Spectral indices measurements using miniature fission chambers at the MINERVE zero-power reactor at CEA using calibration data obtained at the BR1 reactor at SCK·CEN

Spectral indices are parameters of major importance to characterize a multiplying medium and to evaluate performances of neutronics calculation tools as well as quality of associated nuclear data. Experiments performed in 2004 at the CEA MINERVE facility loaded with the R1-UO2 lattice [1], using homemade miniature fission chambers and calibration data acquired at the SCK·CEN BR1 facility in 2001 [2], resulted in ambivalent conclusions. On one hand, spectral indices involving only fissile isotopes gave relevant results and consistent discrepancies between calculation and experiment. On the other hand, spectral indices involving both fissile and fertile isotopes showed inconsistent results depending on the type of calibration data used. In particular, the 238U(n, f)/235U(n, f) spectral indices resulted in values far from the calculations in all cases. In 2004, those differences between measurement and calculation were mainly attributed to an imprecise knowledge of the isotopic composition of the deposits loaded in the 238U chambers. At that time, priorities in the development of programs in MINERVE and other considerations about the manufacturing and calibration of new fission chambers did not allow to further investigate these discrepancies. In 2009, the preparation of a new program at the EOLE facility [1] and changes in the regulatory framework for management of fission chambers motivated the manufacturing of new detectors. At the same time, the re-installation of the R1-UO2 lattice in MINERVE gave the opportunity to carry out a new spectral indices measurement campaign. Nevertheless, although the isotopic composition of active deposits was better known than previously, the comparison between experimental results and calculations still resulted in inconsistent discrepancies. A sensitivity study indicated afterwards that, among all parameters to be considered for the analysis and the assessment of final uncertainty, calibration data have the greatest impact. In April 2010, a new calibration campaign conducted again at the BR1 facility allowed the CEA to reanalyze the spectral indices measurements performed in 2009. With this new set of calibration data, the experimental values of spectral indices finally matched the calculation results within the uncertainty margins. In addition, the confrontation of data from 2001 and 2010 calibration campaigns, both performed with the same experimental conditions, allowed to shed on the light dead time correction issues in the monitoring of calibration measurements. At the end, this also explained the inconsistencies observed during the 2004 campaign. This paper first, presents the two different analysis methods for getting the experimental value of a spectral index and, lists main parameters that influence the determination of the experimental value. It then details the calibration processes that are set-up at the MINERVE and BR1 facilities, describes the various experiments carried out and finally comments the results obtained.