Search

• Gamma noise experiments using CeBr 3 detectors were conducted in the CROCUS reactor. • Comparison to neutron noise experiments and Serpent 2 IFP predictions are shown. • Lower level threshold effects on gamma APSDs and CPSDs are reported. • Uncertainties were estimated using a bootstrap procedure. We experimentally demonstrate the advantages of gamma detection for noise measurements to determine the prompt neutron decay constant α of a nuclear reactor using the power spectral density (PSD) method, coupled to a new uncertainty estimation scheme based on bootstrapping. We compare the relative precision of gamma noise measurements and neutron noise measurements performed in the CROCUS research reactor. The reference neutron noise experiment employs two large 235U fission chambers, whereas the new gamma noise experiment is based on two CeBr 3 scintillators. The results demonstrate that gamma noise is superior to neutron noise with respect to precision, giving an advantage with regards to measurement time to reach desired uncertainty levels: the uncertainty budget shows a relative uncertainty of 3.6% and 1.3% on the prompt neutron decay constant with neutron and gamma noise, respectively. Moreover, thanks to bootstrapping, the reporting of full distributions – rather than assuming Gaussian spreads – of kinetic parameters is shown to be relevant in order to provide an accurate representation of the experimental results in some cases. In addition, we notably discuss the lower level discrimination threshold and its effect on the gamma PSDs. Gamma ray cross PSDs exhibit a behaviour contrary to that of a standard neutron acquisition system, and we discuss the implications. Gamma noise could prove to be a cheaper, more precise, and more flexible method to determine reactor kinetics. [ABSTRACT FROM AUTHOR]

The transfer function is an important characteristic quantity of a nuclear reactor, since it contains the kinetic parameters. It expresses the response of a nuclear reactor to a disturbance of a certain frequency. If it is determined experimentally, it can be used to draw conclusions about the kinetic parameters. This article presents results of measurements of the zero power transfer function of the AKR-2 reactor at TU Dresden together with new data analysis methods. These measurements are compared to the theoretical zero power transfer function with kinetic parameters obtained via Monte Carlo simulations with MCNP and Serpent. To this end, advanced data analysis techniques based on a bootstrapping algorithms are employed. These techniques suppress the signal outside multiples of the fundamental frequency and additionally allow to obtain the full probability distribution of a peak in the frequency domain. This allowed for a reliable estimation of the mean value and uncertainty estimates of measured data of the zero power transfer function and the quantification of deviations between the experiments and the computations. It also made it possible to determine the phase of the zero power transfer function of AKR-2 for the first time. The experiments and computations are in agreement within the estimated uncertainties. [ABSTRACT FROM AUTHOR]

In this paper we present the methodology developed to analyze the detector response of neutron noise experiments performed at EPFL CROCUS and TUD AKR-2 reactors in the framework of the H2020 CORTEX project. The noise, or modulation, is induced by controlled oscillations of neutron absorbers or reactor fuel elements. The quantities of interest are detector response spectral power and phase shift with respect to a selected reference detector. For the analysis we employ data normalization and bootstrapping with replacement. It acts as a frequency response comb, letting through only the harmonics of the base oscillation frequency, and eliminates biases due to correlations of consecutive oscillations while maintaining the frequency resolution of the original data set. The correlations stemming from a common neutron noise inducing device are also considered in the analysis. • Neutron noise analysis. • Spectral power and phase angle. • Oscillation cycle bootstrapping with replacement. • Non-linear noise effects. [ABSTRACT FROM AUTHOR]