A test reaches its final seconds
Published 12 October 2026
On April 25, 1986, staff began reducing the power of Unit 4. The later commission chronology puts the start at 01:06 and records a substantial operating reactivity margin at that point. The unit was being prepared for a turbine rundown test. The intended question was whether a generator, after the steam supply was cut, could provide power during the interval before other supplies were available. The test joined the electrical equipment to the reactor's cooling arrangements because some circulating pumps would be supplied by the slowing turbine. A question about the power available during rundown was therefore not independent of water flow through the core. The equipment test took place inside a functioning reactor system whose state had to remain controllable while the electrical conditions changed. C021, pub.iaea.org, The Chernobyl Accident: Updating of INSAG-1
The April 25 log shows preparations continuing at roughly half thermal power. One turbogenerator was disconnected at 13:05. At 14:00, the emergency core cooling system was disconnected from the circulation circuit, and the Kiev grid controller requested postponement of the test. The reactor remained at reduced power for a prolonged period rather than following the original shutdown schedule without interruption. INSAG-7 later distinguished the authorised disconnection for the test from the extended operation that followed. It accepted that the initial disconnection had been permitted, but regarded keeping a vital safety system unavailable during approximately eleven hours of operation as evidence of weak safety practice. The later criticism is precise: permission for one step did not justify treating a long delay as though it were simply the planned test. C021, pub.iaea.org, The Chernobyl Accident: Updating of INSAG-1
The chronology records power reduction continuing at 23:10. After midnight, the thermal power stood at 720 megawatts, then fell towards the level at which a transfer between automatic control systems occurred. At 00:28, during the move from local to global control at about 500 megawatts, power dropped to roughly thirty megawatts. That fall was outside the test programme. The log says measures were then taken to increase power. Later accounts did not agree completely on the immediate cause of the collapse. INSAG-1 had described operator error, whereas the 1992 reassessment cited newer reports referring to an unknown cause or difficulty controlling the system. The sequence is clearer than the allocation of blame for that particular moment. Power fell sharply, and staff continued towards the test rather than completing shutdown. C021, pub.iaea.org, The Chernobyl Accident: Updating of INSAG-1
Recovering power after the drop required withdrawing rods to compensate for the additional negative reactivity associated with xenon poisoning. The Soviet commission later judged that this brought the reactor into a condition outside the operating regulations and one in which emergency protection could no longer guarantee termination of the reaction. It did not establish that the personnel understood those consequences. At about 01:03, power had been brought to 200 megawatts and stabilised. The seventh main circulating pump was started at 01:03 and the eighth at 01:07, putting all eight into use. The configuration combined low thermal power with high circulation and many rods withdrawn. Each element had a technical consequence, and their combination mattered more than a description of the test as simply being run at low power. C021, pub.iaea.org, The Chernobyl Accident: Updating of INSAG-1
The operating reactivity margin immediately before the test was reconstructed after the accident. A record made at 01:22:30 on magnetic tape was later processed at Smolensk, yielding a margin equivalent to eight manual control rods. The commission says the operating calculation was not performed at that moment at Chernobyl and the control-room staff did not have its results. A later calculation cannot be described as a warning displayed to them in real time. This does not remove the requirement to maintain the margin; it distinguishes the rule from the information actually available during the final preparations. INSAG retained the finding that the minimum margin had been violated, while also explaining that the operators did not adequately understand its safety significance. C021, pub.iaea.org, The Chernobyl Accident: Updating of INSAG-1
The reactor's internal distribution was unfavourable. The commission describes a double-peaked axial field with more neutron flux in the upper region, extensive rod withdrawal and greater xenon poisoning in the middle. Water entering the core was only slightly below the boiling point. Under those conditions, a small power rise could cause a larger change in steam content in the lower core. The commission called this an increased susceptibility to runaway. It linked that state to both low reactor power and the unusually high flow caused by eight pumps. This was not yet a statement that the reactor was visibly racing at the console. It described the unstable potential of a configuration in which a later disturbance could grow faster than the protection system could safely control. C021, pub.iaea.org, The Chernobyl Accident: Updating of INSAG-1
