Analysis and monitoring of in-core dynamics in pressurized water reactors

The applicability of the core-exit coolant temperature and the neutron power noise signals for core coolant flow rate estimation and for monitoring the sign of the moderator temperature coefficient of reactivity in pressurized water reactors (PWRs) are studied. Three kinds of models are developed to investigate the mechanisms involved in the noise signals: the phenomenological model, the distributed parameter model and the multi-nodal model. The Loss-of-Fluid Test (LOFT) reactor is selected as the baseline reactor for this study. The phase relationship between the core-exit coolant temperature and the neutron power signals is found to have a linear phase-frequency behavior in a certain frequency region (about 0.1-2 Hz) with lower and upper bounds which are functions of the system dynamics parameters. The slope of this phase is inversely proportional to the core flow rate. The phase at zero frequency extrapolated from this linear phase-frequency region is -180 degrees for a negative moderator temperature coefficient of reactivity and zero degrees for a positive moderator temperature coefficient.

The feasibility of using a stochastic estimator to estimate the flow rate using the core-exit coolant temperature and the neutron noise signals is also studied. The extended Kalman filter is developed and the validity of the method is tested against the 65% and 100% flow rate noise data from the LOFT reactor. The performance of the maximum likelihood function for flow estimation using noise signals is also investigated using the ambiguity function analysis technique. The sensitivity of the estimation to the variations of other system parameters is evaluated.