Published September 2019 | Version v1
Journal article

Neutron noise analysis using the basic element method

  • 1. University of Chinese Academy of Sciences, Beijing (China)
  • 2. Institute of Modern Physics, Chinese Academy of Sciences, Lanzhou (China)
  • 3. Joint Institute for Nuclear Research, LIT, Dubna, Moscow (Russian Federation)
  • 4. Joint Institute for Nuclear Research, FLNP, Dubna, Moscow (Russian Federation)
  • 5. Institute for Physics and Nuclear Engineering, DFCTI, Bucharest (Romania)

Description

Highlights: • Algorithm for disentangling neutron noise from reactor-operations' power changes. • Algorithm tested on both static and dynamic states of the reactor. • Algorithm optimized for the best separation between neutron noise and power changes. • Algorithm speed adequate for real-time monitoring. - Abstract: Neutron noise spectra in nuclear reactors are a convolution of multiple induced reactivities. Under normal IBR-2M reactor operating conditions the full noise range can reach ±22% with respect to average power. Slow changes in average power (induced for instance by the control rods) are thereby much smaller than the neutron noise, however they are of importance in conducting the safe operation of the reactor. The problem we address is the separation of neutron noise from the slow power-changes. For this task we used an algorithm based on the basic element method (BEM) developed at Laboratory of Information Technologies – LIT JINR. The algorithm was applied to both static and dynamic states of the reactor in the 0–2 MW power range. The average processing time for one point on a x8664 Intel Core i5-4570 Sandy Bridge processor at 3.20 GHz, was approximately 50 μs, which allows the use of the MSPPA-6 algorithm in real time.

Availability note (English)

Available from http://dx.doi.org/10.1016/j.anucene.2019.04.008

Additional details

Identifiers

DOI
10.1016/j.anucene.2019.04.008;
PII
S0306454919301860;

Publishing Information

Journal Title
Annals of Nuclear Energy (Oxford)
Journal Volume
131
Journal Page Range
p. 475-482
ISSN
0306-4549
CODEN
ANENDJ

Optional Information

Notes
© 2019 Elsevier Ltd. All rights reserved.