Cross section adjustment method based on random sampling technique

Abstract

A cross section adjustment method based on the random sampling technique is proposed. In the proposed method, correlations among cross sections and core parameters are used instead of sensitivity coefficients of cross sections, which are necessary in the conventional method. The correlations are statistically estimated by the random sampling technique. The proposed method is theoretically consistent with the conventional method and provides comparable adjusted cross sections when sufficient number of random sampling is taken into account. The proposed method would be suitable for practical light water reactor (LWR) core analysis since estimation of sensitivity coefficients, which requires considerable computational cost in typical LWR problems, is not necessary. Through a benchmark problem in simple pin-cell geometry, adjusted cross sections by the present and the conventional cross section adjustment method are compared. The adjusted cross sections by the present method well reproduce the conventional ones, thus the feasibility of the present method is confirmed.

Confidence interval estimation by bootstrap method for uncertainty quantification using random sampling method

“…Note that the bootstrap method has capability to estimate confidence intervals for other sample estimates: not only the variance for one kind of target parameter but also the covariance (or the correlation coefficient) between two different kinds of target parameters. Moreover, even in the case of the cross-section adjustment method using the RS method [11], the bootstrap method is applicable to estimate the confidence intervals for the adjusted cross-section and covariance data.…”

“…The square root matrix A is numerically calculated by the singular value decomposition (SVD) for [9,11]. Namely, using the SVD, a real symmetric matrix can be decomposed as follows:…”

“…covariance data) of cross-section data [3][4][5][6] could be a major factor. In order to estimate the uncertainties of target safety parameters due to covariance data of cross-section, the RS method is one of the simple and effective methodologies, since a sophisticated sensitivity analysis code, such as a generalized adjoint analysis code [7,8], is not necessary in the RS method [9][10][11]. Especially for light water reactor (LWR) analysis, which employs a complicated multi-step calculation (lattice-core calculation) considering burnup and thermal-hydraulics feedback effects, the RS method is a practical approach to quantify uncertainties of target parameters due to cross-section covariance data.…”

Recent activities in the field of reactor physics

“…Those subjects have been well studied, especially for fast reactors (FRs) [20] followed by recent research works on light water reactors [21][22][23] and accelerator-driven systems (ADSs) [24]. With recent improvements in the computing environment, new approaches for sensitivity analysis and cross section adjustments have been studied; stochastic methods such as the "Total Monte Carlo" or "Random Sampling" approach have been proposed for adjustment of cross section data [25,26] and uncertainty analysis [27]. A reduced-order modeling based on the subspace method [28] as a "generalized perturbation theory (GPT) free" approach has been proposed for the problems where GPT does not work because many responses are required.…”

Bias factor method using random sampling technique

“…analysis, the authors proposed the cross-section adjustment technique on the basis of the random sampling (RS) technique [6,7]. Generally, a complicated twostep calculation scheme (e.g.…”