Classification: Documented

Chernobyl: the night of Reactor Four and the decades that followed

In the early hours of 26 April 1986, a test at Chernobyl’s Reactor Four ended in explosions and a release of radioactive material. This dossier follows the technical sequence, delayed public warning, emergency work, measured health effects and uncertain projections, through to the continuing challenges of confinement.

Pripyat, Oblast de Kyiv, Ucrânia · See stories from this place and nearby

A city still asleep

In the early hours of 26 April 1986, the Chernobyl plant stood near Pripyat, then in the Ukrainian Soviet Socialist Republic. The city of workers and families was only a few kilometres from Reactor Four. When explosions destroyed the core, people outside the plant could not see radiation. The International Atomic Energy Agency’s (IAEA) International Chernobyl Project technical report says Pripyat residents received no public protective instructions during 26 April. Children remained outdoors as officials prepared an evacuation that would not begin until the next day.

Present-day photographs of empty apartment buildings often suggest that Pripyat instantly became a ghost city. The chronology is different: there was a living community, a night of disaster, a day of scarce information and then buses. The atlas coordinate identifies the plant approximately, not a home address or a place to visit. This dossier follows what the records can establish and what still depends on models.

A test pushed into the night

The crew intended to test how long the turbine, still spinning after shutdown, could power circulation pumps until backup systems took over. They were not testing a nuclear explosion. According to the IAEA’s INSAG-7 report, power reduction began on 25 April, but Kyiv’s electricity grid asked for the test programme to be delayed. The operation continued into the following night. At 00:28, a change in automatic control caused power to fall far below the planned level. Attempts to raise it then took place under difficult conditions for stability.

Xenon-135, a fission product that absorbs neutrons, made recovering power after the fall more difficult. Withdrawing rods to counter that absorption reduced the operating safety margin. INSAG-7 describes operating actions that created an unsafe state, while stressing design faults and the information withheld from operators. The phrase “someone pressed the wrong button” erases a chain of technical and organisational decisions.

The problem concealed in the shutdown rods

Reactor Four was an RBMK-1000, graphite moderated and water cooled. Under certain conditions, more steam bubbles in the cooling water meant more reactivity: a chain reaction tending to grow instead of subside, known as a positive void coefficient. In that night’s unstable configuration it favoured a rapid power surge. INSAG-7 describes a less intuitive fault as well: at the start of insertion, shutdown rods could temporarily increase reactivity in part of the core before reducing it.

The AZ-5 shutdown command was activated shortly before destruction. The report says the rods’ initial effect, combined with the steam feedback, may have been decisive in the severity of the excursion. It does not pretend every second was directly observed. According to the IAEA review, the rod defect had been identified at the RBMK plant at Ignalina in 1983, yet promised changes had not been implemented before Chernobyl. A device intended for protection could worsen that exceptional combination of conditions.

01:23: explosions, fire and release

Around 01:23 on 26 April, power surged and explosions destroyed Reactor Four. Fires and the open core released radionuclides into the atmosphere over days. This was not an atomic-bomb detonation: reactor dynamics, steam, mechanical damage and fire produced a radiological release on an enormous scale. Plant staff and firefighters faced flames and wreckage without a full grasp of the invisible danger in the first moments.

The release included iodine-131, important for thyroid dose in the first weeks, and caesium-137, significant for long-lived environmental contamination. UNSCEAR records detectable material across the northern hemisphere, but detection does not mean identical doses or risks everywhere. The plume depended on weather, rainfall and wind direction. Deposition maps, food testing and exposure histories tell more than a single shaded area drawn over Europe.

The delayed warning and Pripyat’s buses

The IAEA technical report says the decision to evacuate Pripyat was confirmed on the morning of 27 April, with departure planned for about 2 p.m. More than a thousand buses were mobilised. Nearly 50,000 residents left the city. UNSCEAR estimates about 115,000 evacuated in 1986 from nearby areas and around 220,000 relocated after 1986 across Ukraine, Belarus and Russia. These figures describe different phases; they are not a single journey out of Pripyat.

The delay in public information crossed borders. Radioactive material was detected in Scandinavia before the Soviet Union publicly acknowledged the severity of the event. It would be simplistic to describe an absolute absence of response: technical records show internal measurements and preparations while the population still lacked clear instructions. Communication failure was part of the displaced people’s experience, making it harder to understand risks or plan a return, work and family life.

