Development of computational methods and their applications for the analysis of nuclear power plants

Paper in proceeding, 2008

In this paper, the modelling of nuclear reactors at the system level is presented from different viewpoints. A specificity of nuclear reactors is their multi-physics and multi-scale character. The multi-physics nature comes from the interdependency between different fields governing the physics of such systems (neutron transport, heat transfer and fluid dynamics). The coupling between physical phenomena across various characteristic lengths, varying from the micro-scales to the macro-scales, requires a multi-scale treatment. Specific modelling techniques are thus required for the simulation of nuclear reactors and are presented in this paper. The use of such techniques for both time-independent simulations and time-dependent simulations are dealt with, and examples of such simulations performed at the Department of Nuclear Engineering, Chalmers University of Technology are presented. Time-independent simulations are mostly carried out for in-core fuel management purposes, i.e. for designing a core loading allowing running the reactor in a safe and economical manner. For the time-dependent simulations, two classes of problems are encountered. If the system undergoes small stationary fluctuations whereas the mean values of the variables remain constant, linear theory can be used to find the governing equations of such fluctuations. The analysis of measurements for such small fluctuations is referred to as noise analysis. Due to the stationary character of the fluctuations, such calculations are more easily performed in the frequency-domain. If the system undergoes large fluctuations and/or if the mean values of the variables are changing with time, the equations are to be solved in the time-domain. Such class of problems is usually referred to as transient analysis. These two approaches have complementary fields of applicability. On the one hand, transient analysis allows verifying that the system always fulfils its design criteria in all foreseeable situations. One the other hand, noise analysis allows interpreting noise measurements aimed at monitoring the plant on-line and at early detecting anomalies.