DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Extremely accurate sequential verification of RELAP5-3D

Abstract

Large computer programs like RELAP5-3D solve complex systems of governing, closure and special process equations to model the underlying physics of nuclear power plants. Further, these programs incorporate many other features for physics, input, output, data management, user-interaction, and post-processing. For software quality assurance, the code must be verified and validated before being released to users. For RELAP5-3D, verification and validation are restricted to nuclear power plant applications. Verification means ensuring that the program is built right by checking that it meets its design specifications, comparing coding to algorithms and equations and comparing calculations against analytical solutions and method of manufactured solutions. Sequential verification performs these comparisons initially, but thereafter only compares code calculations between consecutive code versions to demonstrate that no unintended changes have been introduced. Recently, an automated, highly accurate sequential verification method has been developed for RELAP5-3D. The method also provides to test that no unintended consequences result from code development in the following code capabilities: repeating a timestep advancement, continuing a run from a restart file, multiple cases in a single code execution, and modes of coupled/uncoupled operation. In conclusion, mathematical analyses of the adequacy of the checks used in the comparisons are provided.

Authors:
 [1];  [2];  [2]
  1. Idaho National Lab. (INL), Idaho Falls, ID (United States)
  2. Bettis Atomic Power Lab., West Mifflin, PA (United States)
Publication Date:
Research Org.:
Idaho National Lab. (INL), Idaho Falls, ID (United States)
Sponsoring Org.:
USDOE
OSTI Identifier:
1255242
Report Number(s):
INL/JOU-14-33443
Journal ID: ISSN 0029-5639
Grant/Contract Number:  
AC07-05ID14517
Resource Type:
Accepted Manuscript
Journal Name:
Nuclear Science and Engineering
Additional Journal Information:
Journal Volume: 182; Journal Issue: 1; Journal ID: ISSN 0029-5639
Publisher:
American Nuclear Society
Country of Publication:
United States
Language:
English
Subject:
22 GENERAL STUDIES OF NUCLEAR REACTORS; 97 MATHEMATICS AND COMPUTING; RELAP5-3D; verification; governing equations; coverage; hypothesis testing

Citation Formats

Mesina, George L., Aumiller, David L., and Buschman, Francis X. Extremely accurate sequential verification of RELAP5-3D. United States: N. p., 2015. Web. doi:10.13182/NSE14-151.
Mesina, George L., Aumiller, David L., & Buschman, Francis X. Extremely accurate sequential verification of RELAP5-3D. United States. https://doi.org/10.13182/NSE14-151
Mesina, George L., Aumiller, David L., and Buschman, Francis X. Thu . "Extremely accurate sequential verification of RELAP5-3D". United States. https://doi.org/10.13182/NSE14-151. https://www.osti.gov/servlets/purl/1255242.
@article{osti_1255242,
title = {Extremely accurate sequential verification of RELAP5-3D},
author = {Mesina, George L. and Aumiller, David L. and Buschman, Francis X.},
abstractNote = {Large computer programs like RELAP5-3D solve complex systems of governing, closure and special process equations to model the underlying physics of nuclear power plants. Further, these programs incorporate many other features for physics, input, output, data management, user-interaction, and post-processing. For software quality assurance, the code must be verified and validated before being released to users. For RELAP5-3D, verification and validation are restricted to nuclear power plant applications. Verification means ensuring that the program is built right by checking that it meets its design specifications, comparing coding to algorithms and equations and comparing calculations against analytical solutions and method of manufactured solutions. Sequential verification performs these comparisons initially, but thereafter only compares code calculations between consecutive code versions to demonstrate that no unintended changes have been introduced. Recently, an automated, highly accurate sequential verification method has been developed for RELAP5-3D. The method also provides to test that no unintended consequences result from code development in the following code capabilities: repeating a timestep advancement, continuing a run from a restart file, multiple cases in a single code execution, and modes of coupled/uncoupled operation. In conclusion, mathematical analyses of the adequacy of the checks used in the comparisons are provided.},
doi = {10.13182/NSE14-151},
journal = {Nuclear Science and Engineering},
number = 1,
volume = 182,
place = {United States},
year = {Thu Nov 19 00:00:00 EST 2015},
month = {Thu Nov 19 00:00:00 EST 2015}
}