Development of an integrated neutronic/thermal-hydraulic tool for estimating fluctuations in PWRs

Sammanfattning: Monitoring and diagnostics of a reactor core at a given state is very important for detecting potential problems and anomalies at an early stage. Several methods exist for these purposes, but the majority of them require a disruption of the plant operation, e.g. by shut-down or by perturbation of the system. The knowledge of the noise, i.e. the fluctuations of a parameter around its time-averaged value, can be used for on-line/off-line core monitoring and diagnostics without any disruption of normal operation. Measurements of reactor parameters can be used for estimating dynamical parameters, as the Moderator Temperature Coefficient (MTC) in a Pressurized Water Reactor (PWR), or for detecting anomalies, as unseated fuel assemblies in Boiling Water Reactors (BWRs) or flow blockages. Since the interpretation of the neutron noise aims at finding the driving perturbation responsible for the observed noise in the system and the limited number of detectors makes this attempt difficult, it would be beneficial to be able to numerically estimate in-core noise for any arbitrary noise source so that the comparison between the measured and calculated noise would allow finding the root cause of the observed fluctuations. The intention of this work is thus to calculate the in-core noise for commercial systems using coupled thermal-hydraulic/neutronic models. This includes the determination of the neutron noise, as well as of fluctuations in fuel temperature, moderator density and flow velocity. The work is divided into two main parts, neutron kinetics and thermal-hydraulics. On the neutron kinetic side, P1 and diffusion theories were compared with the conclusion that diffusion theory was accurate enough for noise calculations. The spatial discretization was also tested through the finite difference method and the Analytical Nodal Method (ANM), where it was concluded that both could be used for calculating the neutron noise. The comparisons between those two methods revealed that finite differences seems to be accurate enough for most practical applications. Both methods were successfully validated against analytical solutions. The other part covered in this work was the development for PWRs of a thermal-hydraulic model coupled to the neutronic model. This model was validated against calculations by the commercial Computational Fluid Dynamics (CFD) code FLUENT. The integrated neutronic/thermal-hydraulic model was proven viable and was also benchmarked against a RELAP5/PARCS model of a commercial PWR with mostly good results.