Bias in calculated k{sub eff} from subcritical measurements by the {sup 252}Cf-source-driven noise analysis method

Abstract

The development of MCNP-DSP, which allows direct calculation of the measured time and frequency analysis parameters from subcritical measurements using the {sup 252}Cf-source-driven noise analysis method, permits the validation of calculational methods for criticality safety with in-plant subcritical measurements. In addition, a method of obtaining the bias in the calculations, which is essential to the criticality safety specialist, is illustrated using the results of measurements with 17.771-cm-diam, enriched (93.15), unreflected, and unmoderated uranium metal cylinders. For these uranium metal cylinders the bias obtained using MCNP-DSP and ENDF/B-V cross-section data increased with subcriticality. For a critical experiment [height (h) = 12.629 cm], it was {minus}0.0061 {+-} 0.0003. For a 10.16-cm-high cylinder (k {approx} 0.93), it was 0.0060 {+-} 0.0016, and for a subcritical cylinder (h = 8.13 cm, k {approx} 0.85), the bias was {minus}0.0137 {+-} 0.0037, more than a factor of 2 larger in magnitude. This method allows the nuclear criticality safety specialist to establish the bias in calculational methods for criticality safety from in-plant subcritical measurements by the {sup 252}Cf-source-driven noise analysis method.