American Nuclear Society

Home / Publications / Journals / Nuclear Technology / Volume 201 / Number 3

Upgrading Limiting Peak-Power Analysis Techniques with Modern Validation and Uncertainty Quantification for the Advanced Test Reactor

Samuel E. Bays, Cliff B. Davis, Periann A. Archibald

Nuclear Technology / Volume 201 / Number 3 / March 2018 / Pages 209-227

Technical Paper /

Received:May 22, 2017
Accepted:December 6, 2017
Published:February 16, 2018

This work supports the acceptability of the two-dimensional deterministic transport code HELIOS to replace the legacy diffusion code PDQ for computing the peak-power performance parameters of the Advanced Test Reactor (ATR). The 95% Confidence Rule, commonly used in the commercial reactor sector, is explored to develop the so-called reliability factors that provide statistical confidence that the peak-power limits within the hottest location along a fuel plate, referred to as the hot stripe, will not be exceeded. Additionally, an alternative “legacy” methodology was explored that attempts to mimic the exact PDQ analysis process used for defining the peak-power limits. The legacy methodology involves interpolating power between regions at azimuthal boundaries subtending the regions of interest.

In order to apply the 95% Confidence Rule, a statistically significant calculation-to-measurement bias must first be established. Unlike the commercial world, where thousands of power observations can be collected every cycle using online flux-mapping instrumentation, the ATR power distribution must be measured during “depressurized” zero-power measurements using fission wires in polyethylene wands. In 2012, fission wire activation data were collected during a flux run in the Advanced Test Reactor Critical Facility. Also to improve statistical validity, archival data from ATR zero-power flux runs from 1977, 1986, and 1994 were digitized from scanned reports and used to create new benchmark models. Borrowing from least-squares adjustment methods commonly used for neutron activation spectroscopy, adjusted fission wire powers were calculated for all four data sets. The mean and standard deviation of the bias between a priori calculated and adjusted wire powers were then taken as the bias and uncertainty used in the 95% Confidence Rule.

Questions or comments about the site? Contact the ANS Webmaster.