The accident at Chernobyl Unit 4, on 26 April 1986, did not occur during normal operation of the reactor. It happened during a test designed to assess the reactor's safety margin in a particular set of circumstances. The test, which had to be performed at less than full reactor power, was scheduled to coincide with a routine shut-down of the reactor.

The test

Nuclear power stations not only produce electricity, they also consume electricity, for example to power the pumps that circulate the coolant. This electricity is usually supplied from the grid. If the source of electricity should fail, most reactors are able to derive the required electricity from their own production. However, if the reactor is operating but not producing power, for example when in the process of shutting down, some other source of supply is required. Generators are generally used to supply the required power, but there is a time delay while they are started.

The test carried out at Chernobyl-4 was designed to demonstrate that a coasting turbine would provide sufficient power to pump coolant through the reactor core while waiting for electricity from the diesel generators. The circulation of coolant was expected to be sufficient to give the reactor an adequate safety margin.

Main factors in the accident

Simplified sequence of events

A number of reports have been published that have given summaries of the events leading up to the accident. Since the reactor was destroyed, these summaries have been based on interpretation of evidence. They have not been consistent. There are three reasons for this: Different researchers have interpreted the same evidence in different ways. With the passage of time more evidence has become available. Some authors of reports have been biased. The sequence of events which follows has been compiled following a review of a large number of reports and it represents what we consider to be the most likely sequence of events.

April 25: Prelude

01:06	The scheduled shutdown of the reactor started. Gradual lowering of the power level began .
03:47	Lowering of reactor power halted at 1600 MW(t).
14:00	The emergency core cooling system (ECCS) was isolated (part of the test procedure) to prevent it from interrupting the test later.
The fact that the ECCS was isolated did not contribute to the accident; however, had it been available it might have reduced the impact slightly.
14:00	The power was due to be lowered further; however, the controller of the electricity grid in Kiev requested the reactor operator to keep supplying electricity to enable demand to be met. Consequently, the reactor power level was maintained at 1600 MW(t) and the experiment was delayed.
Without this delay, the test would have been conducted during `day shift'.
23:10	Power reduction recommenced.
24:00	Shift change.
April 26: Preparation for the test
00:05	Power level had been decreased to 720 MW(t) and continued to be reduced.
It is now recognised that the safe operating level for a pre-accident configuration RBMK was about 700 Mwt because of the positive void coefficient.
00:28	Power level was now 500 MW(t).
Control was transferred from the local to the automatic regulating system. Either the operator failed to give the `hold power at required level' signal or the regulating system failed to respond to this signal. This led to an unexpected fall in power, which rapidly dropped to 30 MW(t).
00:32	(approximate time). In response, the operator retracted a number of control rods in an attempt to restore the power level.
Station safety procedures required that approval of the chief engineer be obtained to operate the reactor with fewer than the effective equivalent of 26 control rods. It is estimated that there were less than this number remaining in the reactor at this time.
01:00	The reactor power had risen to 200 MW(t).
01:03	An additional pump was switched into the left hand cooling circuit in order to increase the water flow to the core (part of the test procedure).
01:07	An additional pump was switched into the right hand cooling circuit (part of the test procedure).
Operation of additional pumps removed heat from the core more quickly. This reduced the water level in the steam separator.
01:15	Automatic trip systems to the steam separator were deactivated by the operator to permit continued operation of the reactor.
01:18	Operator increased feed water flow in an attempt to address the problems in the cooling system.
01:19	Some manual control rods withdrawn to increase power and raise the temperature and pressure in the steam separator.
Operating policy required that a minimum effective equivalent of 15 manual control rods be inserted in the reactor at all times. At this point it is likely that the number of manual rods was reduced to less than this (probably eight). However, automatic control rods were in place, thereby increasing the total number.
01:21:40	Feed water flow rate reduced to below normal by the operator to stabilise steam separator water level, decreasing heat removal from the core.
01:22:10	Spontaneous generation of steam in the core began.
01:22:45	Indications received by the operator, although abnormal, gave the appearance that the reactor was stable.
The test
01:23:04	Turbine feed valves closed to start turbine coasting. This was the beginning of the actual test.
01:23:10	Automatic control rods withdrawn from the core. An approximately 10 second withdrawal was the normal response to compensate for a decrease in the reactivity following the closing of the turbine feed valves.
Usually this decrease is caused by an increase in pressure in the cooling system and a consequent decrease in the quantity of steam in the core. The expected decrease in steam quantity did not occurdue to reduced feedwater to the core.
01:23:21	Steam generation increased to a point where, owing to the reactor's positive void coefficient, a further increase of steam generation would lead to a rapid increase in power.
01:23:35	Steam in the core begins to increase uncontrollably.
01:23:40	The emergency button (AZ-5) was pressed by the operator. Control rods started to enter the core.
The insertion of the rods from the top concentrated all of the reactivity in the bottom of the core.
01:23:44	Reactor power rose to a peak of about 100 times the design value.
01:23:45	Fuel pellets started to shatter, reacting with the cooling water to produce a pulse of high pressure in the fuel channels.
01:23:49	Fuel channels ruptured.
01:24	Two explosions occurred. One was a steam explosion; the other resulted from the expansion of fuel vapour.
The explosions lifted the pile cap, allowing the entry of air. The air reacted with the graphite moderator blocks to form carbon monoxide. This flammable gas ignited and a reactor fire resulted.
Results
Some 8 of the 140 tonnes of fuel, which contained plutonium and other highly radioactive materials (fission products), were ejected from the reactor along with a portion of the graphite moderator, which was also radioactive. These materials were scattered around the site. In addition, caesium and iodine vapours were released both by the explosion and during the subsequent fire.