Simulation of Noise Experiments Using a Stochastic Differential Equation Model

  Eshed Magali  

In this work, OpenMC, a community-developed Monte Carlo code, is used to model the Australian OPAL reactor's core. The OpenMC model is used to simulate experiments performed in OPAL, and the measured results are used to benchmark both the model and the computation.
As for the modelling process, OpenMC proves to be a convenient and flexible tool with which intricacies of many geometrical features were simply and accurately described. It also proves to be efficient in the generation of many different core states and to be simple to use in tallying the flux in several regions around the core.
As for the accuracy of the computation, measurements of the control rods' worth, measurements of critical configurations and measurements of the thermal flux are compared to simulated results. A decent agreement is found between simulations and measurements concerning various safety related quantities. The integral worth values of each control rod, and therefore the total anti-reactivity worth of the safety rods and the shutdown margin, are found to be predicted within \(5\%\) accuracy. Moreover, the differential worth values (and thus, approximately, the axial flux peaking factor in the region of the control rods) are found to be predicted within \(5\%\) accuracy as well (up to an exception). Furthermore, the axial flux peaking factor values in the outskirts of the core are found to be predicted within \(10\%\) accuracy. Finally, the OpenMC simulation produces a prediction which is the closest on average to the measured naught critical reactivity, compared to the results of other groups worldwide, such that it is also relatively consistent.