skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Variance Estimation in Monte Carlo Eigenvalue Simulations Using Spectral Analysis Method

Authors:
ORCiD logo [1]
  1. ORNL
Publication Date:
Research Org.:
Oak Ridge National Lab. (ORNL), Oak Ridge, TN (United States)
Sponsoring Org.:
USDOE
OSTI Identifier:
1376412
DOE Contract Number:
AC05-00OR22725
Resource Type:
Conference
Resource Relation:
Conference: 2017 ANS Annual Meeting - San Francisco, Washington, United States of America - 6/11/2017 12:00:00 AM-6/15/2017 12:00:00 AM
Country of Publication:
United States
Language:
English

Citation Formats

BANERJEE, KAUSHIK . Variance Estimation in Monte Carlo Eigenvalue Simulations Using Spectral Analysis Method. United States: N. p., 2017. Web.
BANERJEE, KAUSHIK . Variance Estimation in Monte Carlo Eigenvalue Simulations Using Spectral Analysis Method. United States.
BANERJEE, KAUSHIK . Thu . "Variance Estimation in Monte Carlo Eigenvalue Simulations Using Spectral Analysis Method". United States. doi:. https://www.osti.gov/servlets/purl/1376412.
@article{osti_1376412,
title = {Variance Estimation in Monte Carlo Eigenvalue Simulations Using Spectral Analysis Method},
author = {BANERJEE, KAUSHIK .},
abstractNote = {},
doi = {},
journal = {},
number = ,
volume = ,
place = {United States},
year = {Thu Jun 01 00:00:00 EDT 2017},
month = {Thu Jun 01 00:00:00 EDT 2017}
}

Conference:
Other availability
Please see Document Availability for additional information on obtaining the full-text document. Library patrons may search WorldCat to identify libraries that hold this conference proceeding.

Save / Share:
  • The number of tallies performed in a given Monte Carlo calculation is limited in most modern Monte Carlo codes by the amount of memory that can be allocated on a single processor. By using domain decomposition, the calculation is now limited by the total amount of memory available on all processors, allowing for significantly more tallies to be performed. However, decomposing the problem geometry introduces significant issues with the way tally statistics are conventionally calculated. In order to deal with the issue of calculating tally variances in domain decomposed environments for the Shift hybrid Monte Carlo code, this paper presentsmore » an alternative approach for reactor scenarios in which an assumption is made that once a particle leaves a domain, it does not reenter the domain. Particles that reenter the domain are instead treated as separate independent histories. This assumption introduces a bias that inevitably leads to under-prediction of the calculated variances for tallies within a few mean free paths of the domain boundaries. However, through the use of different decomposition strategies, primarily overlapping domains, the negative effects of such an assumption can be significantly reduced to within reasonable levels.« less
  • The reactor physics codes used at the Savannah River Site (SRS) to predict reactor behavior have been continually benchmarked against experimental and operational data. A particular benchmark variable is the observed initial critical control rod position. Historically, there has been some difficulty predicting this position because of the difficulties inherent in using computer codes to model experimental or operational data. The Monte Carlo method is applied in this paper to study the initial critical control rod positions for the SRS K Reactor. A three-dimensional, full-core MCNP model of the reactor was developed for this analysis.
  • Monte Carlo eigenvalue calculations using a fixed number of histories per generation are biased. Previous work has shown that the bias in eigenvalue is generally smaller than observed standard deviations. In this paper, the authors propose a method for bounding [delta][psi] the bias in the unnormalized fission source. In addition, the authors construct a plausibility argument suggesting that the bias is generally insignificant. Results presented are for discretized diffusion calculations based on a Monte Carlo-generated source. Additional calculations have been run for true transport problems using a very fast, but more conventional, Monte Carlo code.
  • CCFE perform Monte-Carlo transport simulations on large and complex tokamak models such as ITER. Such simulations are challenging since streaming and deep penetration effects are equally important. In order to make such simulations tractable, both variance reduction (VR) techniques and parallel computing are used. It has been found that the application of VR techniques in such models significantly reduces the efficiency of parallel computation due to 'long histories'. VR in MCNP can be accomplished using energy-dependent weight windows. The weight window represents an 'average behaviour' of particles, and large deviations in the arriving weight of a particle give rise tomore » extreme amounts of splitting being performed and a long history. When running on parallel clusters, a long history can have a detrimental effect on the parallel efficiency - if one process is computing the long history, the other CPUs complete their batch of histories and wait idle. Furthermore some long histories have been found to be effectively intractable. To combat this effect, CCFE has developed an adaptation of MCNP which dynamically adjusts the WW where a large weight deviation is encountered. The method effectively 'de-optimises' the WW, reducing the VR performance but this is offset by a significant increase in parallel efficiency. Testing with a simple geometry has shown the method does not bias the result. This 'long history method' has enabled CCFE to significantly improve the performance of MCNP calculations for ITER on parallel clusters, and will be beneficial for any geometry combining streaming and deep penetration effects. (authors)« less
  • Monte Carlo is potentially the most accurate method for modeling particle transport since it allows detailed geometry representation and use of pointwise cross sections. Its statistical nature, however, introduces additional concerns related to the reliability of the results. This is a special concern in eigenvalue Monte Carlo simulations because the possibility of significantly underestimating the eigenvalue is real. A series of test problems is introduced and utilized to demonstrate and clarify some of the source convergence issues. Analysis of the results is intended to help formulate improved diagnostic methods for identifying or preventing false source convergence.