Reactivity modulation experiments for nuclear data in CROCUS within a CEA-EPFL collaboration

¹ CEA, DES, IRESNE, DER, SPESI, LP2E, Cadarache F-13108 Saint-Paul-Lez-Durance, France
² Laboratory for Reactor Physics and Systems Behaviour (LRS), Ecole polytechnique fédérale de Lausanne (EPFL), 1015 Lausanne, Switzerland

^* benoit.geslot@cea.fr

Published online: 21 November 2023

Abstract

In nuclear research reactors, integral experiments are powerful tools to measure integral core parameters, such as the delayed neutron fraction. Within the scope of the point kinetic approximation, reactivity modulation experiments can be used for probing the reactor transfer function and then infer integral parameters of the core.

In this context, Commissariat à l’Energie Atomique et aux Energies Alternatives (CEA) and Ecole Polytechnique Fédérale de Lausanne (EPFL) have been collaborating for developing a probe device (PISTIL) and measurement setup adapted to the CROCUS zero power research reactor operated by EPFL (Lausanne, Switzerland).

Despite some mechanical limitations of PISTIL, its maximum reactivity worth was measured with a good precision and repeatability using different methods (8.82 ±0.07 pcm), and its value is found rather close to the simulated one using TRIPOLI-4 (9.4 ± 0.4 pcm with JEFF-3.3). Above 1 Hz, the shape of the used modulation is pseudo-sinusoidal, with only a few well defined harmonics of excitation. The strongest harmonic only was analyzed using standard signal processing algorithms such as the Fourier transform and the Bartlett estimator. Twelve data points were produced in the range 0.5 Hz to 200 Hz, with uncertainty ranging from 1 % to 15 %. The prompt decay constant was measured at 150 ± 3 rad/s. Below 1 Hz, stepwise modulations were used with pseudo-random time sequences, which allowed exciting at once a large number of frequencies. Around 150 data points were produced in this particularly interesting frequency domain, between 1.6 mHz and 0.75 Hz, thanks to the use of three distinctive sequences with different base frequencies and overlapping ranges. The amplitude and phase of the RTF were measured satisfactorily, with uncertainties below 1 % for the strongest harmonics. The shape of the RTF was found consistent with the predictions of both JEFF-3.3 and ENDF/B-VII.1 libraries.