Those who received the largest doses

Acute harm was concentrated among people near the reactor in the first hours. UNSCEAR records 134 plant staff and emergency workers with acute radiation syndrome; 28 died of it in 1986. IAEA sources also record two workers who died in the immediate aftermath, a different category from the 28 syndrome deaths. That is why UNSCEAR’s short summary mentions 30 workers dying within weeks. None of these figures counts all potential future deaths associated with the disaster.

Hundreds of thousands undertook containment, clean-up and reconstruction tasks and became known as liquidators. Their doses and duties varied greatly. Studies indicate increased leukaemia and cataracts among more highly exposed groups, but cannot attribute every later illness of each participant to radiation. Their history also includes hazardous work, medical follow-up and social losses that do not fit a table of acute mortality.

Iodine’s route to the thyroid

One of the best-documented health effects came not from a glowing reactor, but from an ordinary food: milk contaminated with iodine-131. Animals grazed in affected areas; children who drank fresh milk received relatively high thyroid doses. The World Health Organization (WHO) and UNSCEAR link childhood exposure to a sharp increase in thyroid cancer in Belarus, Ukraine and affected Russian regions. Timely action on milk and stable iodine could reduce doses under appropriate conditions; the response varied by place and time.

The registries require careful reading. By 2005, UNSCEAR counted more than 6,000 cases among people who were children or adolescents in 1986. Its 2017 analysis collected 19,233 diagnoses from 1991 to 2015 in the defined cohort and regions. Those 19,233 diagnoses are not 19,233 individually radiation-caused cases. Incidence also changed as the cohort aged and with greater screening and better diagnostics. For non-evacuees, UNSCEAR estimated a radiation-attributable fraction of around 25%, with wide uncertainty; that percentage cannot simply be applied to every case in the total.

Why there is no single death toll

A death from acute radiation syndrome can be diagnosed clinically. A cancer decades later has no label revealing its individual cause. Researchers compare groups, doses and expected background rates to estimate statistical excesses. The Chernobyl Forum, publicised by WHO in 2005, projected up to about 4,000 eventual radiation-related deaths among the higher-exposure groups. That was a projection for a defined population, not deaths already observed in 2005 and not a total for all Europe. WHO’s cancer research agency used another population and time horizon for different projections.

UNSCEAR chose not to turn low doses spread across very large populations into a single absolute projected count because of uncertainty. This does not mean “zero effect”, just as a projection does not prove each predicted death occurred. Studies continue following workers and people exposed as children. This dossier distinguishes what was observed, what was epidemiologically attributed, and what was modelled.

Uneven land and displaced lives

Contamination did not form a uniform circle around the plant. The IAEA describes acute effects in the most highly exposed areas, recovery of some plants and animals, and persistent caesium-137 in certain soils and food chains. In forests, mushrooms, berries and game can retain radiocaesium longer than many cultivated crops. Animals returning to the exclusion zone do not prove radiation harmless: the retreat of farming, industry and human presence also transformed the ecosystem. A place looking green is not necessarily risk free.

The UN Development Programme (UNDP) recovery report treats displacement, lost income, stigma and perceived health as part of the disaster. Thousands left homes and neighbour networks; others continued in areas with restrictions. Reducing every resident to an “irradiated victim” overlooks agency, unequal doses and economic needs. The lasting question is how to provide health monitoring and trustworthy information without defining a whole population as permanently incapacitated.

From sarcophagus to long-term responsibility

An improvised shelter was built over the ruined reactor in 1986. Decades later the New Safe Confinement, a vast arch assembled away from the more contaminated area to reduce workers’ exposure, was moved over Unit Four in 2016 and commissioned in 2019, according to the European Bank for Reconstruction and Development (EBRD). It contains material and enables eventual dismantling of the old shelter; it has not removed all fuel or ended monitoring. In February 2025, a drone strike damaged the new arch. The EBRD reported acceptable radiation readings afterward but also significant damage requiring repair.

Documented: the accident, evacuation, acute radiation syndrome among workers, the rise in thyroid cancer linked to childhood exposure, and the long containment project. Estimated with uncertainty: the final number of additional cancers and deaths across broad populations. Ongoing: epidemiological research, radioactive material management and repair of the damaged structure. To pursue the documentary trail, compare INSAG-7 with UNSCEAR health reports, inspect dated images in the IAEA archive and follow the EBRD timeline. Chernobyl was one night of technical and human failures, but also a test of how societies manage consequences across generations.

Open in Google Maps