and seismic measurements are also consistent with this new timeline, lead author Lars-Erik De Geer says.