VALIDATION OF NEUTRON PROPAGATION CALCULATIONS USING THE DORT AND DOTSYN CODES AND THE SPECIAL DOSIMETRY BENCHMARK EXPERIMENT AT THE FRENCHST LAURENT REACTOR
C. Garat et Cy. Rieg, VALIDATION OF NEUTRON PROPAGATION CALCULATIONS USING THE DORT AND DOTSYN CODES AND THE SPECIAL DOSIMETRY BENCHMARK EXPERIMENT AT THE FRENCHST LAURENT REACTOR, Nuclear Engineering and Design, 168(1-3), 1997, pp. 281-291
The St. Laurent B1 unit is a 900 MWe PWR that was loaded with MOX fuel
at the beginning of its 8th cycle. In order to provide the means of v
alidating the calculation of neutron propagation in a power reactor, t
he French national electric utility EDF installed numerous neutron dos
imeters in special surveillance capsules and in three ex-vessel chains
in the reactor pit. These dosimeters were irradiated for one cycle an
d then removed, providing measured reaction rates at locations both in
side and outside the reactor vessel. This paper describes the analysis
performed by Framatome using the DORT + DOTSYN 3-D synthesis modules
to obtain calculated dosimeter responses at each instrument location.
Several 2-D and 1-D models were developed, paying particular attention
to the geometrical representation of the reactor: the variable mesh c
apability of the DORT code was used and a special variable mesh genera
tor was written for it, along with a graphical interface for visual ch
ecks. A 100 energy group transport cross-section library derived from
the master ENDF/B-IV and ENDF/B-VI libraries was used. The results sho
w very good agreement between the calculated and measured fast neutron
reaction rates at the in-vessel locations, with an average calculated
(C) to measured (M) ratio, C/M, of 0.94 and a small dispersion in the
C/M distribution. The C/M values at the ex-vessel dosimeter locations
averaged 1.12, which indicates that the transport cross-section libra
ry slightly underestimates the attenuation of the fast neutrons passin
g through the reactor vessel. These results are comparable to those fr
om a similar analysis carried out independently based on the same St.
Laurent dosimetry benchmark, using a Monte Carlo code which is reporte
d elsewhere. (C) 1997 Elsevier Science S.A.