The 1986
THE town of Pripyat 3km away from the Chernobyl complex grew up to service the power plant and had 49,000 inhabitants prior to the accident. The old town of Cherobyl (or Chornobyl as it is known as in Ukraine) 15km to the southeast, had a population of 12,500.

On 25 April, 1986, reactor four was to be shut down for routine maintenance, so it was decided to take advantage of this to run a test. Ironically, the test was designed to improve safety. The reactor's cooling pumps relied on electrical power, so the operators wanted to see how long the turbines could produce sufficient energy to keep the pumps running in the event of a loss of power. The reactor's emergency cooling system was deliberately disabled, as they didn't want it cutting in when the main pumps slowed. To reduce cooling requirements, the reactor was to be run at low power, despite the fact that these reactors were known to be unstable at low power settings. The test had been attempted on two previous occasions but never completed.

There were two main cooling systems excluding the back-up, each with four main pumps. Four of these pumps were powered by the generator that was to 'fail.' Prior to the experiment, with reactor power reduced and all eight pumps operating, water flow exceeded permitted levels. The amount of water in the steam-raising circuit reduced steam production. Additionally, the extra water was absorbing neutrons and causing power to fall. Power fell to less than 1% of capacity, so the operators manually removed control rods to compensate, switched off automatic regulators and eventually stabilised the reactor at the planned test power level. At one point only six-eight control rods were being used. According to procedure, at least 30 are required to maintain control, and if there are any less the reactor should have been shut down. They allowed the test to continue, despite knowing that about 20 seconds would be required to lower all the rods and shut down the reactor in the event of a power surge.

Then both generators were shut down to start the test. The cooling pumps slowed, reducing water flow in the core and producing more steam. The excess water had up until then been absorbing neutrons, so the formation of steam pockets caused neutron flux to increase (the positive void coefficient). At 01:23 hours on 26 April, reactor power increased exponentially, up to an estimated 100 times nominal. The control rods could not be re-inserted in time; the fuel overheated and some of the rods ruptured.

The resulting explosion, thought to be caused mainly by steam pressure and chemical reaction with the exposed fuel, blew the 1000-tonne lid clear of the core. A second explosion threw out fragments of burning fuel and graphite from the core and allowed air to rush in, causing the graphite moderator to burst into flames. The exact cause of the second explosion remains unknown, but it is thought that hydrogen may have played a part.