At 01:23:04, the emergency stop valves of turbogenerator No.8 were closed and the rundown began. Four pumps were now depending on the turbine as it slowed. The total coolant flow began falling, while the other pumps' contribution increased slightly. The commission estimates that thirty-five seconds into the transient, total flow had fallen by ten to fifteen percent. That decrease increased steam content, partly offset by rising pressure after the turbine valves closed. These were not unrelated measurements from different days. They were elements of one short test sequence. The hydraulic change brought the positive void relationship into play, while the rod configuration and the shape of power inside the core determined what the shutdown system would do when called upon. C021, pub.iaea.org, The Chernobyl Accident: Updating of INSAG-1
The annexed commission account says it did not find evidence of a hidden runaway before the emergency-stop command. It concluded that neither recorded power nor the other recorded parameters required intervention between the start of the test and the pressing of the button. Its modelling found only a small void-reactivity release during the turbine rundown, one that could have been offset by limited rod insertion. That account differs in emphasis from a simple story in which the operators recognised an escalating emergency and pressed the button only as a last desperate response. The commission instead identified an extended vulnerable state in which a positive-reactivity insertion from any cause could produce a power excursion and the emergency system could fail to perform its protective function. C021, pub.iaea.org, The Chernobyl Accident: Updating of INSAG-1
At 01:23:40, the senior reactor control engineer pressed the manual emergency-stop button, described in the annex as EPS-5. The commission could not establish why it was pressed. That uncertainty should remain in the story rather than being replaced with an invented command, a private thought or a reconstructed conversation presented as fact. The command sent the emergency and manual rods into the core. In the upper layers, absorbing material began reducing neutron flux. Lower down, graphite displacement of water added reactivity. The commission's reconstruction places a rapid distortion of the power profile after the movement began. A shutdown signal and a local increase in power were therefore not mutually exclusive events in this reactor design. C021, pub.iaea.org, The Chernobyl Accident: Updating of INSAG-1
The instrument chronology shows the sequence accelerating after that command. At 01:23:43, protection signals indicated a rapid power rise and recorded power above 530 thermal megawatts. Seconds later, pump flows and pressures changed sharply and measuring-system failures appeared. At 01:23:49, the record includes a reactor-space pressure signal associated with a ruptured fuel channel and a loss of voltage to rod drives. The operating log then records severe shocks and rods stopping before their lower limits. These are the times and signals collected by the later commission. They are not a complete film of the core. The recording systems had limits in speed, measured parameters and synchronisation, so the physical reconstruction combines recorded evidence with calculations. The precision of a printed time must not be mistaken for certainty about every internal process in that second. C021, pub.iaea.org, The Chernobyl Accident: Updating of INSAG-1
The later safety group did not claim to know with certainty what initiated the destructive excursion. It regarded some positive reactivity from growing steam voids as likely and additional reactivity from rod insertion as probably decisive. It also emphasised that the defect in the rods had been known before the accident. These conclusions preserve both the physical explanation and the limits of the reconstruction. They do not reduce the event to a normal explosion in electrical machinery, and they do not make the name of the person pressing the emergency button a sufficient account of the cause. The reactor had been put into a state outside specified safe conditions, but the design and the communication of its dangerous characteristics helped determine what happened when the protection system was activated. C021, pub.iaea.org, The Chernobyl Accident: Updating of INSAG-1
INSAG's reassessment would later change several allegations about the test. Sustained operation below 700 thermal megawatts had not actually been prohibited by the design or operating instructions, although the group believed such a prohibition should have existed. Blocking the turbine trip had been carried out under the procedures, and the later commission did not assign blame for it. In light of positive scram, INSAG no longer accepted the earlier statement that this trip would necessarily have saved the reactor. The minimum reactivity margin, by contrast, had been violated. These distinctions prevent an account of the final hours from turning every questionable act into a breach of a rule that existed at the time. The failure was serious without needing invented prohibitions to make it so. C021, pub.iaea.org, The Chernobyl Accident: Updating of INSAG-1