Calculating Neutron Dosimeter Activation in Vver‐440 Surveillance Chains with Serpent
26th Symposium of AER on VVER Reactor Physics and Reactor Safety (2016, Helsinki, Finland)
Reactor physics experiments and code validation continues
Abstract
CALCULATING NEUTRON DOSIMETER ACTIVATION IN VVER-440 SURVEILLANCE CHAINS WITH SERPENTTuomas Viitanen and Jaakko Leppänen VTT Technical Research Centre of Finland tuomas.viitanen@vtt.fiABSTRACTThe structural integrity of reactor pressure vessels (RPVs) can be studied by preparing test specimens from the RPV material, irradiating the specimens in the surveillance position or positions at the reactor periphery and measuring the material properties of the irradiated samples. To match the measurements of the test specimens to the state of the reactor pressure vessel at a certain moment of time, the neutron exposure of the irradiated test specimens as well as the reactor pressure vessel need to be determined.The exposure, more precisely the neutron fluence, is usually calculated using either deterministic or Monte Carlo calculation codes. Accuracy of the computational estimates can be increased by means of neutron dosimetry, i.e. by normalizing or adjusting the computational results to match the measured activation of neutron dosimeters. For this reason, also the surveillance specimens of Loviisa-1 and Loviisa-2 VVER-440 units are always irradiated together with several neutron dosimeters: The axial profile in the neutron fluence is monitored using Fe/Ni dosimeter discs, in addition to which the fluence spectrum is measured using separate wire dosimeters.In the current work, the measured activities from neutron dosimeters irradiated in the surveillance position of Loviisa-1 unit are used to validate Monte Carlo reactor physics code Serpent for calculations at the reactor periphery, for example surveillance position or RPV. The neutron source in the Serpent calculation is generated based on a full-core power distribution from simulator code HEXBU-3D, and Serpent is only used to calculate the neutron transport from the source points in the reactor core to detector locations. Since the neutron flux decreases by orders of magnitude between the reactor core and the locations of interest, the convergence of the Monte Carlo transport solution needs to be accelerated using new weight-window based variance reduction techniques of Serpent 2.1.27. 26th Symposium of AER on VVER Reactor Physics and Reactor Safety 14 10 – 14 October 2016, Helsinki, Finland