26.02.2013 Views

Program - Brookhaven National Laboratory

Program - Brookhaven National Laboratory

Program - Brookhaven National Laboratory

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

DE 7 5:30 PM<br />

Application of the PSI-NUSS Tool for Estimation of Nuclear Data Related K-eff<br />

Uncertainties for the OCED/NEA WPNCS UACSA Phase 1 Benchmark<br />

Ting Zhu, Alexander Vasiliev, Hakim Ferroukhi<br />

Paul Scherrer Institut<br />

At the Paul Scherrer Institut, a methodology for the propagation of nuclear data uncertainties into criticality<br />

safety calculations with the Monte Carlo code MCNP(X), titled as PSI-NUSS, is under development.<br />

The primary purpose is to provide a complementary and innovative option for the assessment of nuclear<br />

data related uncertainties versus the traditional approach commonly applied in many state-of-the-art criticality<br />

safety evaluation (CSE) methodologies and which relies on estimating biases/uncertainties based on<br />

comprehensive validation studies against representative critical benchmark experiments. A description of<br />

the PSI-NUSS methodology, as well as its assessment against the more rigorous and theoretically grounded<br />

Total Monte Carlo methodology developed at NRG, is presented in an accompanying paper. In the present<br />

paper, the PSI-NUSS methodology is applied to quantify nuclear data uncertainties for the OCED/NEA<br />

WPNCS UACSA Phase 1 benchmark. One underlying reason is that the previous PSI CSE contribution for<br />

this benchmark was based on using a more conventional approach involving ’engineering guesses’ in order to<br />

estimate uncertainties in the calculated effective multiplication factor caused by nuclear data uncertainties.<br />

Therefore, as the PSI-NUSS methodology aims precisely at integrating a more rigorous treatment of these<br />

types of uncertainties for CSE, its application to the UACSA benchmark Phase I is conducted here with<br />

two main objectives. First, the nuclear-data related uncertainty component is estimated and compared<br />

to results obtained by other participants using different codes/libraries and methodologies. Secondly, the<br />

estimated k-eff uncertainty is compared to the results obtained with the previously employed approach.<br />

Outlook for further verification and validation studies for the PSI-NUSS methodology is presented and the<br />

eventual updates of the overall PSI CSE methodology are discussed.<br />

DE 8 5:45 PM<br />

SENS1D: A New Sensitivity And Uncertainty Analysis Code<br />

Wenming Wang,Haicheng Wu<br />

China Nuclear Data Center, China Institute of Atomic Energy<br />

As the development of modern reactor technology, the need of safety and reliability plays an increasing role<br />

of reactor design and maintainance, which calls for more precise nuclear data and more accurate validation<br />

of uncertainty analysis. In order to quantify the design uncertainty of key integral parameters due to<br />

input fundamental nuclear data, the perturbation method was applied to the transport process, basing<br />

on which many codes coupled with different transport system were developed from late 1970s, such as<br />

FORSS system, SUSD3D and TSUNAMI in the latest SCALE 6 system. In China Nuclear Data Center,<br />

written in Fortran95, a new S/U analysis code SENS1D was developed, coupled with XSDRN and ANISN<br />

basing on 1st order perturbation theory and MCNP with the perturbation card method. SENS1D has the<br />

ability to analyze the sensitivity and uncertainty of keff by nuclear and reaction, which includes all main<br />

reaction channel such as total, fission, absorption (all neutron disappearing channel), scattering (elastic<br />

and in-elastic), nu-bar and fission spectrum. As testing case, SENS1D were implemented into several 1-<br />

D benchmark facilities, such as Jezebel, Godiva and Flattop. The group constants were generated from<br />

CENDL3.1 and ENDF/B-VII, with the covariance data from JENDL4. Both XSDRN and ANISN were<br />

used to offer neutron flux information. Coupled with MCNP code, SENS1D was also used for the S/U<br />

73

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!