Core response of a PWR to a slug of under-borated water

A parameter study, incorporating stationary and transient core calculations, was carried out for a hypothetical boron dilution event in a pressurized water reactor, using the three-dimensional (313) neutron kinetics core model DYN3D in combination with a fast running semi-analytical coolant mixing model. It was assumed that a slug of deborated water was formed in one of the loops, due to a secondary-to-primary steam generator leakage during outage. It was further assumed that this slug is not recognised and that the first main coolant pump is started, in preparation to returning to power. At the initiation of pump start-up, the reactor is still in the cold and deeply sub-critical state. By varying the initial slug volume, it was found in stationary calculations that, for the given core configuration, slugs of less than 14 m(3) do not lead to re-criticality. Transient core calculations with larger slug volumes show a significant reactivity insertion and over-criticality. However, according to the calculations, even an over-criticality of about 2$ did not lead to safety-relevant consequences. The power excursion is mitigated and stopped by Doppler feedback. The influence of the cross-section library on the course of the transient was investigated, applying an alternative library. Differences in the global core parameters lead to quantitative differences in the time and height of the power excursion. In addition, it is shown that numerical diffusion has to be suppressed in order to describe the boron transport correctly, especially at low flow velocities. Otherwise the neutron kinetic core behaviour cannot be correctly modelled. (C) 2004 Elsevier B.V. All rights reserved.