System identification of nuclear reactor by dynamic data system methodology

A new modeling technique called Dynamic Data System (DDS) is used to identify the nuclear reactor dynamics using operating neutron flux data. The identification of the reactor dynamic characteristics is carried out by properly choosing the sampling interval and sample length. The internal lumped temperature feed-back time constant ??t can be identified from the operating data for the High Flux Isotope Reactor (HFIR) and the Oconee Power Reactor. The ??t value for the HFIR is found increasing from .08 sec. to .35 sec. during a 23 day fuel cycle operation. The ??t value for the Oconee reactor is found to be approximately 3 seconds. In addition, two abnormal frequency peaks are identified during the control rod bearing failure in HFIR. One frequency is always exactly twice as large in value as the other, e.g., 5.4 Hz vs. 2.7 Hz. These two frequencies are found shifting during the cycle operation.