2008 article

Uncertainty Quantification, Sensitivity Analysis, and Data Assimilation for Nuclear Systems Simulation

Abdel-Khalik, H., Turinsky, P., Jessee, M., Elkins, J., Stover, T., & Iqbal, M. (2008, December). NUCLEAR DATA SHEETS, Vol. 109, pp. 2785–2790.

By: H. Abdel-Khalik  n, P. Turinsky n, M. Jessee n, J. Elkins n, T. Stover n & M. Iqbal  n

co-author countries: United States of America πŸ‡ΊπŸ‡Έ
Source: Web Of Science
Added: August 6, 2018

Reliable evaluation of nuclear data will play a major role in reduction of nuclear systems simulation uncertainties via the use of advanced sensitivity analysis (SA), uncertainty quantification (UQ), and data assimilation (DA) methodologies. This follows since nuclear data have proven to constitute a major source of neutronics uncertainties. This paper will overview the use of the Efficient Subspace Method (ESM), developed at NCSU, to overcome one of the main deficiencies of existing methodologies for SA/UQ/DA, that is the ability to handle codes with large input and output (I/O) data streams, where neither the forward nor the adjoint approach alone are appropriate. We demonstrate the functionality of ESM for an LWR core, a boiling water reactor, and a fast reactor benchmark experiment, the ZPR6/7A assembly. This work demonstrates the capability of adjusting cross section data thereby providing guidance to cross section evaluation efforts by identification of key cross sections and associated energy ranges that contribute the most to the propagated core attributes uncertainties.