Studies on sources and transmission characteristics of fluctuations of variables in BWR

It is the purpose of this paper to clarify if one can determine some changes in (1) noise source distribution, and in (2) dynamic parameters of a BWR system by the observation of random fluctuation signals, it requires the statistical characteristics about noise sources. Therefore, unknown noise sources in a BWR have been an object of investigation to the noise analysts. To estimate unknown noise sources, the author applied multivariate autoregressive (MAR) model fitting to the fluctuation signals in a BWR plant system. With this model, statistical characteristics (power spectra) of the noise sources and the frequency response functions of the system were distinctively analysed. The experiments were performed in the Japan Power Demonstration Reactor (JPDR)-II. Fluctuation signals dealt with were the neutron density, the vessel pressure, the pump flow, the steam flow, the feed-water flow, and the channel flows, the fuel center temperature and the coolant inlet temperature of the instrumented fuel assemblies (IFAs). With the result of MAR model fitting, the power spectra of independent noise sources in the coolant channel flows, the pump flow, the fuel center temperature, the inlet temperature and the pressure were estimated. The frequency response functions among the process variables were also estimated with the MAR model, and compared with a theoretical model (Miida-Suda model). It is concluded through the experiments that this method has potential usefulness for (1) evaluation of theoretical noise models, and for (2) monitoring abnormal noise sources and parameters of the transfer functions of a power reactor plant system. (author)