In the situation which resulted from independent actions of the personnel, the reactor got (on a heat carrier expense) to such state when even little change of power leads to increase in volume steam content, many times over bigger, than at the rated power. Growth of volume steam content caused emergence of positive reactivity. Fluctuations of power finally could lead to its further growth.
Xenon which is called sometimes "a neutron kernel", has a half-life period about nine hours and constantly is present at an active zone of the reactor. But during the normal operation of the device it partially burns out under the influence of the same neutrons therefore practically the amount of xenon remains at one level.
The stop of the reactor of the 4th power unit was planned in the afternoon on April 25, therefore, other, not night personnel prepared for tests. In the afternoon at station at station there are heads, the main experts, and, so there is an opportunity to exercise more reliable control of the course of experiments. However there was "discrepancy". The dispatcher of "Kiyevenergo" did not allow to stop the reactor in time planned for the CNPP as in an integrated power grid there was no electric power because at other power plant the power unit unexpectedly failed.
23 hours 10 minutes — are got permission to a reactor stop. Further decrease in its power to 1000 — 700 MW began (thermal), as well as it was provided by the program of tests. But the operator did not manage to drive therefore the power of the device reduced almost to zero. In such cases the reactor has to be suppressed. But the personnel was not counted with this requirement. Began rise in power.
As a result of researches it became clear that the extent of zones with the increased concentration of plutonium was insignificant, and chemical forms and the sizes of particles in which it appeared, easily was late respirators.
Before a stop tests of one of turbogenerators in the mode of a vybeg with loading of own needs of the block were planned. The essence of experiment consists in modeling of a situation when the turbogenerator can remain without the driving force, that is without steam supply. The special mode according to which at shutdown of steam due to inertial rotation of a rotor the generator some time continued to develop the electric power necessary for own needs, in particular for power supply of the main circulation pulsers was for this purpose developed.
Xenon lowered device power, and for maintenance of its "breath" it was required to bring a large number of cores of SUSA which also absorb neutrons out of an active zone. Thus, the aspiration of the personnel, despite everything, to make experiment conflicted to requirements of regulations.
14 hours 00 minutes — according to the program of tests are disconnected system of emergency cooling of the reactor. As the reactor cannot be operated without system of emergency cooling, it needed to be stopped. However the dispatcher of "Kiyevenergo" did not grant permission for muffling of the device. And the reactor continued to work without SAOR.
Causes of accident on the CNPP, its development were investigated by the leading scientists and experts with use of data on a condition of the reactor and its systems before accident, mathematical models of the power unit and its reactor installation and electronic computer facilities. As a result it was succeeded to restore the course of events, to formulate versions about the reasons and development of accident.
Deceleration of power of the reactor as it was already told, began at 1 hour 00 minutes on April 25. Then this process was stopped on request of the dispatcher of a power supply system. And further work on power deceleration began at 23 hours 10 minutes again.
Quality of the program of tests which was not properly prepared and agreed, was low. In it a number of the major provisions of regulations of operation was violated. Besides that additional security measures, in essence, were not provided in the program, it recommended shutdown of the system of emergency cooling of the reactor (SECR). Similar in general it is impossible to do. But here made. And the motivation was. During experiment there could be an automatic operation of SAOR that would prevent completion of tests in the mode of a vybeg. As a result of many hours the 4th reactor was operated without this very important element of a security system.