An Approach to Determining the Uncertainty in Reactor Test Objects Using Deterministic and Monte Carlo Methods

Abstract A procedure for determining an estimate of the uncertainty in test objects components using discrete ordinates and Monte Carlo methods has been developed. The procedure is described, and a review of the advantages and disadvantages of these methods is presented with the emphasis on the advantage of employing both transport methods in analysis. The discrete ordinates-based techniques use the DOORS and DANTSYS radiation transport codes (with the SUSD3D code for the determination of sensitivity of the response parameter to cross sections), and the Monte Carlo-based techniques primarily employ the MCNP code (using the general MCNP perturbation operator in a manner that determines the sensitivities to transport cross sections). The focus of the analysis presented is on experiments that are preformed at the Annular Core Research Reactor (ACRR) at Sandia National Laboratories (SNL), and includes uncertainty contributed by the uncertainties in the source spectrum, in response functions, and in transport cross sections.