No abstract

The modern requirements for monitoring the state of reactor equipment during reactor operation are implemented by systems for on-line diagnostics, specifically, vibrational diagnostics system. Effective operation of this system is possible only after the auto-and cross-spectral characteristics of the registered signals have been interpreted. To this end, a noise experiment was performed on the first power-generating unit of the Kalinin nuclear power plant.Detectors. Three types of sensors recorded the noise signal with the aid of a 20-charmel system during the stationary operation of the reactor at 80% nominal power: three extracore ionization chambers (IC7, IC14, IC21), three pressure pulsation sensors (PPS) (Q4, $4, Q3), and 14 intracore direct-charge detectors (DCD) (two assemblies with each assembly containing seven DCDs) (Fig. 1). Experimental conditions were produced for switching the signals from the DCDs chosen by a specially developed method. Approximately 40 of the 20-charmel records permit investigating the following: reactivity and local effects caused by vibrations of the reactor vessel, intravessel apparatus, and specifically vibrations of fuel assemblies which have operated for different periods of time; distribution of the amplitude of different acoustic standing waves inside the reactor vessel; and, transfer functions whose parameters, characterizing the thermohydraulic state of the core, cannot be measured directly.Effects Masking the Vibrations of lntravessel Equipment. The most significant effects are the effects due to the transport of coolant nonuniformities, for example, temperature fluctuations, from the entrance into the core along the fuel channel. This is the so-called source 6Tin. The transport defect is also characteristic of the source 5G --fluctuations of the coolant flow rate. A detailed exposition of the thermohydraulic sources of neutron noise is given in [1]. Such wide-band effects (Fig. 2) cause a delay of one signal from DCD with respect to another by a time to; in spectral language, this is expressed as a linear frequency dependence of the phase of the cross-spectral power density. If the same frequency range contains the vibrational frequency of the intravessel devices, then the expected set of phases (most often in-or antiphaseness of signals from two neutron sensors) will not occur. To eliminate masking of this type, the phase characteristic must be corrected for the linear dependence. There are at least two such dependences. A direct calculation of the coolant velocity with respect to the parameter t o in the range 0-1 Hz is found to be a rough estimate. As shown in [1], the nonuniformity of the power-release field over the vertical coordinate in the core must be taken into account. Transport effects are also observed in the sensor pairs DCD-nearest IC (Fig. 3) and in the pairs DCD-DCD, placed in neighboring fuel assemblies. These effects have a high diagnostic value, but from the standpoint of vibrational diagnostics of intravessel equipment, they are sources of masking ...

No abstract

No abstract

No abstract