Development of a Numerical Tool for Estimating the Neutron Noise in Nuclear Reactors

Licentiate thesis, 2009

The neutron noise, i.e. the time-variations of the neutron flux around its mean value, can be used for obtaining important safety parameters in a nuclear reactor, e.g. the decay ratio for a boiling water reactor and the moderator temperature coefficient for a pressurized water reactor. The main benefit of using the neutron noise is that it can be taken from the existing detectors and there is therefore no need for perturbing reactor operation. It would thus be advantageous to be able to calculate the neutron noise with a relatively simple model with a relatively small computational effort, a work initiated earlier at the Department of Nuclear Engineering, Chalmers University of Technology. The objectives of this report are thus to validate some of the assumptions in the previously developed model and to evaluate various ways of improving it. A comparison is thus conducted between P1 and diffusion theories with two energy groups for a 1-dimensional 2-region reactor. The previous model used the Finite Difference Method, a method which is simple and easy to implement for spatial discretization of the equations. The drawback is that the accuracy is low unless the size of the nodes of the reactor model is small. To enhance the neutronics while keeping the size of the nodes in the range of 15-20 cm, theAnalytical Nodal Method was implemented for the calculation of the neutron noise. The results fromthis study are that there is no need in practical situations for using higher order theory for neutron noise calculations since diffusion theory gives results that are accurate enough. The implemented Analytical Nodal Method also shows high agreement with the analytically calculated neutron noise.