The Legacy of Chernobyl
by Zhores A. Medvedev
Norton, 352 pp., $24.95
The worst nuclear accident ever to take place began with a safety test. At 1:23 AM on Saturday, April 26, 1986, the operator of Reactor No. 4 at the Chernobyl nuclear power station started an experiment to see how long a spinning turbine could provide electricity in the event of a loss of power to the plant. If the power supply failed, it would take more than thirty seconds for the backup generators to come into play. The purpose of the test was to see whether the turbine could provide enough power to pump cooling water over the uranium fuel until the emergency generators took over.
The test was to be conducted just before the reactor was shut down for routine maintenance. At midday on April 25 reactor power was reduced to about 50 percent. The next step was to reduce power to about 30 percent, but this was delayed until 11:10 that night because of unexpectedly high demand for electricity in the Kiev region. As a result, the test was conducted by the night shift which, unlike the day shift, had not been instructed in advance about it and was, besides, less experienced.
The test called for operating the reactor at 30 percent of power, so that if the test failed the first time it could be repeated. When permission was given to reduce power at 11:10 PM, the operator made a mistake in setting the controls, and power fell to 1 percent, too low for the experiment. This caused a buildup of xenon in the uranium fuel and threatened to shut down the reactor completely.
In order to prevent this, the operator pulled out almost all the control rods, bringing the reactor up to about 7 percent of power. This was very dangerous, because some of the control rods are used for emergency shutdown. The reactor was now unstable: the xenon buildup was acting as a brake, pushing the reactor toward shutdown, while the operator was trying to drive up the power by removing the control rods. An automatic system should have shut down the reactor at this point, but the operator wanted to ensure that the test could be carried out, and he therefore blocked a number of the emergency shutdown signals.
At 1:23:04 AM the test began. The turbine was disconnected and its energy was fed to four of the eight main pumps. As the turbine slowed down, so too did the pumps and the flow of cooling water over the fuel elements. Power began to rise, and at 1:23:40 the operator pressed the button for an immediate shutdown of the reactor. The shutdown rods moved down too slowly to prevent a runaway increase in power (and they may even have contributed to it, because of faulty design). In any event, power rose to about one hundred times its normal full level within four seconds. The fuel disintegrated and caused a rapid boiling of the cooling water.