The main problem associated with the operation of nuclear power plants is the problem of correct and timely diagnosis of failure or violation. The mistakes of technologists can lead to severe damage to nuclear power plants, or simply to a reduction in the utilization rate of installed capacity. The personnel make the main mistakes in the conditions of time shortage and being in a stressful situation during the development of an accident, when assessing changes in emergency parameters is not always possible, which leads to incorrect diagnosis of initial events (IE). Any IE leads to the deviation of the monitored parameters from normal values (values that are inherent to normal operation). The operator-technologist needs as soon as possible to determine the parameter (or parameters) for which the change occurs and, by performing a certain sequence of actions, bring their values to the norm. For information support of the operator-technologist is the information computing system (ICS). In the ICS, it is possible to control all parameters affecting the safe operation of the equipment, for each of which there are certain values, the deviation from which may lead to an accident. The paper considers the issues of the need to separate the measured parameters of NPP into separate groups that uniquely characterize the state of the critical security features. In particular, the necessity of consideration as the most important mass parameter of the primary coolant is shown. The issues of level measurement problems in the pressure compensator are discussed as the most important from the point of view of determining the mass of the primary coolant. Methods are proposed for determining the operability of sensors, the method of calculating a reliable level in the volume compensator. Estimates of the computational efficiency of the proposed methods are given.

No abstract

During the design of the nuclear power plant with VVER-1000 units, which today form the basis of the nuclear power industry of Ukraine, the development of information and computing systems (IСS) was insufficient to solve many tasks due to the small capacities of computing systems. Currently, newer ones, the computing capabilities of which are thousands of times greater than their predecessors, have replaced many IСS. IСS download currently does not exceed 1 %...2 %, this is due to the use of simplified models. A model is proposed that adequately describes the considered processes and can work in real time, which significantly increases the range of use. It has a small computational effort, so it can be implemented on the existing IСS. A method has been proposed which, as a result of testing, has shown its effectiveness and, in some cases, higher accuracy and resolution.