PhD abstract

Over the last decades, antineutrino experiments conducted at short and long baselines from nuclear reactors and using detection techniques based on the inverse β-decay (IBD) have revealed a systematic and significant deficit of detected antineutrinos compared to the predicted fluxes. Besides this flux discrepancy, called the reactor antineutrino anomaly (RAA), a difference in the shape of the measured spectra compared to state-of-the-art predictions has been observed. No evidence for an experimental bias has so far been detected as the origin of these discrepancies, and the interpretation of the reactor antineutrino anomaly as a neutrino oscillation with a sterile neutrino state is currently disfavored by recent very short baseline reactor experiments. The validity of the predictions is also questioned as the source of the observed discrepancies, and has motivated a revision of the modelings of reactor antineutrino spectra. In this context, a revisited prediction has been developed and is presented in this PhD dissertation. In a nuclear reactor, antineutrinos are typically emitted during the β-decays of the products originating from the fission of four actinides making up more than 99% of the thermal power released in the core (²³⁵U, ²³⁸U, ²³⁹Pu, ²⁴¹Pu). During a beta decay, an excited nucleus ejects in correlation an electron and an antineutrino. The total emission probability and the energy spectra associated to each of these two particles are characteristic of the parent and daughter nuclei, and depend on their nuclear and atomic structures. The antineutrino spectrum emitted by a reactor core then results from the superposition of thousands of β spectra. The revised prediction is based on the summation method which consists in modeling each of these β transitions. An advanced modeling of the β-decay theory has been used to include various effects due to the Coulomb interaction between the emitted electron and the daughter nucleus through a numerical treatment. The latest evaluated nuclear data are used to model the thousands β-decays contributing to a reactor antineutrino spectrum, including the most recent data from total absorption gamma spectroscopy measurements. A thorough propagation of the uncertainties associated to both the modeling and the nuclear data evaluation has also been investigated, allowing to produce a consistent and conservative uncertainty budget for the revisited spectra. The new summation modeling is finally compared with other state-of-the-art predictions, and its improvements and limitations are discussed in regards to IBD datasets from recent short and long baseline reactor experiments.

Key words

neutrino, reactor, particle, modelling, numerical computation, β decay

PhD thesis