Development and Investigation of Reactivity Measurement Methods in Subcritical Cores

Licentiate thesis, 2005

Subcriticality measurements during core loading and in future accelerator driven systems have a clear safety relevance. In this thesis two subcriticality methods are treated: the Feynman-alpha and the source modulation method. The Feynman-alpha method is a technique to determine the reactivity from the relative variance of the detector counts during a measurement period. The period length is varied to get the full time dependence of the variance-to-mean. The corresponding theoretical formula was known only with stationary sources. In this thesis, due to its relevance for novel reactivity measurement methods, the Feynman-alpha formulae for pulsed sources for both the stochastic and the deterministic cases are treated. Formulae neglecting as well as including the delayed neutrons are derived. The formulae neglecting delayed neutrons are experimentally verified with quite good agreement. The second reactivity measurement technique investigated in this thesis is the so-called source modulation technique. The theory of the method was elaborated on the assumption of point kinetics, but in practice the method will be applied by using the signal from a single local neutron detector. Applicability of the method therefore assumes point kinetic behaviour of the core. Hence, first the conditions of the point kinetic behaviour of subcritical cores was investigated. After that the performance of the source modulation technique in the general case as well as and in the limit of exact point kinetic behaviour was examined. We obtained the unexpected result that the method has a finite, non-negligible error even in the limit of point kinetic behaviour, and a substantial error in the operation range of future accelarator driven subcritical reactors (ADS). In practice therefore the method needs to be calibrated by some other method for on-line applications.