The local power generation in the core of a Boiling Water Reactor (BWR) is directly related to the neutron flux. It depends strongly on the void (steam) fraction of the coolant via the nuclear reactivity feedback. The void oscillations in the core induce power oscillations due to this neutronic feedback. These power oscillations are coupled to oscillations of the thermal-hydraulic flow variables, such as density wave oscillations.The determination of the operating region where such flow-void-power oscillations may appear and the response of the reactor at these pointsis of large importance for BWRs.
In order to address the stability of BWRs, ASCOMP performs the analysis with two different approaches:
Advanced Coupled 3D Neutronics to Thermalhydraulics
This approach is based on a realistic model of the reactor (all core assemblies, lower and upper plenums, riser, steam separators, dome, downcomer) and its recirculation loop. 3D Neutronics equations (two energy groups with six precursor groups), rod conduction equation, and flow conservation equations (non-homogenous, non-equilibrium) constitute the main set of equations. These are discretized in space for evaluation of the core-statics solution. Furthermore the neutron fluxes are decomposed into their harmonics by the well-known nodal-modal method, which allows the analysis of the complex out-of-phase poweroscillations. The fundamental-mode oscillations represents the in-phase power oscillations (or the global oscillations).
Stability analysis results are presented in terms of the stability indicator “Decay Ratio” (DR) and in the classical Nyquist diagram of the system transfer function.
Time Series Analysis
Reactor-power time histories are measured or calculated at the locations of the in-core detectors. These may be Local Power Range Monitors (LPRM) or Average Power Range Monitors (APRM). A time series reflects a stable or unstable oscillatory power evolution at any operational point of the BWR core after a specific or non-specific parameter perturbation.
ASCOMP analyzes the reactor power oscillations using parametric and non-parametric approaches. The parametric approach (whether Auto Regressive (AR) or Auto Regressive Moving-Average (ARMA)) involves determination of the model parameters and model order. The model order may be evaluated by the Akaike Information Criterion (AIC) or other more restrictive criteria that penalize the high model order. Finally an optimized model order is sought among the poles of the transfer function. The outcomes of the time-series analysis are the decay ratio and the resonance frequency related to the specific operation point. These are obtained from the poles of the system transfer function. The resonance frequency could be alternatively evaluated from the power spectral density diagram, during the non-parametric analysis.